| 1 | Guile NEWS --- history of user-visible changes. |
| 2 | Copyright (C) 1996-2015 Free Software Foundation, Inc. |
| 3 | See the end for copying conditions. |
| 4 | |
| 5 | Please send Guile bug reports to bug-guile@gnu.org. |
| 6 | |
| 7 | |
| 8 | \f |
| 9 | Changes in 2.1.1 (changes since the 2.0.x series): |
| 10 | |
| 11 | * Notable changes |
| 12 | |
| 13 | ** Speed |
| 14 | |
| 15 | The biggest change in Guile 2.2 is a complete rewrite of its virtual |
| 16 | machine and compiler internals. The result is faster startup time, |
| 17 | better memory usage, and faster execution of user code. See the |
| 18 | "Performance improvements" section below for more details. |
| 19 | |
| 20 | ** Better thread-safety |
| 21 | |
| 22 | This new release series takes the ABI-break opportunity to fix some |
| 23 | interfaces that were difficult to use correctly from multiple threads. |
| 24 | Notably, weak hash tables are now transparently thread-safe. Ports are |
| 25 | also thread-safe; see "New interfaces" below for details on the changes |
| 26 | to the C interface. |
| 27 | |
| 28 | ** Better space-safety |
| 29 | |
| 30 | It used to be the case that, when calling a Scheme procedure, the |
| 31 | procedure and arguments were always preserved against garbage |
| 32 | collection. This is no longer the case; Guile is free to collect the |
| 33 | procedure and arguments if they become unreachable, or to re-use their |
| 34 | slots for other local variables. Guile still offers good-quality |
| 35 | backtraces by determining the procedure being called from the |
| 36 | instruction pointer instead of from the value in slot 0 of an |
| 37 | application frame, and by using a live variable map that allows the |
| 38 | debugger to know which locals are live at all points in a frame. |
| 39 | |
| 40 | ** Off-main-thread finalization |
| 41 | |
| 42 | Following Guile 2.0.6's change to invoke finalizers via asyncs, Guile |
| 43 | 2.2 takes the additional step of invoking finalizers from a dedicated |
| 44 | finalizer thread, if threads are enabled. This avoids concurrency |
| 45 | issues between finalizers and application code, and also speeds up |
| 46 | finalization. If your application's finalizers are not robust to the |
| 47 | presence of threads, see "Foreign Objects" in the manual for information |
| 48 | on how to disable automatic finalization and instead run finalizers |
| 49 | manually. |
| 50 | |
| 51 | ** Better locale support in Guile scripts |
| 52 | |
| 53 | When Guile is invoked directly, either from the command line or via a |
| 54 | hash-bang line (e.g. "#!/usr/bin/guile"), it now installs the current |
| 55 | locale via a call to `(setlocale LC_ALL "")'. For users with a unicode |
| 56 | locale, this makes all ports unicode-capable by default, without the |
| 57 | need to call `setlocale' in your program. This behavior may be |
| 58 | controlled via the GUILE_INSTALL_LOCALE environment variable; see the |
| 59 | manual for more. |
| 60 | |
| 61 | ** Complete Emacs-compatible Elisp implementation |
| 62 | |
| 63 | Thanks to the work of BT Templeton, Guile's Elisp implementation is now |
| 64 | fully Emacs-compatible, implementing all of Elisp's features and quirks |
| 65 | in the same way as the editor we know and love. |
| 66 | |
| 67 | ** Dynamically expandable stacks |
| 68 | |
| 69 | Instead of allocating fixed stack sizes for running Scheme code, Guile |
| 70 | now starts off each thread with only one page of stack, and expands and |
| 71 | shrinks it dynamically as needed. Guile will throw an exception for |
| 72 | stack overflows if growing the stack fails. It is also possible to |
| 73 | impose a stack limit during the extent of a function call. See "Stack |
| 74 | Overflow" in the manual, for more. |
| 75 | |
| 76 | This change allows users to write programs that use the stack as a data |
| 77 | structure for pending computations, as it was meant to be, without |
| 78 | reifying that data out to the heap. Where you would previously make a |
| 79 | loop that collect its results in reverse order only to re-reverse them |
| 80 | at the end, now you can just recurse without worrying about stack |
| 81 | overflows. |
| 82 | |
| 83 | Using the stack also allows more code to be continuation-safe. For |
| 84 | example, returning multiple times from a `map' procedure in Guile 2.0 |
| 85 | would change the value of previously returned result lists, because |
| 86 | `map' built its result list in reverse order then used `reverse!' to |
| 87 | return the proper result. Now in Guile 2.2, `map' is implemented using |
| 88 | straightforward recursion, which eliminates this bug while maintaining |
| 89 | good performance as well as good space complexity. |
| 90 | |
| 91 | ** Out-of-memory improvements |
| 92 | |
| 93 | Instead of aborting, failures to allocate memory will now raise an |
| 94 | unwind-only `out-of-memory' exception, and cause the corresponding |
| 95 | `catch' expression to run garbage collection in order to free up memory. |
| 96 | |
| 97 | ** GOOPS core reimplemented in Scheme |
| 98 | |
| 99 | Guile's object orientation system, GOOPS, has been mostly reimplemented |
| 100 | in Scheme. This decreases its maintenance burden on the rest of Guile, |
| 101 | while also makes it possible to implement new features in the future, |
| 102 | such as method combinations or `eqv?' specializers. |
| 103 | |
| 104 | * Performance improvements |
| 105 | |
| 106 | ** Faster programs via new virtual machine |
| 107 | |
| 108 | Guile now compiles programs to instructions for a new virtual machine. |
| 109 | The new virtual machine's instructions can address their source and |
| 110 | destination operands by "name" (slot). This makes access to named |
| 111 | temporary values much faster than in Guile 2.0, and removes a lot of |
| 112 | value-shuffling that the old virtual machine had to do. The end result |
| 113 | is that loop-heavy code can be two or three times as fast with Guile 2.2 |
| 114 | as in 2.0. Your mileage may vary, of course; see "A Virtual Machine for |
| 115 | Guile" in the manual for the nitties and the gritties. |
| 116 | |
| 117 | ** Better startup time, memory usage with ELF object file format |
| 118 | |
| 119 | Guile now uses the standard ELF format for its compiled code. (Guile |
| 120 | has its own loader and linker, so this does not imply a dependency on |
| 121 | any particular platform's ELF toolchain.) The benefit is that Guile is |
| 122 | now able to statically allocate more data in the object files. ELF also |
| 123 | enables more sharing of data between processes, and decreases startup |
| 124 | time (about 40% faster than the already fast startup of the Guile 2.0 |
| 125 | series). Guile also uses DWARF for some of its debugging information. |
| 126 | Much of the debugging information can be stripped from the object files |
| 127 | as well. See "Object File Format" in the manual, for full details. |
| 128 | |
| 129 | ** Better optimizations via compiler rewrite |
| 130 | |
| 131 | Guile's compiler now uses a Continuation-Passing Style (CPS) |
| 132 | intermediate language, allowing it to reason easily about temporary |
| 133 | values and control flow. Examples of optimizations that this permits |
| 134 | are optimal contification, optimal common subexpression elimination, |
| 135 | dead code elimination, parallel moves with at most one temporary, |
| 136 | allocation of stack slots using precise liveness information, and |
| 137 | closure optimization. For more, see "Continuation-Passing Style" in the |
| 138 | manual. |
| 139 | |
| 140 | ** Faster interpreter |
| 141 | |
| 142 | Combined with a number of optimizations to the interpreter itself, |
| 143 | simply compiling `eval.scm' with the new compiler yields an interpreter |
| 144 | that is consistently two or three times faster than the one in Guile |
| 145 | 2.0. |
| 146 | |
| 147 | ** Allocation-free dynamic stack |
| 148 | |
| 149 | Guile now implements the dynamic stack with an actual stack instead of a |
| 150 | list of heap objects, avoiding most allocation. This speeds up prompts, |
| 151 | the `scm_dynwind_*' family of functions, fluids, and `dynamic-wind'. |
| 152 | |
| 153 | ** Optimized UTF-8 and Latin-1 ports, symbols, and strings |
| 154 | |
| 155 | Guile 2.2 is faster at reading and writing UTF-8 and Latin-1 strings |
| 156 | from ports, and at converting symbols and strings to and from these |
| 157 | encodings. |
| 158 | |
| 159 | ** Optimized hash functions |
| 160 | |
| 161 | Guile 2.2 now uses Bob Jenkins' `hashword2' (from his `lookup3.c') for |
| 162 | its string hash, and Thomas Wang's integer hash function for `hashq' and |
| 163 | `hashv'. These functions produce much better hash values across all |
| 164 | available fixnum bits. |
| 165 | |
| 166 | ** Optimized generic array facility |
| 167 | |
| 168 | Thanks to work by Daniel Llorens, the generic array facility is much |
| 169 | faster now, as it is internally better able to dispatch on the type of |
| 170 | the underlying backing store. |
| 171 | |
| 172 | * New interfaces |
| 173 | |
| 174 | ** New `cond-expand' feature: `guile-2.2' |
| 175 | |
| 176 | Use this feature if you need to check for Guile 2.2 from Scheme code. |
| 177 | |
| 178 | ** New predicate: `nil?' |
| 179 | |
| 180 | See "Nil" in the manual. |
| 181 | |
| 182 | ** New compiler modules |
| 183 | |
| 184 | Since the compiler was rewritten, there are new modules for the back-end |
| 185 | of the compiler and the low-level loader and introspection interfaces. |
| 186 | See the "Guile Implementation" chapter in the manual for all details. |
| 187 | |
| 188 | ** New functions: `scm_to_intptr_t', `scm_from_intptr_t' |
| 189 | ** New functions: `scm_to_uintptr_t', `scm_from_uintptr_t' |
| 190 | |
| 191 | See "Integers" in the manual, for more. |
| 192 | |
| 193 | ** New thread-safe port API |
| 194 | |
| 195 | For details on `scm_c_make_port', `scm_c_make_port_with_encoding', |
| 196 | `scm_c_lock_port', `scm_c_try_lock_port', `scm_c_unlock_port', |
| 197 | `scm_c_port_type_ref', `scm_c_port_type_add_x', `SCM_PORT_DESCRIPTOR', |
| 198 | and `scm_dynwind_lock_port', see XXX. |
| 199 | |
| 200 | There is now a routine to atomically adjust port "revealed counts". See |
| 201 | XXX for more on `scm_adjust_port_revealed_x' and |
| 202 | `adjust-port-revealed!', |
| 203 | |
| 204 | All other port API now takes the lock on the port if needed. There are |
| 205 | some C interfaces if you know that you don't need to take a lock; see |
| 206 | XXX for details on `scm_get_byte_or_eof_unlocked', |
| 207 | `scm_peek_byte_or_eof_unlocked' `scm_c_read_unlocked', |
| 208 | `scm_getc_unlocked' `scm_unget_byte_unlocked', `scm_ungetc_unlocked', |
| 209 | `scm_ungets_unlocked', `scm_fill_input_unlocked' `scm_putc_unlocked', |
| 210 | `scm_puts_unlocked', and `scm_lfwrite_unlocked'. |
| 211 | |
| 212 | ** New inline functions: `scm_new_smob', `scm_new_double_smob' |
| 213 | |
| 214 | These can replace many uses of SCM_NEWSMOB, SCM_RETURN_NEWSMOB2, and the |
| 215 | like. See XXX in the manual, for more. |
| 216 | |
| 217 | ** New low-level type accessors |
| 218 | |
| 219 | For more on `SCM_HAS_TYP7', `SCM_HAS_TYP7S', `SCM_HAS_TYP16', see XXX. |
| 220 | |
| 221 | `SCM_HEAP_OBJECT_P' is now an alias for the inscrutable `SCM_NIMP'. |
| 222 | |
| 223 | `SCM_UNPACK_POINTER' and `SCM_PACK_POINTER' are better-named versions of |
| 224 | the old `SCM2PTR' and `PTR2SCM'. Also, `SCM_UNPACK_POINTER' yields a |
| 225 | void*. |
| 226 | |
| 227 | ** <standard-vtable>, standard-vtable-fields |
| 228 | |
| 229 | See "Structures" in the manual for more on these |
| 230 | |
| 231 | ** Convenience utilities for ports and strings. |
| 232 | |
| 233 | See XXX for more on `scm_from_port_string', `scm_from_port_stringn', |
| 234 | `scm_to_port_string', and `scm_to_port_stringn'. |
| 235 | |
| 236 | ** New expressive PEG parser |
| 237 | |
| 238 | See "PEG Parsing" in the manual for more. Thanks to Michael Lucy for |
| 239 | originally writing these, and to Noah Lavine for integration work. |
| 240 | |
| 241 | ** `make-stack' now also works on delimited continuations |
| 242 | |
| 243 | ** Better URI-reference support |
| 244 | |
| 245 | The `(web uri)' module now has interfaces for handling URI references, |
| 246 | which might not have a scheme. The Location header of a web request or |
| 247 | response is now a URI reference instead of a URI. Also, |
| 248 | `request-absolute-uri' now has an optional default scheme argument. See |
| 249 | "Web" in the manual for full details. |
| 250 | |
| 251 | ** formal-name->char, char->formal-name |
| 252 | |
| 253 | See "Characters", in the manual. |
| 254 | |
| 255 | * Incompatible changes |
| 256 | |
| 257 | ** ASCII is not ISO-8859-1 |
| 258 | |
| 259 | In Guile 2.0, if a user set "ASCII" or "ANSI_X3.4-1968" as the encoding |
| 260 | of a port, Guile would treat it as ISO-8859-1. While these encodings |
| 261 | are the same for codepoints 0 to 127, ASCII does not extend past that |
| 262 | range, whereas ISO-8859-1 goes up to 255. Guile 2.2 no longer treats |
| 263 | ASCII as ISO-8859-1. This is likely to be a problem only if the user's |
| 264 | locale is set to ASCII, and the user or a program writes non-ASCII |
| 265 | codepoints to a port. |
| 266 | |
| 267 | ** String ports default to UTF-8 |
| 268 | |
| 269 | Guile 2.0 would use the `%default-port-encoding' when creating string |
| 270 | ports. This resulted in ports that could only accept a subset of valid |
| 271 | characters, which was surprising to users. Now string ports default to |
| 272 | the UTF-8 encoding. Sneaky users can still play encoding conversion |
| 273 | games with string ports by explicitly setting the encoding of a port |
| 274 | after it is open. See "Ports" in the manual for more. |
| 275 | |
| 276 | ** `scm_from_stringn' and `scm_to_stringn' encoding arguments are never NULL |
| 277 | |
| 278 | These functions now require a valid `encoding' argument, and will abort |
| 279 | if given `NULL'. |
| 280 | |
| 281 | ** All r6rs ports are both textual and binary |
| 282 | |
| 283 | Because R6RS ports are a thin layer on top of Guile's ports, and Guile's |
| 284 | ports are both textual and binary, Guile's R6RS ports are also both |
| 285 | textual and binary, and thus both kinds have port transcoders. This is |
| 286 | an incompatibility with respect to R6RS. |
| 287 | |
| 288 | ** Vtable hierarchy changes |
| 289 | |
| 290 | In an attempt to make Guile's structure and record types integrate |
| 291 | better with GOOPS by unifying the vtable hierarchy, `make-vtable-vtable' |
| 292 | is now deprecated. Instead, users should just use `make-vtable' with |
| 293 | appropriate arguments. See "Structures" in the manual for all of the |
| 294 | details. As such, `record-type-vtable' and `%condition-type-vtable' now |
| 295 | have a parent vtable and are no longer roots of the vtable hierarchy. |
| 296 | |
| 297 | ** Syntax parameters are a distinct type |
| 298 | |
| 299 | Guile 2.0's transitional implementation of `syntax-parameterize' was |
| 300 | based on the `fluid-let-syntax' interface inherited from the psyntax |
| 301 | expander. This interface allowed any binding to be dynamically rebound |
| 302 | -- even bindings like `lambda'. This is no longer the case in Guile |
| 303 | 2.2. Syntax parameters must be defined via `define-syntax-parameter', |
| 304 | and only such bindings may be parameterized. See "Syntax Parameters" in |
| 305 | the manual for more. |
| 306 | |
| 307 | ** Defined identifiers scoped in the current module |
| 308 | |
| 309 | Sometimes Guile's expander would attach incorrect module scoping |
| 310 | information for top-level bindings made by an expansion. For example, |
| 311 | given the following R6RS library: |
| 312 | |
| 313 | (library (defconst) |
| 314 | (export defconst) |
| 315 | (import (guile)) |
| 316 | (define-syntax-rule (defconst name val) |
| 317 | (begin |
| 318 | (define t val) |
| 319 | (define-syntax-rule (name) t)))) |
| 320 | |
| 321 | Attempting to use it would produce an error: |
| 322 | |
| 323 | (import (defconst)) |
| 324 | (defconst foo 42) |
| 325 | (foo) |
| 326 | =| Unbound variable: t |
| 327 | |
| 328 | It wasn't clear that we could fix this in Guile 2.0 without breaking |
| 329 | someone's delicate macros, so the fix is only coming out now. |
| 330 | |
| 331 | ** Pseudo-hygienically rename macro-introduced bindings |
| 332 | |
| 333 | Bindings introduced by macros, like `t' in the `defconst' example above, |
| 334 | are now given pseudo-fresh names. This allows |
| 335 | |
| 336 | (defconst foo 42) |
| 337 | (defconst bar 37) |
| 338 | |
| 339 | to introduce different bindings for `t'. These pseudo-fresh names are |
| 340 | made in such a way that if the macro is expanded again, for example as |
| 341 | part of a simple recompilation, the introduced identifiers get the same |
| 342 | pseudo-fresh names. See "Hygiene and the Top-Level" in the manual, for |
| 343 | details. |
| 344 | |
| 345 | ** Fix literal matching for module-bound literals |
| 346 | |
| 347 | `syntax-rules' and `syntax-case' macros can take a set of "literals": |
| 348 | bound or unbound keywords that the syntax matcher treats specially. |
| 349 | Before, literals were always matched symbolically (by name). Now they |
| 350 | are matched by binding. This allows literals to be reliably bound to |
| 351 | values, renamed by imports or exports, et cetera. See "Syntax-rules |
| 352 | Macros" in the manual for more on literals. |
| 353 | |
| 354 | ** `dynamic-wind' doesn't check that guards are thunks |
| 355 | |
| 356 | Checking that the dynamic-wind out-guard procedure was actually a thunk |
| 357 | before doing the wind was slow, unreliable, and not strictly needed. |
| 358 | |
| 359 | ** All deprecated code removed |
| 360 | |
| 361 | All code deprecated in Guile 2.0 has been removed. See older NEWS, and |
| 362 | check that your programs can compile without linker warnings and run |
| 363 | without runtime warnings. See "Deprecation" in the manual. |
| 364 | |
| 365 | ** Remove miscellaneous unused interfaces |
| 366 | |
| 367 | We have removed accidentally public, undocumented interfaces that we |
| 368 | think are not used, and not useful. This includes `scm_markstream', |
| 369 | `SCM_FLUSH_REGISTER_WINDOWS', `SCM_THREAD_SWITCHING_CODE', `SCM_FENCE', |
| 370 | `scm_call_generic_0', `scm_call_generic_1', `scm_call_generic_2' |
| 371 | `scm_call_generic_3', `scm_apply_generic', and `scm_program_source'. |
| 372 | `scm_async_click' was renamed to `scm_async_tick', and `SCM_ASYNC_TICK' |
| 373 | was made private (use `SCM_TICK' instead). |
| 374 | |
| 375 | ** Many internal compiler / VM changes |
| 376 | |
| 377 | As the compiler and virtual machine were re-written, there are many |
| 378 | changes in the back-end of Guile to interfaces that were introduced in |
| 379 | Guile 2.0. These changes are only only of interest if you wrote a |
| 380 | language on Guile 2.0 or a tool using Guile 2.0 internals. If this is |
| 381 | the case, drop by the IRC channel to discuss the changes. |
| 382 | |
| 383 | ** Defining a SMOB or port type no longer mucks exports of `(oop goops)' |
| 384 | |
| 385 | It used to be that defining a SMOB or port type added an export to |
| 386 | GOOPS, for the wrapper class of the smob type. This violated |
| 387 | modularity, though, so we have removed this behavior. |
| 388 | |
| 389 | ** Bytecode replaces objcode as a target language |
| 390 | |
| 391 | One way in which people may have used details of Guile's runtime in |
| 392 | Guile 2.0 is in compiling code to thunks for later invocation. Instead |
| 393 | of compiling to objcode and then calling `make-program', now the way to |
| 394 | do it is to compile to `bytecode' and then call `load-thunk-from-memory' |
| 395 | from `(system vm loader)'. |
| 396 | |
| 397 | ** Weak pairs removed |
| 398 | |
| 399 | Weak pairs were not safe to access with `car' and `cdr', and so were |
| 400 | removed. |
| 401 | |
| 402 | ** Weak alist vectors removed |
| 403 | |
| 404 | Use weak hash tables instead. |
| 405 | |
| 406 | ** Weak vectors may no longer be accessed via `vector-ref' et al |
| 407 | |
| 408 | Weak vectors may no longer be accessed with the vector interface. This |
| 409 | was a source of bugs in the 2.0 Guile implementation, and a limitation |
| 410 | on using vectors as building blocks for other abstractions. Vectors in |
| 411 | Guile are now a concrete type; for an abstract interface, use the |
| 412 | generic array facility (`array-ref' et al). |
| 413 | |
| 414 | ** scm_t_array_implementation removed |
| 415 | |
| 416 | This interface was introduced in 2.0 but never documented. It was a |
| 417 | failed attempt to layer the array implementation that actually |
| 418 | introduced too many layers, as it prevented the "vref" and "vset" |
| 419 | members of scm_t_array_handle (called "ref" and "set" in 1.8, not |
| 420 | present in 2.0) from specializing on array backing stores. |
| 421 | |
| 422 | Notably, the definition of scm_t_array_handle has now changed, to not |
| 423 | include the (undocumented) "impl" member. We are sorry for any |
| 424 | inconvenience this may cause. |
| 425 | |
| 426 | ** `scm_make' is now equivalent to Scheme `make' |
| 427 | |
| 428 | It used to be that `scm_make' only implemented a hard-wired object |
| 429 | allocation and initialization protocol. This was because `scm_make' was |
| 430 | used while GOOPS booted its own, more complete `make' implementation in |
| 431 | Scheme. Now that we've re-implemented everything in Scheme, the C |
| 432 | `scm_make' now dispatches directly to Scheme `make', which implements |
| 433 | the full protocol. This change is incompatible in some ways, but on the |
| 434 | whole is good news for GOOPS users. |
| 435 | |
| 436 | ** GOOPS slot definitions are now objects |
| 437 | |
| 438 | Slot definitions are now instances of a <slot> class, instead of being |
| 439 | specially formatted lists. To most user code, this is transparent, as |
| 440 | the slot definition accessors like `slot-definition-name' continue to |
| 441 | work. However, code that for example uses `car' to get the name of a |
| 442 | slot definition will need to be updated to use the accessors. |
| 443 | |
| 444 | ** Class slot changes |
| 445 | |
| 446 | Class objects no longer have a `default-slot-definition-class' slot, |
| 447 | which was never used. They also no longer have slots for hashsets |
| 448 | (`h0', `h1', and so on up to `h7'), which have been unused since Guile |
| 449 | 2.0 and were not a great idea. |
| 450 | |
| 451 | There is a new class option, `#:static-slot-allocation?'. See the |
| 452 | manual for details. |
| 453 | |
| 454 | ** Removal of internal, unintentionally exposed GOOPS C interfaces |
| 455 | |
| 456 | These include: `scm_sys_fast_slot_ref', `scm_sys_fast_slot_set_x' |
| 457 | `scm_basic_basic_make_class', `scm_sys_compute_slots', |
| 458 | `scm_sys_prep_layout_x' `scm_t_method', `SCM_METHOD', |
| 459 | `scm_s_slot_set_x', `SCM_CLASS_CLASS_LAYOUT', `scm_si_slotdef_class', |
| 460 | `scm_si_generic_function', `scm_si_specializers', `scm_si_procedure', |
| 461 | `scm_si_formals', `scm_si_body', `scm_si_make_procedure', |
| 462 | `SCM_CLASS_CLASS_LAYOUT', `SCM_INSTANCE_HASH', `SCM_SET_HASHSET', `union |
| 463 | scm_t_debug_info', `scm_pure_generic_p', `SCM_PUREGENERICP', |
| 464 | `SCM_VALIDATE_PUREGENERIC', `SCM_VTABLE_FLAG_GOOPS_PURE_GENERIC', |
| 465 | `SCM_CLASSF_PURE_GENERIC', `scm_c_extend_primitive_generic', |
| 466 | `scm_sys_initialize_object', `SCM_CLASS_CLASS_LAYOUT', |
| 467 | `scm_si_redefined', `scm_si_direct_supers', `scm_si_direct_slots', |
| 468 | `scm_si_direct_subclasses', `scm_si_direct_methods', `scm_si_cpl' |
| 469 | `scm_si_slots', `scm_si_getters_n_setters', `SCM_N_CLASS_SLOTS', |
| 470 | `SCM_OBJ_CLASS_REDEF', `SCM_INST', `SCM_ACCESSORS_OF', |
| 471 | `scm_sys_allocate_instance', and `scm_sys_invalidate_class_x'. |
| 472 | |
| 473 | * New deprecations |
| 474 | |
| 475 | ** SCM_WTA_DISPATCH_0, SCM_WTA_DISPATCH_1, SCM_WTA_DISPATCH_2, SCM_WTA_DISPATCH_N |
| 476 | ** SCM_GASSERT0, SCM_GASSERT1, SCM_GASSERT2, SCM_GASSERTn |
| 477 | ** SCM_WTA_DISPATCH_1_SUBR |
| 478 | |
| 479 | These macros were used in dispatching primitive generics. They can be |
| 480 | replaced by using C functions (the same name but in lower case), if |
| 481 | needed, but this is a hairy part of Guile that perhaps you shouldn't be |
| 482 | using. |
| 483 | |
| 484 | ** scm_compute_applicable_methods and scm_find_method |
| 485 | |
| 486 | Use `compute-applicable-methods' from Scheme instead. |
| 487 | |
| 488 | ** scm_no_applicable_method |
| 489 | |
| 490 | Fetch no-applicable-method from the GOOPS exports if you need it. |
| 491 | |
| 492 | ** scm_class_boolean, scm_class_char, scm_class_pair |
| 493 | ** scm_class_procedure, scm_class_string, scm_class_symbol |
| 494 | ** scm_class_primitive_generic, scm_class_vector, scm_class_null |
| 495 | ** scm_class_real, scm_class_complex, scm_class_integer |
| 496 | ** scm_class_fraction, scm_class_unknown, scm_class_top |
| 497 | ** scm_class_object, scm_class_class, scm_class_applicable |
| 498 | ** scm_class_applicable_struct, scm_class_applicable_struct_with_setter |
| 499 | ** scm_class_generic, scm_class_generic_with_setter, scm_class_accessor |
| 500 | ** scm_class_extended_generic, scm_class_extended_generic_with_setter |
| 501 | ** scm_class_extended_accessor, scm_class_method |
| 502 | ** scm_class_accessor_method, scm_class_procedure_class |
| 503 | ** scm_class_applicable_struct_class, scm_class_number, scm_class_list |
| 504 | ** scm_class_keyword, scm_class_port, scm_class_input_output_port |
| 505 | ** scm_class_input_port, scm_class_output_port, scm_class_foreign_slot |
| 506 | ** scm_class_self, scm_class_protected, scm_class_hidden |
| 507 | ** scm_class_opaque, scm_class_read_only, scm_class_protected_hidden |
| 508 | ** scm_class_protected_opaque, scm_class_protected_read_only |
| 509 | ** scm_class_scm, scm_class_int, scm_class_float, scm_class_double |
| 510 | ** scm_port_class, scm_smob_class |
| 511 | |
| 512 | These class exports are now deprecated. Instead, look up the ones you |
| 513 | need from the GOOPS module, or use `scm_class_of' on particular values. |
| 514 | |
| 515 | ** scm_get_keyword |
| 516 | |
| 517 | Instead from Scheme use kw-arg-ref or real keyword arguments, and from C |
| 518 | use `scm_c_bind_keyword_arguments'. |
| 519 | |
| 520 | ** scm_slot_ref_using_class, scm_slot_set_using_class_x |
| 521 | ** scm_slot_bound_using_class_p, scm_slot_exists_using_class_p |
| 522 | |
| 523 | Instead use the normal `scm_slot_ref' and similar procedures. |
| 524 | |
| 525 | * Changes to the distribution |
| 526 | |
| 527 | ** New minor version |
| 528 | |
| 529 | The "effective version" of Guile is now 2.2, which allows parallel |
| 530 | installation with other effective versions (for example, the older Guile |
| 531 | 2.0). See "Parallel Installations" in the manual for full details. |
| 532 | Notably, the `pkg-config' file is now `guile-2.2'. |
| 533 | |
| 534 | ** Bump required libgc version to 7.2, released March 2012. |
| 535 | |
| 536 | ** The readline extension is now installed in the extensionsdir |
| 537 | |
| 538 | The shared library that implements Guile's readline extension is no |
| 539 | longer installed to the libdir. This change should be transparent to |
| 540 | users, but packagers may be interested. |
| 541 | |
| 542 | |
| 543 | \f |
| 544 | Changes in 2.0.11 (since 2.0.10): |
| 545 | |
| 546 | This release fixes an embarrassing regression introduced in the C |
| 547 | interface to SRFI-4 vectors. See |
| 548 | <https://lists.gnu.org/archive/html/guile-devel/2014-03/msg00047.html> |
| 549 | for details. |
| 550 | |
| 551 | \f |
| 552 | Changes in 2.0.10 (since 2.0.9): |
| 553 | |
| 554 | * Notable changes |
| 555 | |
| 556 | ** New GDB extension to support Guile |
| 557 | |
| 558 | Guile now comes with an extension for GDB 7.8 or later (unreleased at |
| 559 | the time of writing) that simplifies debugging of C code that uses |
| 560 | Guile. See "GDB Support" in the manual. |
| 561 | |
| 562 | ** Improved integration between R6RS and native Guile exceptions |
| 563 | |
| 564 | R6RS exception handlers, established using 'with-exception-handler' or |
| 565 | 'guard', are now able to catch native Guile exceptions, which are |
| 566 | automatically converted into appropriate R6RS condition objects. |
| 567 | |
| 568 | ** Support for HTTP proxies |
| 569 | |
| 570 | Guile's built-in web client now honors the 'http_proxy' environment |
| 571 | variable, as well as the new 'current-http-proxy' parameter. See |
| 572 | "Web Client" in the manual for details. |
| 573 | |
| 574 | ** Lexical syntax improvements |
| 575 | |
| 576 | *** Support |...| symbol notation. |
| 577 | |
| 578 | Guile's core reader and printer now support the R7RS |...| notation |
| 579 | for writing symbols with arbitrary characters, as a more portable and |
| 580 | attractive alternative to Guile's native #{...}# notation. To enable |
| 581 | this notation by default, put one or both of the following in your |
| 582 | ~/.guile: |
| 583 | |
| 584 | (read-enable 'r7rs-symbols) |
| 585 | (print-enable 'r7rs-symbols) |
| 586 | |
| 587 | *** Support '#true' and '#false' notation for booleans. |
| 588 | |
| 589 | The booleans '#t' and '#f' may now be written as '#true' and '#false' |
| 590 | for improved readability, per R7RS. |
| 591 | |
| 592 | *** Recognize '#\escape' character name. |
| 593 | |
| 594 | The escape character '#\esc' may now be written as '#\escape', per R7RS. |
| 595 | |
| 596 | *** Accept "\|" in string literals. |
| 597 | |
| 598 | The pipe character may now be preceded by a backslash, per R7RS. |
| 599 | |
| 600 | ** Custom binary input ports now support 'setvbuf'. |
| 601 | |
| 602 | Until now, ports returned by 'make-custom-binary-input-port' were always |
| 603 | full-buffered. Now, their buffering mode can be changed using 'setvbuf'. |
| 604 | |
| 605 | ** SRFI-4 predicates and length accessors no longer accept arrays. |
| 606 | |
| 607 | Given that the SRFI-4 accessors don't work for arrays, the fact that the |
| 608 | predicates and length accessors returned true for arrays was a bug. |
| 609 | |
| 610 | ** GUILE_PROGS now supports specifying a minimum required version. |
| 611 | |
| 612 | The 'GUILE_PROGS' autoconf macro in guile.m4 now allows an optional |
| 613 | argument to specify a minimum required Guile version. By default, it |
| 614 | requires Guile >= 2.0. A micro version can also be specified, e.g.: |
| 615 | GUILE_PROGS([2.0.10]) |
| 616 | |
| 617 | ** Error reporting improvements |
| 618 | |
| 619 | *** Improved run-time error reporting in (ice-9 match). |
| 620 | |
| 621 | If no pattern matches in a 'match' form, the datum that failed to match |
| 622 | is printed along with the location of the failed 'match' invocation. |
| 623 | |
| 624 | *** Print the faulty object upon invalid-keyword errors. |
| 625 | *** Improved error reporting of procedures defined by define-inlinable. |
| 626 | *** Improved error reporting for misplaced ellipses in macro definitions. |
| 627 | *** Improved error checking in 'define-public' and 'module-add!'. |
| 628 | *** Improved error when 'include' form with relative path is not in a file. |
| 629 | |
| 630 | ** Speed improvements |
| 631 | |
| 632 | *** 'scm_c_read' on ISO-8859-1 (e.g. binary) unbuffered ports is faster. |
| 633 | *** New inline asm for VM fixnum multiply, for faster overflow checking. |
| 634 | *** New inline asm for VM fixnum operations on ARM and 32-bit x86. |
| 635 | *** 'positive?' and 'negative?' are now compiled to VM primitives. |
| 636 | *** Numerical comparisons with more than 2 arguments are compiled to VM code. |
| 637 | *** Several R6RS bitwise operators have been optimized. |
| 638 | |
| 639 | ** Miscellaneous |
| 640 | |
| 641 | *** Web: 'content-disposition' headers are now supported. |
| 642 | *** Web: 'uri-encode' hexadecimal percent-encoding is now uppercase. |
| 643 | *** Size argument to 'make-doubly-weak-hash-table' is now optional. |
| 644 | *** Timeout for 'unlock-mutex' and SRFI-18 'mutex-unlock!' may now be #f. |
| 645 | |
| 646 | ** Gnulib update |
| 647 | |
| 648 | Guile's copy of Gnulib was updated to v0.1-92-g546ff82. The following |
| 649 | modules were imported from Gnulib: copysign, fsync, isfinite, link, |
| 650 | lstat, mkdir, mkstemp, readlink, rename, rmdir, and unistd. |
| 651 | |
| 652 | * New interfaces |
| 653 | |
| 654 | ** Cooperative REPL servers |
| 655 | |
| 656 | This new facility supports REPLs that run at specified times within an |
| 657 | existing thread, for example in programs utilizing an event loop or in |
| 658 | single-threaded programs. This allows for safe access and mutation of |
| 659 | a program's data structures from the REPL without concern for thread |
| 660 | synchronization. See "Cooperative REPL Servers" in the manual for |
| 661 | details. |
| 662 | |
| 663 | ** SRFI-43 (Vector Library) |
| 664 | |
| 665 | Guile now includes SRFI-43, a comprehensive library of vector operations |
| 666 | analogous to the SRFI-1 list library. See "SRFI-43" in the manual for |
| 667 | details. |
| 668 | |
| 669 | ** SRFI-64 (A Scheme API for test suites) |
| 670 | |
| 671 | Guile now includes SRFI-64, a flexible framework for creating test |
| 672 | suites. The reference implementation of SRFI-64 has also been updated |
| 673 | to fully support earlier versions of Guile. |
| 674 | |
| 675 | ** SRFI-111 (Boxes) |
| 676 | |
| 677 | See "SRFI-111" in the manual. |
| 678 | |
| 679 | ** 'define-values' |
| 680 | |
| 681 | See "Binding multiple return values" in the manual. |
| 682 | |
| 683 | ** Custom ellipsis identifiers using 'with-ellipsis' or SRFI-46. |
| 684 | |
| 685 | Guile now allows macro definitions to use identifiers other than '...' |
| 686 | as the ellipsis. This is convenient when writing macros that generate |
| 687 | macro definitions. The desired ellipsis identifier can be given as the |
| 688 | first operand to 'syntax-rules', as specified in SRFI-46 and R7RS, or by |
| 689 | using the new 'with-ellipsis' special form in procedural macros. With |
| 690 | this addition, Guile now fully supports SRFI-46. |
| 691 | |
| 692 | See "Specifying a Custom Ellipsis Identifier" and "Custom Ellipsis |
| 693 | Identifiers for syntax-case Macros" in the manual for details. |
| 694 | |
| 695 | ** R7RS 'syntax-error' |
| 696 | |
| 697 | Guile now supports 'syntax-error', as specified by R7RS, allowing for |
| 698 | improved compile-time error reporting from 'syntax-rules' macros. See |
| 699 | "Reporting Syntax Errors in Macros" in the manual for details. |
| 700 | |
| 701 | ** New procedures to convert association lists into hash tables |
| 702 | |
| 703 | Guile now includes the convenience procedures 'alist->hash-table', |
| 704 | 'alist->hashq-table', 'alist->hashv-table', and 'alist->hashx-table'. |
| 705 | See "Hash Table Reference" in the manual. |
| 706 | |
| 707 | ** New predicates: 'exact-integer?' and 'scm_is_exact_integer' |
| 708 | |
| 709 | See "Integers" in the manual. |
| 710 | |
| 711 | ** 'weak-vector-length', 'weak-vector-ref', and 'weak-vector-set!' |
| 712 | |
| 713 | These should now be used to access weak vectors, instead of |
| 714 | 'vector-length', 'vector-ref', and 'vector-set!'. |
| 715 | |
| 716 | * Manual updates |
| 717 | |
| 718 | ** Improve docs for 'eval-when'. |
| 719 | |
| 720 | Each 'eval-when' condition is now explained in detail, including |
| 721 | 'expand' which was previously undocumented. (expand load eval) is now |
| 722 | the recommended set of conditions, instead of (compile load eval). |
| 723 | See "Eval When" in the manual, for details. |
| 724 | |
| 725 | ** Update the section on SMOBs and memory management. |
| 726 | |
| 727 | See "Defining New Types (Smobs)" in the manual. |
| 728 | |
| 729 | ** Fixes |
| 730 | |
| 731 | *** GOOPS: #:dsupers is the init keyword for the dsupers slot. |
| 732 | *** 'unfold-right' takes a tail, not a tail generator. |
| 733 | *** Clarify that 'append!' and 'reverse!' might not mutate. |
| 734 | *** Fix doc that incorrectly claimed (integer? +inf.0) => #t. |
| 735 | (http://bugs.gnu.org/16356) |
| 736 | *** Document that we support SRFI-62 (S-expression comments). |
| 737 | *** Document that we support SRFI-87 (=> in case clauses). |
| 738 | *** Document 'equal?' in the list of R6RS incompatibilities. |
| 739 | *** Remove outdated documentation of LTDL_LIBRARY_PATH. |
| 740 | *** Fix 'weak-vector?' doc: Weak hash tables are not weak vectors. |
| 741 | *** Fix 'my-or' examples to use let-bound variable. |
| 742 | (http://bugs.gnu.org/14203) |
| 743 | |
| 744 | * New deprecations |
| 745 | |
| 746 | ** General 'uniform-vector' interface |
| 747 | |
| 748 | This interface lacked both generality and specificity. The general |
| 749 | replacements are 'array-length', 'array-ref', and friends on the scheme |
| 750 | side, and the array handle interface on the C side. On the specific |
| 751 | side of things, there are the specific bytevector, SRFI-4, and bitvector |
| 752 | interfaces. |
| 753 | |
| 754 | ** Use of the vector interface on arrays |
| 755 | ** 'vector-length', 'vector-ref', and 'vector-set!' on weak vectors |
| 756 | ** 'vector-length', 'vector-ref', and 'vector-set!' as primitive-generics |
| 757 | |
| 758 | Making the vector interface operate only on a single representation will |
| 759 | allow future versions of Guile to compile loops involving vectors to |
| 760 | more efficient native code. |
| 761 | |
| 762 | ** 'htons', 'htonl', 'ntohs', 'ntohl' |
| 763 | |
| 764 | These procedures, like their C counterpart, were used to convert numbers |
| 765 | to/from network byte order, typically in conjunction with the |
| 766 | now-deprecated uniform vector API. |
| 767 | |
| 768 | This functionality is now covered by the bytevector and binary I/O APIs. |
| 769 | See "Interpreting Bytevector Contents as Integers" in the manual. |
| 770 | |
| 771 | ** 'gc-live-object-stats' |
| 772 | |
| 773 | It hasn't worked in the whole 2.0 series. There is no replacement, |
| 774 | unfortunately. |
| 775 | |
| 776 | ** 'scm_c_program_source' |
| 777 | |
| 778 | This internal VM function was not meant to be public. Use |
| 779 | 'scm_procedure_source' instead. |
| 780 | |
| 781 | * Build fixes |
| 782 | |
| 783 | ** Fix build with Clang 3.4. |
| 784 | |
| 785 | ** MinGW build fixes |
| 786 | *** Do not add $(EXEEXT) to guild or guile-tools. |
| 787 | *** tests: Use double quotes around shell arguments, for Windows. |
| 788 | *** tests: Don't rely on $TMPDIR and /tmp on Windows. |
| 789 | *** tests: Skip FFI tests that use `qsort' when it's not accessible. |
| 790 | *** tests: Remove symlink only when it exists. |
| 791 | *** tests: Don't rely on `scm_call_2' being visible. |
| 792 | |
| 793 | ** Fix computation of LIBLOBJS so dependencies work properly. |
| 794 | (http://bugs.gnu.org/14193) |
| 795 | |
| 796 | * Bug fixes |
| 797 | |
| 798 | ** Web: Fix web client with methods other than GET. |
| 799 | (http://bugs.gnu.org/15908) |
| 800 | ** Web: Add Content-Length header for empty bodies. |
| 801 | ** Web: Accept "UTC" as the zone offset in date headers. |
| 802 | (http://bugs.gnu.org/14128) |
| 803 | ** Web: Don't throw if a response is longer than its Content-Length says. |
| 804 | ** Web: Write out HTTP Basic auth headers correctly. |
| 805 | (http://bugs.gnu.org/14370) |
| 806 | ** Web: Always print a path component in 'write-request-line'. |
| 807 | ** Fix 'define-public' from (ice-9 curried-definitions). |
| 808 | ** psyntax: toplevel variable definitions discard previous syntactic binding. |
| 809 | (http://bugs.gnu.org/11988) |
| 810 | ** Fix thread-unsafe lazy initializations. |
| 811 | ** Make (ice-9 popen) thread-safe. |
| 812 | (http://bugs.gnu.org/15683) |
| 813 | ** Make guardians thread-safe. |
| 814 | ** Make regexp_exec thread-safe. |
| 815 | (http://bugs.gnu.org/14404) |
| 816 | ** vm: Gracefully handle stack overflows. |
| 817 | (http://bugs.gnu.org/15065) |
| 818 | ** Fix 'rationalize'. |
| 819 | (http://bugs.gnu.org/14905) |
| 820 | ** Fix inline asm for VM fixnum operations on x32. |
| 821 | ** Fix 'SCM_SYSCALL' to really swallow EINTR. |
| 822 | ** Hide EINTR returns from 'accept'. |
| 823 | ** SRFI-19: Update the table of leap seconds. |
| 824 | ** Add missing files to the test-suite Makefile. |
| 825 | ** Make sure 'ftw' allows directory traversal when running as root. |
| 826 | ** Fix 'hash-for-each' for weak hash tables. |
| 827 | ** SRFI-18: Export 'current-thread'. |
| 828 | (http://bugs.gnu.org/16890) |
| 829 | ** Fix inlining of tail list to apply. |
| 830 | (http://bugs.gnu.org/15533) |
| 831 | ** Fix bug in remqueue in threads.c when removing last element. |
| 832 | ** Fix build when '>>' on negative integers is not arithmetic. |
| 833 | ** Fix 'bitwise-bit-count' for negative arguments. |
| 834 | (http://bugs.gnu.org/14864) |
| 835 | ** Fix VM 'ash' for right shifts by large amounts. |
| 836 | (http://bugs.gnu.org/14864) |
| 837 | ** Fix rounding in scm_i_divide2double for negative arguments. |
| 838 | ** Avoid lossy conversion from inum to double in numerical comparisons. |
| 839 | ** Fix numerical comparison of fractions to infinities. |
| 840 | ** Allow fl+ and fl* to accept zero arguments. |
| 841 | (http://bugs.gnu.org/14869) |
| 842 | ** flonum? returns false for complex number objects. |
| 843 | (http://bugs.gnu.org/14866) |
| 844 | ** flfinite? applied to a NaN returns false. |
| 845 | (http://bugs.gnu.org/14868) |
| 846 | ** Flonum operations always return flonums. |
| 847 | (http://bugs.gnu.org/14871) |
| 848 | ** min and max: NaNs beat infinities, per R6RS errata. |
| 849 | (http://bugs.gnu.org/14865) |
| 850 | ** Fix 'fxbit-count' for negative arguments. |
| 851 | ** 'gcd' and 'lcm' support inexact integer arguments. |
| 852 | (http://bugs.gnu.org/14870) |
| 853 | ** Fix R6RS 'fixnum-width'. |
| 854 | (http://bugs.gnu.org/14879) |
| 855 | ** tests: Use shell constructs that /bin/sh on Solaris 10 can understand. |
| 856 | (http://bugs.gnu.org/14042) |
| 857 | ** Fix display of symbols containing backslashes. |
| 858 | (http://bugs.gnu.org/15033) |
| 859 | ** Fix truncated-print for uniform vectors. |
| 860 | ** Define `AF_UNIX' only when Unix-domain sockets are supported. |
| 861 | ** Decompiler: fix handling of empty 'case-lambda' expressions. |
| 862 | ** Fix handling of signed zeroes and infinities in 'numerator' and 'denominator'. |
| 863 | ** dereference-pointer: check for null pointer. |
| 864 | ** Optimizer: Numerical comparisons are not negatable, for correct NaN handling. |
| 865 | ** Compiler: Evaluate '-' and '/' in left-to-right order. |
| 866 | (for more robust floating-point arithmetic) |
| 867 | ** snarf.h: Declare static const function name vars as SCM_UNUSED. |
| 868 | ** chars.c: Remove duplicate 'const' specifiers. |
| 869 | ** Modify SCM_UNPACK type check to avoid warnings in clang. |
| 870 | ** Arrange so that 'file-encoding' does not truncate the encoding name. |
| 871 | (http://bugs.gnu.org/16463) |
| 872 | ** Improve error checking in bytevector->uint-list and bytevector->sint-list. |
| 873 | (http://bugs.gnu.org/15100) |
| 874 | ** Fix (ash -1 SCM_I_FIXNUM_BIT-1) to return a fixnum instead of a bignum. |
| 875 | ** i18n: Fix null pointer dereference when locale info is missing. |
| 876 | ** Fix 'string-copy!' to work properly with overlapping src/dest. |
| 877 | ** Fix hashing of vectors to run in bounded time. |
| 878 | ** 'port-position' works on CBIPs that do not support 'set-port-position!'. |
| 879 | ** Custom binary input ports sanity-check the return value of 'read!'. |
| 880 | ** bdw-gc.h: Check SCM_USE_PTHREAD_THREADS using #if not #ifdef. |
| 881 | ** REPL Server: Don't establish a SIGINT handler. |
| 882 | ** REPL Server: Redirect warnings to client socket. |
| 883 | ** REPL Server: Improve robustness of 'stop-server-and-clients!'. |
| 884 | ** Add srfi-16, srfi-30, srfi-46, srfi-62, srfi-87 to %cond-expand-features. |
| 885 | ** Fix trap handlers to handle applicable structs. |
| 886 | (http://bugs.gnu.org/15691) |
| 887 | ** Fix optional end argument in `uniform-vector-read!'. |
| 888 | (http://bugs.gnu.org/15370) |
| 889 | ** Fix brainfuck->scheme compiler. |
| 890 | ** texinfo: Fix newline preservation in @example with lines beginning with @ |
| 891 | |
| 892 | ** C standards conformance improvements |
| 893 | |
| 894 | Improvements and bug fixes were made to the C part of Guile's run-time |
| 895 | support (libguile). |
| 896 | |
| 897 | *** Don't use the identifier 'noreturn'. |
| 898 | (http://bugs.gnu.org/15798) |
| 899 | *** Rewrite SCM_I_INUM to avoid unspecified behavior when not using GNU C. |
| 900 | *** Improve fallback implemention of SCM_SRS to avoid unspecified behavior. |
| 901 | *** SRFI-60: Reimplement 'rotate-bit-field' on inums to be more portable. |
| 902 | *** Improve compliance with C standards regarding signed integer shifts. |
| 903 | *** Avoid signed overflow in random.c. |
| 904 | *** VM: Avoid signed overflows in 'add1' and 'sub1'. |
| 905 | *** VM: Avoid overflow in ASM_ADD when the result is most-positive-fixnum. |
| 906 | *** read: Avoid signed integer overflow in 'read_decimal_integer'. |
| 907 | |
| 908 | |
| 909 | \f |
| 910 | Changes in 2.0.9 (since 2.0.7): |
| 911 | |
| 912 | Note: 2.0.8 was a brown paper bag release that was never announced, but |
| 913 | some mirrors may have picked it up. Please do not use it. |
| 914 | |
| 915 | * Notable changes |
| 916 | |
| 917 | ** New keyword arguments for procedures that open files |
| 918 | |
| 919 | The following procedures that open files now support keyword arguments |
| 920 | to request binary I/O or to specify the character encoding for text |
| 921 | files: `open-file', `open-input-file', `open-output-file', |
| 922 | `call-with-input-file', `call-with-output-file', `with-input-from-file', |
| 923 | `with-output-to-file', and `with-error-to-file'. |
| 924 | |
| 925 | It is also now possible to specify whether Guile should scan files for |
| 926 | Emacs-style coding declarations. This scan was done by default in |
| 927 | versions 2.0.0 through 2.0.7, but now must be explicitly requested. |
| 928 | |
| 929 | See "File Ports" in the manual for details. |
| 930 | |
| 931 | ** Rewritten guile.m4 |
| 932 | |
| 933 | The `guile.m4' autoconf macros have been rewritten to use `guild' and |
| 934 | `pkg-config' instead of the deprecated `guile-config' (which itself |
| 935 | calls pkg-config). |
| 936 | |
| 937 | There is also a new macro, `GUILE_PKG', which allows packages to select |
| 938 | the version of Guile that they want to compile against. See "Autoconf |
| 939 | Macros" in the manual, for more information. |
| 940 | |
| 941 | ** Better Windows support |
| 942 | |
| 943 | Guile now correctly identifies absolute paths on Windows (MinGW), and |
| 944 | creates files on that platform according to its path conventions. See |
| 945 | "File System" in the manual, for all details. |
| 946 | |
| 947 | In addition, the new Gnulib imports provide `select' and `poll' on |
| 948 | Windows builds. |
| 949 | |
| 950 | As an incompatible change, systems that are missing <sys/select.h> were |
| 951 | previously provided a public `scm_std_select' C function that defined a |
| 952 | version of `select', but unhappily it also provided its own incompatible |
| 953 | definitions for FD_SET, FD_ZERO, and other system interfaces. Guile |
| 954 | should not be setting these macros in public API, so this interface was |
| 955 | removed on those plaforms (basically only MinGW). |
| 956 | |
| 957 | ** Numerics improvements |
| 958 | |
| 959 | `number->string' now reliably outputs enough digits to produce the same |
| 960 | number when read back in. Previously, it mishandled subnormal numbers |
| 961 | (printing them as "#.#"), and failed to distinguish between some |
| 962 | distinct inexact numbers, e.g. 1.0 and (+ 1.0 (expt 2.0 -52)). These |
| 963 | problems had far-reaching implications, since the compiler uses |
| 964 | `number->string' to serialize numeric constants into .go files. |
| 965 | |
| 966 | `sqrt' now produces exact rational results when possible, and handles |
| 967 | very large or very small numbers more robustly. |
| 968 | |
| 969 | A number (ahem) of operations involving exact rationals have been |
| 970 | optimized, most notably `integer-expt' and `expt'. |
| 971 | |
| 972 | `exact->inexact' now performs correct IEEE rounding. |
| 973 | |
| 974 | ** New optimizations |
| 975 | |
| 976 | There were a number of improvements to the partial evaluator, allowing |
| 977 | complete reduction of forms such as: |
| 978 | |
| 979 | ((let ((_ 10)) (lambda () _))) |
| 980 | |
| 981 | ((lambda _ _)) |
| 982 | |
| 983 | (apply (lambda _ _) 1 2 3 '(4)) |
| 984 | |
| 985 | (call-with-values (lambda () (values 1 2)) (lambda _ _)) |
| 986 | |
| 987 | `string-join' now handles huge lists efficiently. |
| 988 | |
| 989 | `get-bytevector-some' now uses buffered input, which is much faster. |
| 990 | |
| 991 | Finally, `array-ref', `array-set!' on arrays of rank 1 or 2 is now |
| 992 | faster, because it avoids building a rest list. Similarly, the |
| 993 | one-argument case of `array-for-each' and `array-map!' has been |
| 994 | optimized, and `array-copy!' and `array-fill!' are faster. |
| 995 | |
| 996 | ** `peek-char' no longer consumes EOF |
| 997 | |
| 998 | As required by the R5RS, if `peek-char' returns EOF, then the next read |
| 999 | will also return EOF. Previously `peek-char' would consume the EOF. |
| 1000 | This makes a difference for terminal devices where it is possible to |
| 1001 | read past an EOF. |
| 1002 | |
| 1003 | ** Gnulib update |
| 1004 | |
| 1005 | Guile's copy of Gnulib was updated to v0.0-7865-ga828bb2. The following |
| 1006 | modules were imported from Gnulib: select, times, pipe-posix, fstat, |
| 1007 | getlogin, poll, and c-strcase. |
| 1008 | |
| 1009 | ** `include' resolves relative file names relative to including file |
| 1010 | |
| 1011 | Given a relative file name, `include' will look for it relative to the |
| 1012 | directory of the including file. This harmonizes the behavior of |
| 1013 | `include' with that of `load'. |
| 1014 | |
| 1015 | ** SLIB compatibility restored |
| 1016 | |
| 1017 | Guile 2.0.8 is now compatible with SLIB. You will have to use a |
| 1018 | development version of SLIB, however, until a new version of SLIB is |
| 1019 | released. |
| 1020 | |
| 1021 | ** Better ,trace REPL command |
| 1022 | |
| 1023 | Sometimes the ,trace output for nested function calls could overflow the |
| 1024 | terminal width, which wasn't useful. Now there is a limit to the amount |
| 1025 | of space the prefix will take. See the documentation for ",trace" for |
| 1026 | more information. |
| 1027 | |
| 1028 | ** Better docstring syntax supported for `case-lambda' |
| 1029 | |
| 1030 | Docstrings can now be placed immediately after the `case-lambda' or |
| 1031 | `case-lambda*' keyword. See "Case-lambda" in the manual. |
| 1032 | |
| 1033 | ** Improved handling of Unicode byte order marks |
| 1034 | |
| 1035 | See "BOM Handling" in the manual for details. |
| 1036 | |
| 1037 | ** Update predefined character sets to Unicode 6.2 |
| 1038 | |
| 1039 | ** GMP 4.2 or later required |
| 1040 | |
| 1041 | Guile used to require GMP at least version 4.1 (released in May 2002), |
| 1042 | and now requires at least version 4.2 (released in March 2006). |
| 1043 | |
| 1044 | * Manual updates |
| 1045 | |
| 1046 | ** Better SXML documentation |
| 1047 | |
| 1048 | The documentation for SXML modules was much improved, though there is |
| 1049 | still far to go. See "SXML" in manual. |
| 1050 | |
| 1051 | ** Style updates |
| 1052 | |
| 1053 | Use of "iff" was replaced with standard English. Keyword arguments are |
| 1054 | now documented consistently, along with their default values. |
| 1055 | |
| 1056 | ** An end to the generated-documentation experiment |
| 1057 | |
| 1058 | When Guile 2.0 imported some modules from Guile-Lib, they came with a |
| 1059 | system that generated documentation from docstrings and module |
| 1060 | commentaries. This produced terrible documentation. We finally bit the |
| 1061 | bullet and incorporated these modules into the main text, and will be |
| 1062 | improving them manually over time, as is the case with SXML. Help is |
| 1063 | appreciated. |
| 1064 | |
| 1065 | ** New documentation |
| 1066 | |
| 1067 | There is now documentation for `scm_array_type', and `scm_array_ref', as |
| 1068 | well as for the new `array-length' / 'scm_c_array_length' / |
| 1069 | `scm_array_length' functions. `array-in-bounds?' has better |
| 1070 | documentation as well. The `program-arguments-alist' and |
| 1071 | `program-lambda-list' functions are now documented, as well as `and=>', |
| 1072 | `exit', and `quit'. The (system repl server) module is now documented |
| 1073 | (see REPL Servers). Finally, the GOOPS class hierarchy diagram has been |
| 1074 | regenerated for the web and print output formats. |
| 1075 | |
| 1076 | * New deprecations |
| 1077 | |
| 1078 | ** Deprecate generalized vector interface |
| 1079 | |
| 1080 | The generalized vector interface, introduced in 1.8.0, is simply a |
| 1081 | redundant, verbose interface to arrays of rank 1. `array-ref' and |
| 1082 | similar functions are entirely sufficient. Thus, |
| 1083 | `scm_generalized_vector_p', `scm_generalized_vector_length', |
| 1084 | `scm_generalized_vector_ref', `scm_generalized_vector_set_x', and |
| 1085 | `scm_generalized_vector_to_list' are now deprecated. |
| 1086 | |
| 1087 | ** Deprecate SCM_CHAR_CODE_LIMIT and char-code-limit |
| 1088 | |
| 1089 | These constants were defined to 256, which is not the highest codepoint |
| 1090 | supported by Guile. Given that they were useless and incorrect, they |
| 1091 | have been deprecated. |
| 1092 | |
| 1093 | ** Deprecate `http-get*' |
| 1094 | |
| 1095 | The new `#:streaming?' argument to `http-get' subsumes the functionality |
| 1096 | of `http-get*' (introduced in 2.0.7). Also, the `#:extra-headers' |
| 1097 | argument is deprecated in favor of `#:headers'. |
| 1098 | |
| 1099 | ** Deprecate (ice-9 mapping) |
| 1100 | |
| 1101 | This module, present in Guile since 1996 but never used or documented, |
| 1102 | has never worked in Guile 2.0. It has now been deprecated and will be |
| 1103 | removed in Guile 2.2. |
| 1104 | |
| 1105 | ** Deprecate undocumented array-related C functions |
| 1106 | |
| 1107 | These are `scm_array_fill_int', `scm_ra_eqp', `scm_ra_lessp', |
| 1108 | `scm_ra_leqp', `scm_ra_grp', `scm_ra_greqp', `scm_ra_sum', |
| 1109 | `scm_ra_product', `scm_ra_difference', `scm_ra_divide', and |
| 1110 | `scm_array_identity'. |
| 1111 | |
| 1112 | * New interfaces |
| 1113 | |
| 1114 | ** SRFI-41 Streams |
| 1115 | |
| 1116 | See "SRFI-41" in the manual. |
| 1117 | |
| 1118 | ** SRFI-45 exports `promise?' |
| 1119 | |
| 1120 | SRFI-45 now exports a `promise?' procedure that works with its promises. |
| 1121 | Also, its promises now print more nicely. |
| 1122 | |
| 1123 | ** New HTTP client procedures |
| 1124 | |
| 1125 | See "Web Client" for documentation on the new `http-head', `http-post', |
| 1126 | `http-put', `http-delete', `http-trace', and `http-options' procedures, |
| 1127 | and also for more options to `http-get'. |
| 1128 | |
| 1129 | ** Much more capable `xml->sxml' |
| 1130 | |
| 1131 | See "Reading and Writing XML" for information on how the `xml->sxml' |
| 1132 | parser deals with namespaces, processed entities, doctypes, and literal |
| 1133 | strings. Incidentally, `current-ssax-error-port' is now a parameter |
| 1134 | object. |
| 1135 | |
| 1136 | ** New procedures for converting strings to and from bytevectors |
| 1137 | |
| 1138 | See "Representing Strings as Bytes" for documention on the new `(ice-9 |
| 1139 | iconv)' module and its `bytevector->string' and `string->bytevector' |
| 1140 | procedures. |
| 1141 | |
| 1142 | ** Escape continuations with `call/ec' and `let/ec' |
| 1143 | |
| 1144 | See "Prompt Primitives". |
| 1145 | |
| 1146 | ** New procedures to read all characters from a port |
| 1147 | |
| 1148 | See "Line/Delimited" in the manual for documentation on `read-string' |
| 1149 | and `read-string!'. |
| 1150 | |
| 1151 | ** New procedure `sendfile' |
| 1152 | |
| 1153 | See "File System". |
| 1154 | |
| 1155 | ** New procedure `unget-bytevector' |
| 1156 | |
| 1157 | See "R6RS Binary Input". |
| 1158 | |
| 1159 | ** New C helper: `scm_c_bind_keyword_arguments' |
| 1160 | |
| 1161 | See "Keyword Procedures". |
| 1162 | |
| 1163 | ** New command-line arguments: `--language' and `-C' |
| 1164 | |
| 1165 | See "Command-line Options" in the manual. |
| 1166 | |
| 1167 | ** New environment variables: `GUILE_STACK_SIZE', `GUILE_INSTALL_LOCALE' |
| 1168 | |
| 1169 | See "Environment Variables". |
| 1170 | |
| 1171 | ** New procedures for dealing with file names |
| 1172 | |
| 1173 | See "File System" for documentation on `system-file-name-convention', |
| 1174 | `file-name-separator?', `absolute-file-name?', and |
| 1175 | `file-name-separator-string'. |
| 1176 | |
| 1177 | ** `array-length', an array's first dimension |
| 1178 | |
| 1179 | See "Array Procedures". |
| 1180 | |
| 1181 | ** `hash-count', for hash tables |
| 1182 | |
| 1183 | See "Hash Tables". |
| 1184 | |
| 1185 | ** `round-ash', a bit-shifting operator that rounds on right-shift |
| 1186 | |
| 1187 | See "Bitwise Operations". |
| 1188 | |
| 1189 | ** New foreign types: `ssize_t', `ptrdiff_t' |
| 1190 | |
| 1191 | See "Foreign Types". |
| 1192 | |
| 1193 | ** New C helpers: `scm_from_ptrdiff_t', `scm_to_ptrdiff_t' |
| 1194 | |
| 1195 | See "Integers". |
| 1196 | |
| 1197 | ** Socket option `SO_REUSEPORT' now available from Scheme |
| 1198 | |
| 1199 | If supported on the platform, `SO_REUSEPORT' is now available from |
| 1200 | Scheme as well. See "Network Sockets and Communication". |
| 1201 | |
| 1202 | ** `current-language' in default environment |
| 1203 | |
| 1204 | Previously defined only in `(system base language)', `current-language' |
| 1205 | is now defined in the default environment, and is used to determine the |
| 1206 | language for the REPL, and for `compile-and-load'. |
| 1207 | |
| 1208 | ** New procedure: `fluid->parameter' |
| 1209 | |
| 1210 | See "Parameters", for information on how to convert a fluid to a |
| 1211 | parameter. |
| 1212 | |
| 1213 | ** New `print' REPL option |
| 1214 | |
| 1215 | See "REPL Commands" in the manual for information on the new |
| 1216 | user-customizable REPL printer. |
| 1217 | |
| 1218 | ** New variable: %site-ccache-dir |
| 1219 | |
| 1220 | The "Installing Site Packages" and "Build Config" manual sections now |
| 1221 | refer to this variable to describe where users should install their |
| 1222 | `.go' files. |
| 1223 | |
| 1224 | * Build fixes |
| 1225 | |
| 1226 | ** Fix compilation against libgc 7.3. |
| 1227 | ** Fix cross-compilation of `c-tokenize.o'. |
| 1228 | ** Fix warning when compiling against glibc 2.17. |
| 1229 | ** Fix documentation build against Texinfo 5.0. |
| 1230 | ** Fix building Guile from a directory with non-ASCII characters. |
| 1231 | ** Fix native MinGW build. |
| 1232 | ** Fix --disable-posix build. |
| 1233 | ** Fix MinGW builds with networking, POSIX, and thread support. |
| 1234 | |
| 1235 | * Bug fixes |
| 1236 | |
| 1237 | ** Fix inexact number printer. |
| 1238 | (http://bugs.gnu.org/13757) |
| 1239 | ** Fix infinite loop when parsing optional-argument short options (SRFI-37). |
| 1240 | (http://bugs.gnu.org/13176) |
| 1241 | ** web: Support non-GMT date headers in the HTTP client. |
| 1242 | (http://bugs.gnu.org/13544) |
| 1243 | ** web: support IP-literal (IPv6 address) in Host header. |
| 1244 | ** Avoid stack overflows with `par-map' and nested futures in general. |
| 1245 | (http://bugs.gnu.org/13188) |
| 1246 | ** Peek-char no longer consumes EOF. |
| 1247 | (http://bugs.gnu.org/12216) |
| 1248 | ** Avoid swallowing multiple EOFs in R6RS binary-input procedures. |
| 1249 | ** A fork when multiple threads are running will now print a warning. |
| 1250 | ** Allow for spurious wakeups from pthread_cond_wait. |
| 1251 | (http://bugs.gnu.org/10641) |
| 1252 | ** Warn and ignore module autoload failures. |
| 1253 | (http://bugs.gnu.org/12202) |
| 1254 | ** Use chmod portably in (system base compile). |
| 1255 | (http://bugs.gnu.org/10474) |
| 1256 | ** Fix response-body-port for HTTP responses without content-length. |
| 1257 | (http://bugs.gnu.org/13857) |
| 1258 | ** Allow case-lambda expressions with no clauses. |
| 1259 | (http://bugs.gnu.org/9776) |
| 1260 | ** Improve standards conformance of string->number. |
| 1261 | (http://bugs.gnu.org/11887) |
| 1262 | ** Support calls and tail-calls with more than 255 formals. |
| 1263 | ** ,option evaluates its right-hand-side. |
| 1264 | (http://bugs.gnu.org/13076) |
| 1265 | ** Structs with tail arrays are not simple. |
| 1266 | (http://bugs.gnu.org/12808) |
| 1267 | ** Make `SCM_LONG_BIT' usable in preprocessor conditionals. |
| 1268 | (http://bugs.gnu.org/13848) |
| 1269 | ** Fix thread-unsafe lazy initializations. |
| 1270 | ** Allow SMOB mark procedures to be called from parallel markers. |
| 1271 | (http://bugs.gnu.org/13611) |
| 1272 | ** Fix later-bindings-win logic in with-fluids. |
| 1273 | (http://bugs.gnu.org/13843) |
| 1274 | ** Fix duplicate removal of with-fluids. |
| 1275 | (http://bugs.gnu.org/13838) |
| 1276 | ** Support calling foreign functions of 10 arguments or more. |
| 1277 | (http://bugs.gnu.org/13809) |
| 1278 | ** Let reverse! accept arbitrary types as second argument. |
| 1279 | (http://bugs.gnu.org/13835) |
| 1280 | ** Recognize the `x86_64.*-gnux32' triplet. |
| 1281 | ** Check whether a triplet's OS part specifies an ABI. |
| 1282 | ** Recognize mips64* as having 32-bit pointers by default. |
| 1283 | ** Use portable sed constructs. |
| 1284 | (http://bugs.gnu.org/14042) |
| 1285 | ** Remove language/glil/decompile-assembly.scm. |
| 1286 | (http://bugs.gnu.org/10622) |
| 1287 | ** Use O_BINARY in `copy-file', `load-objcode', `mkstemp'. |
| 1288 | ** Use byte-oriented functions in `get-bytevector*'. |
| 1289 | ** Fix abort when iconv swallows BOM from UTF-16 or UTF-32 stream. |
| 1290 | ** Fix compilation of functions with more than 255 local variables. |
| 1291 | ** Fix `getgroups' for when zero supplementary group IDs exist. |
| 1292 | ** Allow (define-macro name (lambda ...)). |
| 1293 | ** Various fixes to the (texinfo) modules. |
| 1294 | ** guild: Gracefully handle failures to install the locale. |
| 1295 | ** Fix format string warnings for ~!, ~|, ~/, ~q, ~Q, and ~^. |
| 1296 | (http://bugs.gnu.org/13485) |
| 1297 | ** Fix source annotation bug in psyntax 'expand-body'. |
| 1298 | ** Ecmascript: Fix conversion to boolean for non-numbers. |
| 1299 | ** Use case-insensitive comparisons for encoding names. |
| 1300 | ** Add missing cond-expand feature identifiers. |
| 1301 | ** A failure to find a module's file does not prevent future loading. |
| 1302 | ** Many (oop goops save) fixes. |
| 1303 | ** `http-get': don't shutdown write end of socket. |
| 1304 | (http://bugs.gnu.org/13095) |
| 1305 | ** Avoid signed integer overflow in scm_product. |
| 1306 | ** http: read-response-body always returns bytevector or #f, never EOF. |
| 1307 | ** web: Correctly detect "No route to host" conditions. |
| 1308 | ** `system*': failure to execvp no longer leaks dangling processes. |
| 1309 | (http://bugs.gnu.org/13166) |
| 1310 | ** More sensible case-lambda* dispatch. |
| 1311 | (http://bugs.gnu.org/12929) |
| 1312 | ** Do not defer expansion of internal define-syntax forms. |
| 1313 | (http://bugs.gnu.org/13509) |
| 1314 | |
| 1315 | |
| 1316 | \f |
| 1317 | Changes in 2.0.7 (since 2.0.6): |
| 1318 | |
| 1319 | * Notable changes |
| 1320 | |
| 1321 | ** SRFI-105 curly infix expressions are supported |
| 1322 | |
| 1323 | Curly infix expressions as described at |
| 1324 | http://srfi.schemers.org/srfi-105/srfi-105.html are now supported by |
| 1325 | Guile's reader. This allows users to write things like {a * {b + c}} |
| 1326 | instead of (* a (+ b c)). SRFI-105 support is enabled by using the |
| 1327 | `#!curly-infix' directive in source code, or the `curly-infix' reader |
| 1328 | option. See the manual for details. |
| 1329 | |
| 1330 | ** Reader options may now be per-port |
| 1331 | |
| 1332 | Historically, `read-options' and related procedures would manipulate |
| 1333 | global options, affecting the `read' procedure for all threads, and all |
| 1334 | current uses of `read'. |
| 1335 | |
| 1336 | Guile can now associate `read' options with specific ports, allowing |
| 1337 | different ports to use different options. For instance, the |
| 1338 | `#!fold-case' and `#!no-fold-case' reader directives have been |
| 1339 | implemented, and their effect is to modify the current read options of |
| 1340 | the current port only; similarly for `#!curly-infix'. Thus, it is |
| 1341 | possible, for instance, to have one port reading case-sensitive code, |
| 1342 | while another port reads case-insensitive code. |
| 1343 | |
| 1344 | ** Futures may now be nested |
| 1345 | |
| 1346 | Futures may now be nested: a future can itself spawn and then `touch' |
| 1347 | other futures. In addition, any thread that touches a future that has |
| 1348 | not completed now processes other futures while waiting for the touched |
| 1349 | future to completed. This allows all threads to be kept busy, and was |
| 1350 | made possible by the use of delimited continuations (see the manual for |
| 1351 | details.) |
| 1352 | |
| 1353 | Consequently, `par-map' and `par-for-each' have been rewritten and can |
| 1354 | now use all cores. |
| 1355 | |
| 1356 | ** `GUILE_LOAD_PATH' et al can now add directories to the end of the path |
| 1357 | |
| 1358 | `GUILE_LOAD_PATH' and `GUILE_LOAD_COMPILED_PATH' can now be used to add |
| 1359 | directories to both ends of the load path. If the special path |
| 1360 | component `...' (ellipsis) is present in these environment variables, |
| 1361 | then the default path is put in place of the ellipsis, otherwise the |
| 1362 | default path is placed at the end. See "Environment Variables" in the |
| 1363 | manual for details. |
| 1364 | |
| 1365 | ** `load-in-vicinity' search for `.go' files in `%load-compiled-path' |
| 1366 | |
| 1367 | Previously, `load-in-vicinity' would look for compiled files in the |
| 1368 | auto-compilation cache, but not in `%load-compiled-path'. This is now |
| 1369 | fixed. This affects `load', and the `-l' command-line flag. See |
| 1370 | <http://bugs.gnu.org/12519> for details. |
| 1371 | |
| 1372 | ** Extension search order fixed, and LD_LIBRARY_PATH preserved |
| 1373 | |
| 1374 | Up to 2.0.6, Guile would modify the `LD_LIBRARY_PATH' environment |
| 1375 | variable (or whichever is relevant for the host OS) to insert its own |
| 1376 | default extension directories in the search path (using GNU libltdl |
| 1377 | facilities was not possible here.) This approach was problematic in two |
| 1378 | ways. |
| 1379 | |
| 1380 | First, the `LD_LIBRARY_PATH' modification would be visible to |
| 1381 | sub-processes, and would also affect future calls to `dlopen', which |
| 1382 | could lead to subtle bugs in the application or sub-processes. Second, |
| 1383 | when the installation prefix is /usr, the `LD_LIBRARY_PATH' modification |
| 1384 | would typically end up inserting /usr/lib before /usr/local/lib in the |
| 1385 | search path, which is often the opposite of system-wide settings such as |
| 1386 | `ld.so.conf'. |
| 1387 | |
| 1388 | Both issues have now been fixed. |
| 1389 | |
| 1390 | ** `make-vtable-vtable' is now deprecated |
| 1391 | |
| 1392 | Programs should instead use `make-vtable' and `<standard-vtable>'. |
| 1393 | |
| 1394 | ** The `-Wduplicate-case-datum' and `-Wbad-case-datum' are enabled |
| 1395 | |
| 1396 | These recently introduced warnings have been documented and are now |
| 1397 | enabled by default when auto-compiling. |
| 1398 | |
| 1399 | ** Optimize calls to `equal?' or `eqv?' with a constant argument |
| 1400 | |
| 1401 | The compiler simplifies calls to `equal?' or `eqv?' with a constant |
| 1402 | argument to use `eq?' instead, when applicable. |
| 1403 | |
| 1404 | * Manual updates |
| 1405 | |
| 1406 | ** SRFI-9 records now documented under "Compound Data Types" |
| 1407 | |
| 1408 | The documentation of SRFI-9 record types has been moved in the "Compound |
| 1409 | Data Types", next to Guile's other record APIs. A new section |
| 1410 | introduces the various record APIs, and describes the trade-offs they |
| 1411 | make. These changes were made in an attempt to better guide users |
| 1412 | through the maze of records API, and to recommend SRFI-9 as the main |
| 1413 | API. |
| 1414 | |
| 1415 | The documentation of Guile's raw `struct' API has also been improved. |
| 1416 | |
| 1417 | ** (ice-9 and-let-star) and (ice-9 curried-definitions) now documented |
| 1418 | |
| 1419 | These modules were missing from the manual. |
| 1420 | |
| 1421 | * New interfaces |
| 1422 | |
| 1423 | ** New "functional record setters" as a GNU extension of SRFI-9 |
| 1424 | |
| 1425 | The (srfi srfi-9 gnu) module now provides three new macros to deal with |
| 1426 | "updates" of immutable records: `define-immutable-record-type', |
| 1427 | `set-field', and `set-fields'. |
| 1428 | |
| 1429 | The first one allows record type "functional setters" to be defined; |
| 1430 | such setters keep the record unchanged, and instead return a new record |
| 1431 | with only one different field. The remaining macros provide the same |
| 1432 | functionality, and also optimize updates of multiple or nested fields. |
| 1433 | See the manual for details. |
| 1434 | |
| 1435 | ** web: New `http-get*', `response-body-port', and `text-content-type?' |
| 1436 | procedures |
| 1437 | |
| 1438 | These procedures return a port from which to read the response's body. |
| 1439 | Unlike `http-get' and `read-response-body', they allow the body to be |
| 1440 | processed incrementally instead of being stored entirely in memory. |
| 1441 | |
| 1442 | The `text-content-type?' predicate allows users to determine whether the |
| 1443 | content type of a response is textual. |
| 1444 | |
| 1445 | See the manual for details. |
| 1446 | |
| 1447 | ** `string-split' accepts character sets and predicates |
| 1448 | |
| 1449 | The `string-split' procedure can now be given a SRFI-14 character set or |
| 1450 | a predicate, instead of just a character. |
| 1451 | |
| 1452 | ** R6RS SRFI support |
| 1453 | |
| 1454 | Previously, in R6RS modules, Guile incorrectly ignored components of |
| 1455 | SRFI module names after the SRFI number, making it impossible to specify |
| 1456 | sub-libraries. This release corrects this, bringing us into accordance |
| 1457 | with SRFI 97. |
| 1458 | |
| 1459 | ** `define-public' is no a longer curried definition by default |
| 1460 | |
| 1461 | The (ice-9 curried-definitions) should be used for such uses. See the |
| 1462 | manual for details. |
| 1463 | |
| 1464 | * Build fixes |
| 1465 | |
| 1466 | ** Remove reference to `scm_init_popen' when `fork' is unavailable |
| 1467 | |
| 1468 | This fixes a MinGW build issue (http://bugs.gnu.org/12477). |
| 1469 | |
| 1470 | ** Fix race between installing `guild' and the `guile-tools' symlink |
| 1471 | |
| 1472 | * Bug fixes |
| 1473 | |
| 1474 | ** Procedures returned by `eval' now have docstrings |
| 1475 | (http://bugs.gnu.org/12173) |
| 1476 | ** web client: correctly handle uri-query, etc. in relative URI headers |
| 1477 | (http://bugs.gnu.org/12827) |
| 1478 | ** Fix docs for R6RS `hashtable-copy' |
| 1479 | ** R6RS `string-for-each' now accepts multiple string arguments |
| 1480 | ** Fix out-of-range error in the compiler's CSE pass |
| 1481 | (http://bugs.gnu.org/12883) |
| 1482 | ** Add missing R6RS `open-file-input/output-port' procedure |
| 1483 | ** Futures: Avoid creating the worker pool more than once |
| 1484 | ** Fix invalid assertion about mutex ownership in threads.c |
| 1485 | (http://bugs.gnu.org/12719) |
| 1486 | ** Have `SCM_NUM2FLOAT' and `SCM_NUM2DOUBLE' use `scm_to_double' |
| 1487 | ** The `scandir' procedure now uses `lstat' instead of `stat' |
| 1488 | ** Fix `generalized-vector->list' indexing bug with shared arrays |
| 1489 | (http://bugs.gnu.org/12465) |
| 1490 | ** web: Change `http-get' to try all the addresses for the given URI |
| 1491 | ** Implement `hash' for structs |
| 1492 | (http://lists.gnu.org/archive/html/guile-devel/2012-10/msg00031.html) |
| 1493 | ** `read' now adds source properties for data types beyond pairs |
| 1494 | ** Improve error reporting in `append!' |
| 1495 | ** In fold-matches, set regexp/notbol unless matching string start |
| 1496 | ** Don't stat(2) and access(2) the .go location before using it |
| 1497 | ** SRFI-19: use zero padding for hours in ISO 8601 format, not blanks |
| 1498 | ** web: Fix uri-encoding for strings with no unreserved chars, and octets 0-15 |
| 1499 | ** More robust texinfo alias handling |
| 1500 | ** Optimize `format' and `simple-format' |
| 1501 | (http://bugs.gnu.org/12033) |
| 1502 | ** Angle of -0.0 is pi, not zero |
| 1503 | |
| 1504 | \f |
| 1505 | Changes in 2.0.6 (since 2.0.5): |
| 1506 | |
| 1507 | * Notable changes |
| 1508 | |
| 1509 | ** New optimization pass: common subexpression elimination (CSE) |
| 1510 | |
| 1511 | Guile's optimizer will now run a CSE pass after partial evaluation. |
| 1512 | This pass propagates static information about branches taken, bound |
| 1513 | lexicals, and effects from an expression's dominators. It can replace |
| 1514 | common subexpressions with their boolean values (potentially enabling |
| 1515 | dead code elimination), equivalent bound lexicals, or it can elide them |
| 1516 | entirely, depending on the context in which they are executed. This |
| 1517 | pass is especially useful in removing duplicate type checks, such as |
| 1518 | those produced by SRFI-9 record accessors. |
| 1519 | |
| 1520 | ** Improvements to the partial evaluator |
| 1521 | |
| 1522 | Peval can now hoist tests that are common to both branches of a |
| 1523 | conditional into the test. This can help with long chains of |
| 1524 | conditionals, such as those generated by the `match' macro. Peval can |
| 1525 | now do simple beta-reductions of procedures with rest arguments. It |
| 1526 | also avoids residualizing degenerate lexical aliases, even when full |
| 1527 | inlining is not possible. Finally, peval now uses the effects analysis |
| 1528 | introduced for the CSE pass. More precise effects analysis allows peval |
| 1529 | to move more code. |
| 1530 | |
| 1531 | ** Run finalizers asynchronously in asyncs |
| 1532 | |
| 1533 | Finalizers are now run asynchronously, via an async. See Asyncs in the |
| 1534 | manual. This allows Guile and user code to safely allocate memory while |
| 1535 | holding a mutex. |
| 1536 | |
| 1537 | ** Update SRFI-14 character sets to Unicode 6.1 |
| 1538 | |
| 1539 | Note that this update causes the Latin-1 characters `§' and `¶' to be |
| 1540 | reclassified as punctuation. They were previously considered to be part |
| 1541 | of `char-set:symbol'. |
| 1542 | |
| 1543 | ** Better source information for datums |
| 1544 | |
| 1545 | When the `positions' reader option is on, as it is by default, Guile's |
| 1546 | reader will record source information for more kinds of datums. |
| 1547 | |
| 1548 | ** Improved error and warning messages |
| 1549 | |
| 1550 | `syntax-violation' errors now prefer `subform' for source info, with |
| 1551 | `form' as fallback. Syntactic errors in `cond' and `case' now produce |
| 1552 | better errors. `case' can now warn on duplicate datums, or datums that |
| 1553 | cannot be usefully compared with `eqv?'. `-Warity-mismatch' now handles |
| 1554 | applicable structs. `-Wformat' is more robust in the presence of |
| 1555 | `gettext'. Finally, various exceptions thrown by the Web modules now |
| 1556 | define appropriate exception printers. |
| 1557 | |
| 1558 | ** A few important bug fixes in the HTTP modules. |
| 1559 | |
| 1560 | Guile's web server framework now checks if an application returns a body |
| 1561 | where it is not permitted, for example in response to a HEAD request, |
| 1562 | and warn or truncate the response as appropriate. Bad requests now |
| 1563 | cause a 400 Bad Request response to be printed before closing the port. |
| 1564 | Finally, some date-printing and URL-parsing bugs were fixed. |
| 1565 | |
| 1566 | ** Pretty-print improvements |
| 1567 | |
| 1568 | When Guile needs to pretty-print Tree-IL, it will try to reconstruct |
| 1569 | `cond', `or`, and other derived syntax forms from the primitive tree-IL |
| 1570 | forms. It also uses the original names instead of the fresh unique |
| 1571 | names, when it is unambiguous to do so. This can be seen in the output |
| 1572 | of REPL commands like `,optimize'. |
| 1573 | |
| 1574 | Also, the `pretty-print' procedure has a new keyword argument, |
| 1575 | `#:max-expr-width'. |
| 1576 | |
| 1577 | ** Fix memory leak involving applicable SMOBs |
| 1578 | |
| 1579 | At some point in the 1.9.x series, Guile began leaking any applicable |
| 1580 | SMOB that was actually applied. (There was a weak-key map from SMOB to |
| 1581 | trampoline functions, where the value had a strong reference on the |
| 1582 | key.) This has been fixed. There was much rejoicing! |
| 1583 | |
| 1584 | ** Support for HTTP/1.1 chunked transfer coding |
| 1585 | |
| 1586 | See "Transfer Codings" in the manual, for more. |
| 1587 | |
| 1588 | ** Micro-optimizations |
| 1589 | |
| 1590 | A pile of micro-optimizations: the `string-trim' function when called |
| 1591 | with `char-set:whitespace'; the `(web http)' parsers; SMOB application; |
| 1592 | conversion of raw UTF-8 and UTF-32 data to and from SCM strings; vlists |
| 1593 | and vhashes; `read' when processing string literals. |
| 1594 | |
| 1595 | ** Incompatible change to `scandir' |
| 1596 | |
| 1597 | As was the original intention, `scandir' now runs the `select?' |
| 1598 | procedure on all items, including subdirectories and the `.' and `..' |
| 1599 | entries. It receives the basename of the file in question instead of |
| 1600 | the full name. We apologize for this incompatible change to this |
| 1601 | function introduced in the 2.0.4 release. |
| 1602 | |
| 1603 | * Manual updates |
| 1604 | |
| 1605 | The manual has been made much more consistent in its naming conventions |
| 1606 | with regards to formal parameters of functions. Thanks to Bake Timmons. |
| 1607 | |
| 1608 | * New interfaces |
| 1609 | |
| 1610 | ** New C function: `scm_to_pointer' |
| 1611 | ** New C inline functions: `scm_new_smob', `scm_new_double_smob' |
| 1612 | ** (ice-9 format): Add ~h specifier for localized number output. |
| 1613 | ** (web response): New procedure: `response-must-not-include-body?' |
| 1614 | ** New predicate: 'supports-source-properties?' |
| 1615 | ** New C helpers: `scm_c_values', `scm_c_nvalues' |
| 1616 | ** Newly public inline C function: `scm_unget_byte' |
| 1617 | ** (language tree-il): New functions: `tree-il=?', `tree-il-hash' |
| 1618 | ** New fluid: `%default-port-conversion-strategy' |
| 1619 | ** New syntax: `=>' within `case' |
| 1620 | ** (web http): `make-chunked-input-port', `make-chunked-output-port' |
| 1621 | ** (web http): `declare-opaque-header!' |
| 1622 | |
| 1623 | Search the manual for these identifiers, for more information. |
| 1624 | |
| 1625 | * New deprecations |
| 1626 | |
| 1627 | ** `close-io-port' deprecated |
| 1628 | |
| 1629 | Use `close-port'. |
| 1630 | |
| 1631 | ** `scm_sym2var' deprecated |
| 1632 | |
| 1633 | In most cases, replace with `scm_lookup' or `scm_module_variable'. Use |
| 1634 | `scm_define' or `scm_module_ensure_local_variable' if the second |
| 1635 | argument is nonzero. See "Accessing Modules from C" in the manual, for |
| 1636 | full details. |
| 1637 | |
| 1638 | ** Lookup closures deprecated |
| 1639 | |
| 1640 | These were never documented. See "Module System Reflection" in the |
| 1641 | manual for replacements. |
| 1642 | |
| 1643 | * Build fixes |
| 1644 | |
| 1645 | ** Fix compilation against uninstalled Guile on non-GNU platforms. |
| 1646 | ** Fix `SCM_I_ERROR' definition for MinGW without networking. |
| 1647 | ** Fix compilation with the Sun C compiler. |
| 1648 | ** Fix check for `clock_gettime' on OpenBSD and some other systems. |
| 1649 | ** Fix build with --enable-debug-malloc. |
| 1650 | ** Honor $(program_transform_name) for the `guile-tools' symlink. |
| 1651 | ** Fix cross-compilation of GOOPS-using code. |
| 1652 | |
| 1653 | * Bug fixes |
| 1654 | |
| 1655 | ** Fix use of unitialized stat buffer in search-path of absolute paths. |
| 1656 | ** Avoid calling `freelocale' with a NULL argument. |
| 1657 | ** Work around erroneous tr_TR locale in Darwin 8 in tests. |
| 1658 | ** Fix `getaddrinfo' test for Darwin 8. |
| 1659 | ** Use Gnulib's `regex' module for better regex portability. |
| 1660 | ** `source-properties' and friends work on any object |
| 1661 | ** Rewrite open-process in C, for robustness related to threads and fork |
| 1662 | ** Fix <TAG>vector-length when applied to other uniform vector types |
| 1663 | ** Fix escape-only prompt optimization (was disabled previously) |
| 1664 | ** Fix a segfault when /dev/urandom is not accessible |
| 1665 | ** Fix flush on soft ports, so that it actually runs. |
| 1666 | ** Better compatibility of SRFI-9 records with core records |
| 1667 | ** Fix and clarify documentation of `sorted?'. |
| 1668 | ** Fix IEEE-754 endianness conversion in bytevectors. |
| 1669 | ** Correct thunk check in the `wind' instruction. |
| 1670 | ** Add @acronym support to texinfo modules |
| 1671 | ** Fix docbook->texi for <ulink> without URL |
| 1672 | ** Fix `setvbuf' to leave the line/column number unchanged. |
| 1673 | ** Add missing public declaration for `scm_take_from_input_buffers'. |
| 1674 | ** Fix relative file name canonicalization with empty %LOAD-PATH entries. |
| 1675 | ** Import newer (ice-9 match) from Chibi-Scheme. |
| 1676 | ** Fix unbound variables and unbound values in ECMAScript runtime. |
| 1677 | ** Make SRFI-6 string ports Unicode-capable. |
| 1678 | |
| 1679 | \f |
| 1680 | Changes in 2.0.5 (since 2.0.4): |
| 1681 | |
| 1682 | This release fixes the binary interface information (SONAME) of |
| 1683 | libguile, which was incorrect in 2.0.4. It does not contain other |
| 1684 | changes. |
| 1685 | |
| 1686 | \f |
| 1687 | Changes in 2.0.4 (since 2.0.3): |
| 1688 | |
| 1689 | * Notable changes |
| 1690 | |
| 1691 | ** Better debuggability for interpreted procedures. |
| 1692 | |
| 1693 | Guile 2.0 came with a great debugging experience for compiled |
| 1694 | procedures, but the story for interpreted procedures was terrible. Now, |
| 1695 | at least, interpreted procedures have names, and the `arity' procedure |
| 1696 | property is always correct (or, as correct as it can be, in the presence |
| 1697 | of `case-lambda'). |
| 1698 | |
| 1699 | ** Support for cross-compilation. |
| 1700 | |
| 1701 | One can now use a native Guile to cross-compile `.go' files for a |
| 1702 | different architecture. See the documentation for `--target' in the |
| 1703 | "Compilation" section of the manual, for information on how to use the |
| 1704 | cross-compiler. See the "Cross building Guile" section of the README, |
| 1705 | for more on how to cross-compile Guile itself. |
| 1706 | |
| 1707 | ** The return of `local-eval'. |
| 1708 | |
| 1709 | Back by popular demand, `the-environment' and `local-eval' allow the |
| 1710 | user to capture a lexical environment, and then evaluate arbitrary |
| 1711 | expressions in that context. There is also a new `local-compile' |
| 1712 | command. See "Local Evaluation" in the manual, for more. Special |
| 1713 | thanks to Mark Weaver for an initial implementation of this feature. |
| 1714 | |
| 1715 | ** Fluids can now have default values. |
| 1716 | |
| 1717 | Fluids are used for dynamic and thread-local binding. They have always |
| 1718 | inherited their values from the context or thread that created them. |
| 1719 | However, there was a case in which a new thread would enter Guile, and |
| 1720 | the default values of all the fluids would be `#f' for that thread. |
| 1721 | |
| 1722 | This has now been fixed so that `make-fluid' has an optional default |
| 1723 | value for fluids in unrelated dynamic roots, which defaults to `#f'. |
| 1724 | |
| 1725 | ** Garbage collector tuning. |
| 1726 | |
| 1727 | The garbage collector has now been tuned to run more often under some |
| 1728 | circumstances. |
| 1729 | |
| 1730 | *** Unmanaged allocation |
| 1731 | |
| 1732 | The new `scm_gc_register_allocation' function will notify the collector |
| 1733 | of unmanaged allocation. This will cause the collector to run sooner. |
| 1734 | Guile's `scm_malloc', `scm_calloc', and `scm_realloc' unmanaged |
| 1735 | allocators eventually call this function. This leads to better |
| 1736 | performance under steady-state unmanaged allocation. |
| 1737 | |
| 1738 | *** Transient allocation |
| 1739 | |
| 1740 | When the collector runs, it will try to record the total memory |
| 1741 | footprint of a process, if the platform supports this information. If |
| 1742 | the memory footprint is growing, the collector will run more frequently. |
| 1743 | This reduces the increase of the resident size of a process in response |
| 1744 | to a transient increase in allocation. |
| 1745 | |
| 1746 | *** Management of threads, bignums |
| 1747 | |
| 1748 | Creating a thread will allocate a fair amount of memory. Guile now does |
| 1749 | some GC work (using `GC_collect_a_little') when allocating a thread. |
| 1750 | This leads to a better memory footprint when creating many short-lived |
| 1751 | threads. |
| 1752 | |
| 1753 | Similarly, bignums can occupy a lot of memory. Guile now offers hooks |
| 1754 | to enable custom GMP allocators that end up calling |
| 1755 | `scm_gc_register_allocation'. These allocators are enabled by default |
| 1756 | when running Guile from the command-line. To enable them in libraries, |
| 1757 | set the `scm_install_gmp_memory_functions' variable to a nonzero value |
| 1758 | before loading Guile. |
| 1759 | |
| 1760 | ** SRFI-39 parameters are available by default. |
| 1761 | |
| 1762 | Guile now includes support for parameters, as defined by SRFI-39, in the |
| 1763 | default environment. See "Parameters" in the manual, for more |
| 1764 | information. `current-input-port', `current-output-port', and |
| 1765 | `current-error-port' are now parameters. |
| 1766 | |
| 1767 | ** Add `current-warning-port'. |
| 1768 | |
| 1769 | Guile now outputs warnings on a separate port, `current-warning-port', |
| 1770 | initialized to the value that `current-error-port' has on startup. |
| 1771 | |
| 1772 | ** Syntax parameters. |
| 1773 | |
| 1774 | Following Racket's lead, Guile now supports syntax parameters. See |
| 1775 | "Syntax parameters" in the manual, for more. |
| 1776 | |
| 1777 | Also see Barzilay, Culpepper, and Flatt's 2011 SFP workshop paper, |
| 1778 | "Keeping it Clean with syntax-parameterize". |
| 1779 | |
| 1780 | ** Parse command-line arguments from the locale encoding. |
| 1781 | |
| 1782 | Guile now attempts to parse command-line arguments using the user's |
| 1783 | locale. However for backwards compatibility with other 2.0.x releases, |
| 1784 | it does so without actually calling `setlocale'. Please report any bugs |
| 1785 | in this facility to bug-guile@gnu.org. |
| 1786 | |
| 1787 | ** One-armed conditionals: `when' and `unless' |
| 1788 | |
| 1789 | Guile finally has `when' and `unless' in the default environment. Use |
| 1790 | them whenever you would use an `if' with only one branch. See |
| 1791 | "Conditionals" in the manual, for more. |
| 1792 | |
| 1793 | ** `current-filename', `add-to-load-path' |
| 1794 | |
| 1795 | There is a new form, `(current-filename)', which expands out to the |
| 1796 | source file in which it occurs. Combined with the new |
| 1797 | `add-to-load-path', this allows simple scripts to easily add nearby |
| 1798 | directories to the load path. See "Load Paths" in the manual, for more. |
| 1799 | |
| 1800 | ** `random-state-from-platform' |
| 1801 | |
| 1802 | This procedure initializes a random seed using good random sources |
| 1803 | available on your platform, such as /dev/urandom. See "Random Number |
| 1804 | Generation" in the manual, for more. |
| 1805 | |
| 1806 | ** Warn about unsupported `simple-format' options. |
| 1807 | |
| 1808 | The `-Wformat' compilation option now reports unsupported format options |
| 1809 | passed to `simple-format'. |
| 1810 | |
| 1811 | ** Manual updates |
| 1812 | |
| 1813 | Besides the sections already mentioned, the following manual sections |
| 1814 | are new in this release: "Modules and the File System", "Module System |
| 1815 | Reflection", "Syntax Transformer Helpers", and "Local Inclusion". |
| 1816 | |
| 1817 | * New interfaces |
| 1818 | |
| 1819 | ** (ice-9 session): `apropos-hook' |
| 1820 | ** New print option: `escape-newlines', defaults to #t. |
| 1821 | ** (ice-9 ftw): `file-system-fold', `file-system-tree', `scandir' |
| 1822 | ** `scm_c_value_ref': access to multiple returned values from C |
| 1823 | ** scm_call (a varargs version), scm_call_7, scm_call_8, scm_call_9 |
| 1824 | ** Some new syntax helpers in (system syntax) |
| 1825 | |
| 1826 | Search the manual for these identifiers and modules, for more. |
| 1827 | |
| 1828 | * Build fixes |
| 1829 | |
| 1830 | ** FreeBSD build fixes. |
| 1831 | ** OpenBSD compilation fixes. |
| 1832 | ** Solaris 2.10 test suite fixes. |
| 1833 | ** IA64 compilation fix. |
| 1834 | ** MinGW build fixes. |
| 1835 | ** Work around instruction reordering on SPARC and HPPA in the VM. |
| 1836 | ** Gnulib updates: added `dirfd', `setenv' modules. |
| 1837 | |
| 1838 | * Bug fixes |
| 1839 | |
| 1840 | ** Add a deprecated alias for $expt. |
| 1841 | ** Add an exception printer for `getaddrinfo-error'. |
| 1842 | ** Add deprecated shim for `scm_display_error' with stack as first argument. |
| 1843 | ** Add warnings for unsupported `simple-format' options. |
| 1844 | ** Allow overlapping regions to be passed to `bytevector-copy!'. |
| 1845 | ** Better function prologue disassembly |
| 1846 | ** Compiler: fix miscompilation of (values foo ...) in some contexts. |
| 1847 | ** Compiler: fix serialization of #nil-terminated lists. |
| 1848 | ** Compiler: allow values bound in non-tail let expressions to be collected. |
| 1849 | ** Deprecate SCM_ASRTGO. |
| 1850 | ** Document invalidity of (begin) as expression; add back-compat shim. |
| 1851 | ** Don't leak file descriptors when mmaping objcode. |
| 1852 | ** Empty substrings no longer reference the original stringbuf. |
| 1853 | ** FFI: Fix `set-pointer-finalizer!' to leave the type cell unchanged. |
| 1854 | ** FFI: Hold a weak reference to the CIF made by `procedure->pointer'. |
| 1855 | ** FFI: Hold a weak reference to the procedure passed to `procedure->pointer'. |
| 1856 | ** FFI: Properly unpack small integer return values in closure call. |
| 1857 | ** Fix R6RS `fold-left' so the accumulator is the first argument. |
| 1858 | ** Fix bit-set*! bug from 2005. |
| 1859 | ** Fix bug in `make-repl' when `lang' is actually a <language>. |
| 1860 | ** Fix bugs related to mutation, the null string, and shared substrings. |
| 1861 | ** Fix <dynwind> serialization. |
| 1862 | ** Fix erroneous check in `set-procedure-properties!'. |
| 1863 | ** Fix generalized-vector-{ref,set!} for slices. |
| 1864 | ** Fix error messages involving definition forms. |
| 1865 | ** Fix primitive-eval to return #<unspecified> for definitions. |
| 1866 | ** HTTP: Extend handling of "Cache-Control" header. |
| 1867 | ** HTTP: Fix qstring writing of cache-extension values |
| 1868 | ** HTTP: Fix validators for various list-style headers. |
| 1869 | ** HTTP: Permit non-date values for Expires header. |
| 1870 | ** HTTP: `write-request-line' writes absolute paths, not absolute URIs. |
| 1871 | ** Hack the port-column of current-output-port after printing a prompt. |
| 1872 | ** Make sure `regexp-quote' tests use Unicode-capable string ports. |
| 1873 | ** Peval: Fix bugs in the new optimizer. |
| 1874 | ** Statistically unique marks and labels, for robust hygiene across sessions. |
| 1875 | ** Web: Allow URIs with empty authorities, like "file:///etc/hosts". |
| 1876 | ** `,language' at REPL sets the current-language fluid. |
| 1877 | ** `primitive-load' returns the value(s) of the last expression. |
| 1878 | ** `scm_from_stringn' always returns unique strings. |
| 1879 | ** `scm_i_substring_copy' tries to narrow the substring. |
| 1880 | ** i18n: Fix gc_malloc/free mismatch on non-GNU systems. |
| 1881 | |
| 1882 | \f |
| 1883 | Changes in 2.0.3 (since 2.0.2): |
| 1884 | |
| 1885 | * Speed improvements |
| 1886 | |
| 1887 | ** Guile has a new optimizer, `peval'. |
| 1888 | |
| 1889 | `Peval' is a partial evaluator that performs constant folding, dead code |
| 1890 | elimination, copy propagation, and inlining. By default it runs on |
| 1891 | every piece of code that Guile compiles, to fold computations that can |
| 1892 | happen at compile-time, so they don't have to happen at runtime. |
| 1893 | |
| 1894 | If we did our job right, the only impact you would see would be your |
| 1895 | programs getting faster. But if you notice slowdowns or bloated code, |
| 1896 | please send a mail to bug-guile@gnu.org with details. |
| 1897 | |
| 1898 | Thanks to William R. Cook, Oscar Waddell, and Kent Dybvig for inspiring |
| 1899 | peval and its implementation. |
| 1900 | |
| 1901 | You can see what peval does on a given piece of code by running the new |
| 1902 | `,optimize' REPL meta-command, and comparing it to the output of |
| 1903 | `,expand'. See "Compile Commands" in the manual, for more. |
| 1904 | |
| 1905 | ** Fewer calls to `stat'. |
| 1906 | |
| 1907 | Guile now stats only the .go file and the .scm file when loading a fresh |
| 1908 | compiled file. |
| 1909 | |
| 1910 | * Notable changes |
| 1911 | |
| 1912 | ** New module: `(web client)', a simple synchronous web client. |
| 1913 | |
| 1914 | See "Web Client" in the manual, for more. |
| 1915 | |
| 1916 | ** Users can now install compiled `.go' files. |
| 1917 | |
| 1918 | See "Installing Site Packages" in the manual. |
| 1919 | |
| 1920 | ** Remove Front-Cover and Back-Cover text from the manual. |
| 1921 | |
| 1922 | The manual is still under the GNU Free Documentation License, but no |
| 1923 | longer has any invariant sections. |
| 1924 | |
| 1925 | ** More helpful `guild help'. |
| 1926 | |
| 1927 | `guild' is Guile's multi-tool, for use in shell scripting. Now it has a |
| 1928 | nicer interface for querying the set of existing commands, and getting |
| 1929 | help on those commands. Try it out and see! |
| 1930 | |
| 1931 | ** New macro: `define-syntax-rule' |
| 1932 | |
| 1933 | `define-syntax-rule' is a shorthand to make a `syntax-rules' macro with |
| 1934 | one clause. See "Syntax Rules" in the manual, for more. |
| 1935 | |
| 1936 | ** The `,time' REPL meta-command now has more precision. |
| 1937 | |
| 1938 | The output of this command now has microsecond precision, instead of |
| 1939 | 10-millisecond precision. |
| 1940 | |
| 1941 | ** `(ice-9 match)' can now match records. |
| 1942 | |
| 1943 | See "Pattern Matching" in the manual, for more on matching records. |
| 1944 | |
| 1945 | ** New module: `(language tree-il debug)'. |
| 1946 | |
| 1947 | This module provides a tree-il verifier. This is useful for people that |
| 1948 | generate tree-il, usually as part of a language compiler. |
| 1949 | |
| 1950 | ** New functions: `scm_is_exact', `scm_is_inexact'. |
| 1951 | |
| 1952 | These provide a nice C interface for Scheme's `exact?' and `inexact?', |
| 1953 | respectively. |
| 1954 | |
| 1955 | * Bugs fixed |
| 1956 | |
| 1957 | See the git log (or the ChangeLog) for more details on these bugs. |
| 1958 | |
| 1959 | ** Fix order of importing modules and resolving duplicates handlers. |
| 1960 | ** Fix a number of bugs involving extended (merged) generics. |
| 1961 | ** Fix invocation of merge-generics duplicate handler. |
| 1962 | ** Fix write beyond array end in arrays.c. |
| 1963 | ** Fix read beyond end of hashtable size array in hashtab.c. |
| 1964 | ** (web http): Locale-independent parsing and serialization of dates. |
| 1965 | ** Ensure presence of Host header in HTTP/1.1 requests. |
| 1966 | ** Fix take-right and drop-right for improper lists. |
| 1967 | ** Fix leak in get_current_locale(). |
| 1968 | ** Fix recursive define-inlinable expansions. |
| 1969 | ** Check that srfi-1 procedure arguments are procedures. |
| 1970 | ** Fix r6rs `map' for multiple returns. |
| 1971 | ** Fix scm_tmpfile leak on POSIX platforms. |
| 1972 | ** Fix a couple of leaks (objcode->bytecode, make-boot-program). |
| 1973 | ** Fix guile-lib back-compatibility for module-stexi-documentation. |
| 1974 | ** Fix --listen option to allow other ports. |
| 1975 | ** Fix scm_to_latin1_stringn for substrings. |
| 1976 | ** Fix compilation of untyped arrays of rank not 1. |
| 1977 | ** Fix unparse-tree-il of <dynset>. |
| 1978 | ** Fix reading of #||||#. |
| 1979 | ** Fix segfault in GOOPS when class fields are redefined. |
| 1980 | ** Prefer poll(2) over select(2) to allow file descriptors above FD_SETSIZE. |
| 1981 | |
| 1982 | \f |
| 1983 | Changes in 2.0.2 (since 2.0.1): |
| 1984 | |
| 1985 | * Notable changes |
| 1986 | |
| 1987 | ** `guile-tools' renamed to `guild' |
| 1988 | |
| 1989 | The new name is shorter. Its intended future use is for a CPAN-like |
| 1990 | system for Guile wizards and journeyfolk to band together to share code; |
| 1991 | hence the name. `guile-tools' is provided as a backward-compatible |
| 1992 | symbolic link. See "Using Guile Tools" in the manual, for more. |
| 1993 | |
| 1994 | ** New control operators: `shift' and `reset' |
| 1995 | |
| 1996 | See "Shift and Reset" in the manual, for more information. |
| 1997 | |
| 1998 | ** `while' as an expression |
| 1999 | |
| 2000 | Previously the return value of `while' was unspecified. Now its |
| 2001 | values are specified both in the case of normal termination, and via |
| 2002 | termination by invoking `break', possibly with arguments. See "while |
| 2003 | do" in the manual for more. |
| 2004 | |
| 2005 | ** Disallow access to handles of weak hash tables |
| 2006 | |
| 2007 | `hash-get-handle' and `hash-create-handle!' are no longer permitted to |
| 2008 | be called on weak hash tables, because the fields in a weak handle could |
| 2009 | be nulled out by the garbage collector at any time, but yet they are |
| 2010 | otherwise indistinguishable from pairs. Use `hash-ref' and `hash-set!' |
| 2011 | instead. |
| 2012 | |
| 2013 | ** More precision for `get-internal-run-time', `get-internal-real-time' |
| 2014 | |
| 2015 | On 64-bit systems which support POSIX clocks, Guile's internal timing |
| 2016 | procedures offer nanosecond resolution instead of the 10-millisecond |
| 2017 | resolution previously available. 32-bit systems now use 1-millisecond |
| 2018 | timers. |
| 2019 | |
| 2020 | ** Guile now measures time spent in GC |
| 2021 | |
| 2022 | `gc-stats' now returns a meaningful value for `gc-time-taken'. |
| 2023 | |
| 2024 | ** Add `gcprof' |
| 2025 | |
| 2026 | The statprof profiler now exports a `gcprof' procedure, driven by the |
| 2027 | `after-gc-hook', to see which parts of your program are causing GC. Let |
| 2028 | us know if you find it useful. |
| 2029 | |
| 2030 | ** `map', `for-each' and some others now implemented in Scheme |
| 2031 | |
| 2032 | We would not mention this in NEWS, as it is not a user-visible change, |
| 2033 | if it were not for one thing: `map' and `for-each' are no longer |
| 2034 | primitive generics. Instead they are normal bindings, which can be |
| 2035 | wrapped by normal generics. This fixes some modularity issues between |
| 2036 | core `map', SRFI-1 `map', and GOOPS. |
| 2037 | |
| 2038 | Also it's pretty cool that we can do this without a performance impact. |
| 2039 | |
| 2040 | ** Add `scm_peek_byte_or_eof'. |
| 2041 | |
| 2042 | This helper is like `scm_peek_char_or_eof', but for bytes instead of |
| 2043 | full characters. |
| 2044 | |
| 2045 | ** Implement #:stop-at-first-non-option option for getopt-long |
| 2046 | |
| 2047 | See "getopt-long Reference" in the manual, for more information. |
| 2048 | |
| 2049 | ** Improve R6RS conformance for conditions in the I/O libraries |
| 2050 | |
| 2051 | The `(rnrs io simple)' module now raises the correct R6RS conditions in |
| 2052 | error cases. `(rnrs io ports)' is also more correct now, though it is |
| 2053 | still a work in progress. |
| 2054 | |
| 2055 | ** All deprecated routines emit warnings |
| 2056 | |
| 2057 | A few deprecated routines were lacking deprecation warnings. This has |
| 2058 | been fixed now. |
| 2059 | |
| 2060 | * Speed improvements |
| 2061 | |
| 2062 | ** Constants in compiled code now share state better |
| 2063 | |
| 2064 | Constants with shared state, like `("foo")' and `"foo"', now share state |
| 2065 | as much as possible, in the entire compilation unit. This cuts compiled |
| 2066 | `.go' file sizes in half, generally, and speeds startup. |
| 2067 | |
| 2068 | ** VLists: optimize `vlist-fold-right', and add `vhash-fold-right' |
| 2069 | |
| 2070 | These procedures are now twice as fast as they were. |
| 2071 | |
| 2072 | ** UTF-8 ports to bypass `iconv' entirely |
| 2073 | |
| 2074 | This reduces memory usage in a very common case. |
| 2075 | |
| 2076 | ** Compiler speedups |
| 2077 | |
| 2078 | The compiler is now about 40% faster. (Note that this is only the case |
| 2079 | once the compiler is itself compiled, so the build still takes as long |
| 2080 | as it did before.) |
| 2081 | |
| 2082 | ** VM speed tuning |
| 2083 | |
| 2084 | Some assertions that were mostly useful for sanity-checks on the |
| 2085 | bytecode compiler are now off for both "regular" and "debug" engines. |
| 2086 | This together with a fix to cache a TLS access and some other tweaks |
| 2087 | improve the VM's performance by about 20%. |
| 2088 | |
| 2089 | ** SRFI-1 list-set optimizations |
| 2090 | |
| 2091 | lset-adjoin and lset-union now have fast paths for eq? sets. |
| 2092 | |
| 2093 | ** `memq', `memv' optimizations |
| 2094 | |
| 2095 | These procedures are now at least twice as fast than in 2.0.1. |
| 2096 | |
| 2097 | * Deprecations |
| 2098 | |
| 2099 | ** Deprecate scm_whash API |
| 2100 | |
| 2101 | `scm_whash_get_handle', `SCM_WHASHFOUNDP', `SCM_WHASHREF', |
| 2102 | `SCM_WHASHSET', `scm_whash_create_handle', `scm_whash_lookup', and |
| 2103 | `scm_whash_insert' are now deprecated. Use the normal hash table API |
| 2104 | instead. |
| 2105 | |
| 2106 | ** Deprecate scm_struct_table |
| 2107 | |
| 2108 | `SCM_STRUCT_TABLE_NAME', `SCM_SET_STRUCT_TABLE_NAME', |
| 2109 | `SCM_STRUCT_TABLE_CLASS', `SCM_SET_STRUCT_TABLE_CLASS', |
| 2110 | `scm_struct_table', and `scm_struct_create_handle' are now deprecated. |
| 2111 | These routines formed part of the internals of the map between structs |
| 2112 | and classes. |
| 2113 | |
| 2114 | ** Deprecate scm_internal_dynamic_wind |
| 2115 | |
| 2116 | The `scm_t_inner' type and `scm_internal_dynamic_wind' are deprecated, |
| 2117 | as the `scm_dynwind' API is better, and this API encourages users to |
| 2118 | stuff SCM values into pointers. |
| 2119 | |
| 2120 | ** Deprecate scm_immutable_cell, scm_immutable_double_cell |
| 2121 | |
| 2122 | These routines are deprecated, as the GC_STUBBORN API doesn't do |
| 2123 | anything any more. |
| 2124 | |
| 2125 | * Manual updates |
| 2126 | |
| 2127 | Andreas Rottman kindly transcribed the missing parts of the `(rnrs io |
| 2128 | ports)' documentation from the R6RS documentation. Thanks Andreas! |
| 2129 | |
| 2130 | * Bugs fixed |
| 2131 | |
| 2132 | ** Fix double-loading of script in -ds case |
| 2133 | ** -x error message fix |
| 2134 | ** iconveh-related cross-compilation fixes |
| 2135 | ** Fix small integer return value packing on big endian machines. |
| 2136 | ** Fix hash-set! in weak-value table from non-immediate to immediate |
| 2137 | ** Fix call-with-input-file & relatives for multiple values |
| 2138 | ** Fix `hash' for inf and nan |
| 2139 | ** Fix libguile internal type errors caught by typing-strictness==2 |
| 2140 | ** Fix compile error in MinGW fstat socket detection |
| 2141 | ** Fix generation of auto-compiled file names on MinGW |
| 2142 | ** Fix multithreaded access to internal hash tables |
| 2143 | ** Emit a 1-based line number in error messages |
| 2144 | ** Fix define-module ordering |
| 2145 | ** Fix several POSIX functions to use the locale encoding |
| 2146 | ** Add type and range checks to the complex generalized vector accessors |
| 2147 | ** Fix unaligned accesses for bytevectors of complex numbers |
| 2148 | ** Fix '(a #{.} b) |
| 2149 | ** Fix erroneous VM stack overflow for canceled threads |
| 2150 | |
| 2151 | \f |
| 2152 | Changes in 2.0.1 (since 2.0.0): |
| 2153 | |
| 2154 | * Notable changes |
| 2155 | |
| 2156 | ** guile.m4 supports linking with rpath |
| 2157 | |
| 2158 | The GUILE_FLAGS macro now sets GUILE_LIBS and GUILE_LTLIBS, which |
| 2159 | include appropriate directives to the linker to include libguile-2.0.so |
| 2160 | in the runtime library lookup path. |
| 2161 | |
| 2162 | ** `begin' expands macros in its body before other expressions |
| 2163 | |
| 2164 | This enables support for programs like the following: |
| 2165 | |
| 2166 | (begin |
| 2167 | (define even? |
| 2168 | (lambda (x) |
| 2169 | (or (= x 0) (odd? (- x 1))))) |
| 2170 | (define-syntax odd? |
| 2171 | (syntax-rules () |
| 2172 | ((odd? x) (not (even? x))))) |
| 2173 | (even? 10)) |
| 2174 | |
| 2175 | ** REPL reader usability enhancements |
| 2176 | |
| 2177 | The REPL now flushes input after a read error, which should prevent one |
| 2178 | error from causing other errors. The REPL also now interprets comments |
| 2179 | as whitespace. |
| 2180 | |
| 2181 | ** REPL output has configurable width |
| 2182 | |
| 2183 | The REPL now defaults to output with the current terminal's width, in |
| 2184 | columns. See "Debug Commands" in the manual for more information on |
| 2185 | the ,width command. |
| 2186 | |
| 2187 | ** Better C access to the module system |
| 2188 | |
| 2189 | Guile now has convenient C accessors to look up variables or values in |
| 2190 | modules and their public interfaces. See `scm_c_public_ref' and friends |
| 2191 | in "Accessing Modules from C" in the manual. |
| 2192 | |
| 2193 | ** Added `scm_call_5', `scm_call_6' |
| 2194 | |
| 2195 | See "Fly Evaluation" in the manual. |
| 2196 | |
| 2197 | ** Added `scm_from_latin1_keyword', `scm_from_utf8_keyword' |
| 2198 | |
| 2199 | See "Keyword Procedures" in the manual, for more. Note that |
| 2200 | `scm_from_locale_keyword' should not be used when the name is a C string |
| 2201 | constant. |
| 2202 | |
| 2203 | ** R6RS unicode and string I/O work |
| 2204 | |
| 2205 | Added efficient implementations of `get-string-n' and `get-string-n!' |
| 2206 | for binary ports. Exported `current-input-port', `current-output-port' |
| 2207 | and `current-error-port' from `(rnrs io ports)', and enhanced support |
| 2208 | for transcoders. |
| 2209 | |
| 2210 | ** Added `pointer->scm', `scm->pointer' to `(system foreign)' |
| 2211 | |
| 2212 | These procedure are useful if one needs to pass and receive SCM values |
| 2213 | to and from foreign functions. See "Foreign Variables" in the manual, |
| 2214 | for more. |
| 2215 | |
| 2216 | ** Added `heap-allocated-since-gc' to `(gc-stats)' |
| 2217 | |
| 2218 | Also fixed the long-standing bug in the REPL `,stat' command. |
| 2219 | |
| 2220 | ** Add `on-error' REPL option |
| 2221 | |
| 2222 | This option controls what happens when an error occurs at the REPL, and |
| 2223 | defaults to `debug', indicating that Guile should enter the debugger. |
| 2224 | Other values include `report', which will simply print a backtrace |
| 2225 | without entering the debugger. See "System Commands" in the manual. |
| 2226 | |
| 2227 | ** Enforce immutability of string literals |
| 2228 | |
| 2229 | Attempting to mutate a string literal now causes a runtime error. |
| 2230 | |
| 2231 | ** Fix pthread redirection |
| 2232 | |
| 2233 | Guile 2.0.0 shipped with headers that, if configured with pthread |
| 2234 | support, would re-define `pthread_create', `pthread_join', and other API |
| 2235 | to redirect to the BDW-GC wrappers, `GC_pthread_create', etc. This was |
| 2236 | unintended, and not necessary: because threads must enter Guile with |
| 2237 | `scm_with_guile', Guile can handle thread registration itself, without |
| 2238 | needing to make the GC aware of all threads. This oversight has been |
| 2239 | fixed. |
| 2240 | |
| 2241 | ** `with-continuation-barrier' now unwinds on `quit' |
| 2242 | |
| 2243 | A throw to `quit' in a continuation barrier will cause Guile to exit. |
| 2244 | Before, it would do so before unwinding to the barrier, which would |
| 2245 | prevent cleanup handlers from running. This has been fixed so that it |
| 2246 | exits only after unwinding. |
| 2247 | |
| 2248 | ** `string->pointer' and `pointer->string' have optional encoding arg |
| 2249 | |
| 2250 | This allows users of the FFI to more easily deal in strings with |
| 2251 | particular (non-locale) encodings, like "utf-8". See "Void Pointers and |
| 2252 | Byte Access" in the manual, for more. |
| 2253 | |
| 2254 | ** R6RS fixnum arithmetic optimizations |
| 2255 | |
| 2256 | R6RS fixnum operations are are still slower than generic arithmetic, |
| 2257 | however. |
| 2258 | |
| 2259 | ** New procedure: `define-inlinable' |
| 2260 | |
| 2261 | See "Inlinable Procedures" in the manual, for more. |
| 2262 | |
| 2263 | ** New procedure: `exact-integer-sqrt' |
| 2264 | |
| 2265 | See "Integer Operations" in the manual, for more. |
| 2266 | |
| 2267 | ** "Extended read syntax" for symbols parses better |
| 2268 | |
| 2269 | In #{foo}# symbols, backslashes are now treated as escapes, as the |
| 2270 | symbol-printing code intended. Additionally, "\x" within #{foo}# is now |
| 2271 | interpreted as starting an R6RS hex escape. This is backward compatible |
| 2272 | because the symbol printer would never produce a "\x" before. The |
| 2273 | printer also works better too. |
| 2274 | |
| 2275 | ** Added `--fresh-auto-compile' option |
| 2276 | |
| 2277 | This allows a user to invalidate the auto-compilation cache. It's |
| 2278 | usually not needed. See "Compilation" in the manual, for a discussion. |
| 2279 | |
| 2280 | * Manual updates |
| 2281 | |
| 2282 | ** GOOPS documentation updates |
| 2283 | |
| 2284 | ** New man page |
| 2285 | |
| 2286 | Thanks to Mark Harig for improvements to guile.1. |
| 2287 | |
| 2288 | ** SRFI-23 documented |
| 2289 | |
| 2290 | The humble `error' SRFI now has an entry in the manual. |
| 2291 | |
| 2292 | * New modules |
| 2293 | |
| 2294 | ** `(ice-9 binary-ports)': "R6RS I/O Ports", in the manual |
| 2295 | ** `(ice-9 eval-string)': "Fly Evaluation", in the manual |
| 2296 | ** `(ice-9 command-line)', not documented yet |
| 2297 | |
| 2298 | * Bugs fixed |
| 2299 | |
| 2300 | ** Fixed `iconv_t' memory leak on close-port |
| 2301 | ** Fixed some leaks with weak hash tables |
| 2302 | ** Export `vhash-delq' and `vhash-delv' from `(ice-9 vlist)' |
| 2303 | ** `after-gc-hook' works again |
| 2304 | ** `define-record-type' now allowed in nested contexts |
| 2305 | ** `exact-integer-sqrt' now handles large integers correctly |
| 2306 | ** Fixed C extension examples in manual |
| 2307 | ** `vhash-delete' honors HASH argument |
| 2308 | ** Make `locale-digit-grouping' more robust |
| 2309 | ** Default exception printer robustness fixes |
| 2310 | ** Fix presence of non-I CPPFLAGS in `guile-2.0.pc' |
| 2311 | ** `read' updates line/column numbers when reading SCSH block comments |
| 2312 | ** Fix imports of multiple custom interfaces of same module |
| 2313 | ** Fix encoding scanning for non-seekable ports |
| 2314 | ** Fix `setter' when called with a non-setter generic |
| 2315 | ** Fix f32 and f64 bytevectors to not accept rationals |
| 2316 | ** Fix description of the R6RS `finite?' in manual |
| 2317 | ** Quotient, remainder and modulo accept inexact integers again |
| 2318 | ** Fix `continue' within `while' to take zero arguments |
| 2319 | ** Fix alignment for structures in FFI |
| 2320 | ** Fix port-filename of stdin, stdout, stderr to match the docs |
| 2321 | ** Fix weak hash table-related bug in `define-wrapped-pointer-type' |
| 2322 | ** Fix partial continuation application with pending procedure calls |
| 2323 | ** scm_{to,from}_locale_string use current locale, not current ports |
| 2324 | ** Fix thread cleanup, by using a pthread_key destructor |
| 2325 | ** Fix `quit' at the REPL |
| 2326 | ** Fix a failure to sync regs in vm bytevector ops |
| 2327 | ** Fix (texinfo reflection) to handle nested structures like syntax patterns |
| 2328 | ** Fix stexi->html double translation |
| 2329 | ** Fix tree-il->scheme fix for <prompt> |
| 2330 | ** Fix compilation of <prompt> in <fix> in single-value context |
| 2331 | ** Fix race condition in ensure-writable-dir |
| 2332 | ** Fix error message on ,disassemble "non-procedure" |
| 2333 | ** Fix prompt and abort with the boot evaluator |
| 2334 | ** Fix `procedure->pointer' for functions returning `void' |
| 2335 | ** Fix error reporting in dynamic-pointer |
| 2336 | ** Fix problems detecting coding: in block comments |
| 2337 | ** Fix duplicate load-path and load-compiled-path in compilation environment |
| 2338 | ** Add fallback read(2) suppport for .go files if mmap(2) unavailable |
| 2339 | ** Fix c32vector-set!, c64vector-set! |
| 2340 | ** Fix mistakenly deprecated read syntax for uniform complex vectors |
| 2341 | ** Fix parsing of exact numbers with negative exponents |
| 2342 | ** Ignore SIGPIPE in (system repl server) |
| 2343 | ** Fix optional second arg to R6RS log function |
| 2344 | ** Fix R6RS `assert' to return true value. |
| 2345 | ** Fix fencepost error when seeking in bytevector input ports |
| 2346 | ** Gracefully handle `setlocale' errors when starting the REPL |
| 2347 | ** Improve support of the `--disable-posix' configure option |
| 2348 | ** Make sure R6RS binary ports pass `binary-port?' regardless of the locale |
| 2349 | ** Gracefully handle unterminated UTF-8 sequences instead of hitting an `assert' |
| 2350 | |
| 2351 | |
| 2352 | \f |
| 2353 | Changes in 2.0.0 (changes since the 1.8.x series): |
| 2354 | |
| 2355 | * New modules (see the manual for details) |
| 2356 | |
| 2357 | ** `(srfi srfi-18)', more sophisticated multithreading support |
| 2358 | ** `(srfi srfi-27)', sources of random bits |
| 2359 | ** `(srfi srfi-38)', External Representation for Data With Shared Structure |
| 2360 | ** `(srfi srfi-42)', eager comprehensions |
| 2361 | ** `(srfi srfi-45)', primitives for expressing iterative lazy algorithms |
| 2362 | ** `(srfi srfi-67)', compare procedures |
| 2363 | ** `(ice-9 i18n)', internationalization support |
| 2364 | ** `(ice-9 futures)', fine-grain parallelism |
| 2365 | ** `(rnrs bytevectors)', the R6RS bytevector API |
| 2366 | ** `(rnrs io ports)', a subset of the R6RS I/O port API |
| 2367 | ** `(system xref)', a cross-referencing facility (FIXME undocumented) |
| 2368 | ** `(ice-9 vlist)', lists with constant-time random access; hash lists |
| 2369 | ** `(system foreign)', foreign function interface |
| 2370 | ** `(sxml match)', a pattern matcher for SXML |
| 2371 | ** `(srfi srfi-9 gnu)', extensions to the SRFI-9 record library |
| 2372 | ** `(system vm coverage)', a line-by-line code coverage library |
| 2373 | ** `(web uri)', URI data type, parser, and unparser |
| 2374 | ** `(web http)', HTTP header parsers and unparsers |
| 2375 | ** `(web request)', HTTP request data type, reader, and writer |
| 2376 | ** `(web response)', HTTP response data type, reader, and writer |
| 2377 | ** `(web server)', Generic HTTP server |
| 2378 | ** `(ice-9 poll)', a poll wrapper |
| 2379 | ** `(web server http)', HTTP-over-TCP web server implementation |
| 2380 | |
| 2381 | ** Replaced `(ice-9 match)' with Alex Shinn's compatible, hygienic matcher. |
| 2382 | |
| 2383 | Guile's copy of Andrew K. Wright's `match' library has been replaced by |
| 2384 | a compatible hygienic implementation by Alex Shinn. It is now |
| 2385 | documented, see "Pattern Matching" in the manual. |
| 2386 | |
| 2387 | Compared to Andrew K. Wright's `match', the new `match' lacks |
| 2388 | `match-define', `match:error-control', `match:set-error-control', |
| 2389 | `match:error', `match:set-error', and all structure-related procedures. |
| 2390 | |
| 2391 | ** Imported statprof, SSAX, and texinfo modules from Guile-Lib |
| 2392 | |
| 2393 | The statprof statistical profiler, the SSAX XML toolkit, and the texinfo |
| 2394 | toolkit from Guile-Lib have been imported into Guile proper. See |
| 2395 | "Standard Library" in the manual for more details. |
| 2396 | |
| 2397 | ** Integration of lalr-scm, a parser generator |
| 2398 | |
| 2399 | Guile has included Dominique Boucher's fine `lalr-scm' parser generator |
| 2400 | as `(system base lalr)'. See "LALR(1) Parsing" in the manual, for more |
| 2401 | information. |
| 2402 | |
| 2403 | * Changes to the stand-alone interpreter |
| 2404 | |
| 2405 | ** Guile now can compile Scheme to bytecode for a custom virtual machine. |
| 2406 | |
| 2407 | Compiled code loads much faster than Scheme source code, and runs around |
| 2408 | 3 or 4 times as fast, generating much less garbage in the process. |
| 2409 | |
| 2410 | ** Evaluating Scheme code does not use the C stack. |
| 2411 | |
| 2412 | Besides when compiling Guile itself, Guile no longer uses a recursive C |
| 2413 | function as an evaluator. This obviates the need to check the C stack |
| 2414 | pointer for overflow. Continuations still capture the C stack, however. |
| 2415 | |
| 2416 | ** New environment variables: GUILE_LOAD_COMPILED_PATH, |
| 2417 | GUILE_SYSTEM_LOAD_COMPILED_PATH |
| 2418 | |
| 2419 | GUILE_LOAD_COMPILED_PATH is for compiled files what GUILE_LOAD_PATH is |
| 2420 | for source files. It is a different path, however, because compiled |
| 2421 | files are architecture-specific. GUILE_SYSTEM_LOAD_COMPILED_PATH is like |
| 2422 | GUILE_SYSTEM_PATH. |
| 2423 | |
| 2424 | ** New read-eval-print loop (REPL) implementation |
| 2425 | |
| 2426 | Running Guile with no arguments drops the user into the new REPL. See |
| 2427 | "Using Guile Interactively" in the manual, for more information. |
| 2428 | |
| 2429 | ** Remove old Emacs interface |
| 2430 | |
| 2431 | Guile had an unused `--emacs' command line argument that was supposed to |
| 2432 | help when running Guile inside Emacs. This option has been removed, and |
| 2433 | the helper functions `named-module-use!' and `load-emacs-interface' have |
| 2434 | been deprecated. |
| 2435 | |
| 2436 | ** Add `(system repl server)' module and `--listen' command-line argument |
| 2437 | |
| 2438 | The `(system repl server)' module exposes procedures to listen on |
| 2439 | sockets for connections, and serve REPLs to those clients. The --listen |
| 2440 | command-line argument allows any Guile program to thus be remotely |
| 2441 | debuggable. |
| 2442 | |
| 2443 | See "Invoking Guile" for more information on `--listen'. |
| 2444 | |
| 2445 | ** Command line additions |
| 2446 | |
| 2447 | The guile binary now supports a new switch "-x", which can be used to |
| 2448 | extend the list of filename extensions tried when loading files |
| 2449 | (%load-extensions). |
| 2450 | |
| 2451 | ** New reader options: `square-brackets', `r6rs-hex-escapes', |
| 2452 | `hungry-eol-escapes' |
| 2453 | |
| 2454 | The reader supports a new option (changeable via `read-options'), |
| 2455 | `square-brackets', which instructs it to interpret square brackets as |
| 2456 | parentheses. This option is on by default. |
| 2457 | |
| 2458 | When the new `r6rs-hex-escapes' reader option is enabled, the reader |
| 2459 | will recognize string escape sequences as defined in R6RS. R6RS string |
| 2460 | escape sequences are incompatible with Guile's existing escapes, though, |
| 2461 | so this option is off by default. |
| 2462 | |
| 2463 | Additionally, Guile follows the R6RS newline escaping rules when the |
| 2464 | `hungry-eol-escapes' option is enabled. |
| 2465 | |
| 2466 | See "String Syntax" in the manual, for more information. |
| 2467 | |
| 2468 | ** Function profiling and tracing at the REPL |
| 2469 | |
| 2470 | The `,profile FORM' REPL meta-command can now be used to statistically |
| 2471 | profile execution of a form, to see which functions are taking the most |
| 2472 | time. See `,help profile' for more information. |
| 2473 | |
| 2474 | Similarly, `,trace FORM' traces all function applications that occur |
| 2475 | during the execution of `FORM'. See `,help trace' for more information. |
| 2476 | |
| 2477 | ** Recursive debugging REPL on error |
| 2478 | |
| 2479 | When Guile sees an error at the REPL, instead of saving the stack, Guile |
| 2480 | will directly enter a recursive REPL in the dynamic context of the |
| 2481 | error. See "Error Handling" in the manual, for more information. |
| 2482 | |
| 2483 | A recursive REPL is the same as any other REPL, except that it |
| 2484 | has been augmented with debugging information, so that one can inspect |
| 2485 | the context of the error. The debugger has been integrated with the REPL |
| 2486 | via a set of debugging meta-commands. |
| 2487 | |
| 2488 | For example, one may access a backtrace with `,backtrace' (or |
| 2489 | `,bt'). See "Interactive Debugging" in the manual, for more |
| 2490 | information. |
| 2491 | |
| 2492 | ** New `guile-tools' commands: `compile', `disassemble' |
| 2493 | |
| 2494 | Pass the `--help' command-line option to these commands for more |
| 2495 | information. |
| 2496 | |
| 2497 | ** Guile now adds its install prefix to the LTDL_LIBRARY_PATH |
| 2498 | |
| 2499 | Users may now install Guile to nonstandard prefixes and just run |
| 2500 | `/path/to/bin/guile', instead of also having to set LTDL_LIBRARY_PATH to |
| 2501 | include `/path/to/lib'. |
| 2502 | |
| 2503 | ** Guile's Emacs integration is now more keyboard-friendly |
| 2504 | |
| 2505 | Backtraces may now be disclosed with the keyboard in addition to the |
| 2506 | mouse. |
| 2507 | |
| 2508 | ** Load path change: search in version-specific paths before site paths |
| 2509 | |
| 2510 | When looking for a module, Guile now searches first in Guile's |
| 2511 | version-specific path (the library path), *then* in the site dir. This |
| 2512 | allows Guile's copy of SSAX to override any Guile-Lib copy the user has |
| 2513 | installed. Also it should cut the number of `stat' system calls by half, |
| 2514 | in the common case. |
| 2515 | |
| 2516 | ** Value history in the REPL on by default |
| 2517 | |
| 2518 | By default, the REPL will save computed values in variables like `$1', |
| 2519 | `$2', and the like. There are programmatic and interactive interfaces to |
| 2520 | control this. See "Value History" in the manual, for more information. |
| 2521 | |
| 2522 | ** Readline tab completion for arguments |
| 2523 | |
| 2524 | When readline is enabled, tab completion works for arguments too, not |
| 2525 | just for the operator position. |
| 2526 | |
| 2527 | ** Expression-oriented readline history |
| 2528 | |
| 2529 | Guile's readline history now tries to operate on expressions instead of |
| 2530 | input lines. Let us know what you think! |
| 2531 | |
| 2532 | ** Interactive Guile follows GNU conventions |
| 2533 | |
| 2534 | As recommended by the GPL, Guile now shows a brief copyright and |
| 2535 | warranty disclaimer on startup, along with pointers to more information. |
| 2536 | |
| 2537 | * Changes to Scheme functions and syntax |
| 2538 | |
| 2539 | ** Support for R6RS libraries |
| 2540 | |
| 2541 | The `library' and `import' forms from the latest Scheme report have been |
| 2542 | added to Guile, in such a way that R6RS libraries share a namespace with |
| 2543 | Guile modules. R6RS modules may import Guile modules, and are available |
| 2544 | for Guile modules to import via use-modules and all the rest. See "R6RS |
| 2545 | Libraries" in the manual for more information. |
| 2546 | |
| 2547 | ** Implementations of R6RS libraries |
| 2548 | |
| 2549 | Guile now has implementations for all of the libraries defined in the |
| 2550 | R6RS. Thanks to Julian Graham for this excellent hack. See "R6RS |
| 2551 | Standard Libraries" in the manual for a full list of libraries. |
| 2552 | |
| 2553 | ** Partial R6RS compatibility |
| 2554 | |
| 2555 | Guile now has enough support for R6RS to run a reasonably large subset |
| 2556 | of R6RS programs. |
| 2557 | |
| 2558 | Guile is not fully R6RS compatible. Many incompatibilities are simply |
| 2559 | bugs, though some parts of Guile will remain R6RS-incompatible for the |
| 2560 | foreseeable future. See "R6RS Incompatibilities" in the manual, for more |
| 2561 | information. |
| 2562 | |
| 2563 | Please contact bug-guile@gnu.org if you have found an issue not |
| 2564 | mentioned in that compatibility list. |
| 2565 | |
| 2566 | ** New implementation of `primitive-eval' |
| 2567 | |
| 2568 | Guile's `primitive-eval' is now implemented in Scheme. Actually there is |
| 2569 | still a C evaluator, used when building a fresh Guile to interpret the |
| 2570 | compiler, so we can compile eval.scm. Thereafter all calls to |
| 2571 | primitive-eval are implemented by VM-compiled code. |
| 2572 | |
| 2573 | This allows all of Guile's procedures, be they interpreted or compiled, |
| 2574 | to execute on the same stack, unifying multiple-value return semantics, |
| 2575 | providing for proper tail recursion between interpreted and compiled |
| 2576 | code, and simplifying debugging. |
| 2577 | |
| 2578 | As part of this change, the evaluator no longer mutates the internal |
| 2579 | representation of the code being evaluated in a thread-unsafe manner. |
| 2580 | |
| 2581 | There are two negative aspects of this change, however. First, Guile |
| 2582 | takes a lot longer to compile now. Also, there is less debugging |
| 2583 | information available for debugging interpreted code. We hope to improve |
| 2584 | both of these situations. |
| 2585 | |
| 2586 | There are many changes to the internal C evalator interface, but all |
| 2587 | public interfaces should be the same. See the ChangeLog for details. If |
| 2588 | we have inadvertantly changed an interface that you were using, please |
| 2589 | contact bug-guile@gnu.org. |
| 2590 | |
| 2591 | ** Procedure removed: `the-environment' |
| 2592 | |
| 2593 | This procedure was part of the interpreter's execution model, and does |
| 2594 | not apply to the compiler. |
| 2595 | |
| 2596 | ** No more `local-eval' |
| 2597 | |
| 2598 | `local-eval' used to exist so that one could evaluate code in the |
| 2599 | lexical context of a function. Since there is no way to get the lexical |
| 2600 | environment any more, as that concept has no meaning for the compiler, |
| 2601 | and a different meaning for the interpreter, we have removed the |
| 2602 | function. |
| 2603 | |
| 2604 | If you think you need `local-eval', you should probably implement your |
| 2605 | own metacircular evaluator. It will probably be as fast as Guile's |
| 2606 | anyway. |
| 2607 | |
| 2608 | ** Scheme source files will now be compiled automatically. |
| 2609 | |
| 2610 | If a compiled .go file corresponding to a .scm file is not found or is |
| 2611 | not fresh, the .scm file will be compiled on the fly, and the resulting |
| 2612 | .go file stored away. An advisory note will be printed on the console. |
| 2613 | |
| 2614 | Note that this mechanism depends on the timestamp of the .go file being |
| 2615 | newer than that of the .scm file; if the .scm or .go files are moved |
| 2616 | after installation, care should be taken to preserve their original |
| 2617 | timestamps. |
| 2618 | |
| 2619 | Auto-compiled files will be stored in the $XDG_CACHE_HOME/guile/ccache |
| 2620 | directory, where $XDG_CACHE_HOME defaults to ~/.cache. This directory |
| 2621 | will be created if needed. |
| 2622 | |
| 2623 | To inhibit automatic compilation, set the GUILE_AUTO_COMPILE environment |
| 2624 | variable to 0, or pass --no-auto-compile on the Guile command line. |
| 2625 | |
| 2626 | ** New POSIX procedures: `getrlimit' and `setrlimit' |
| 2627 | |
| 2628 | Note however that the interface of these functions is likely to change |
| 2629 | in the next prerelease. |
| 2630 | |
| 2631 | ** New POSIX procedure: `getsid' |
| 2632 | |
| 2633 | Scheme binding for the `getsid' C library call. |
| 2634 | |
| 2635 | ** New POSIX procedure: `getaddrinfo' |
| 2636 | |
| 2637 | Scheme binding for the `getaddrinfo' C library function. |
| 2638 | |
| 2639 | ** Multicast socket options |
| 2640 | |
| 2641 | Support was added for the IP_MULTICAST_TTL and IP_MULTICAST_IF socket |
| 2642 | options. See "Network Sockets and Communication" in the manual, for |
| 2643 | more information. |
| 2644 | |
| 2645 | ** `recv!', `recvfrom!', `send', `sendto' now deal in bytevectors |
| 2646 | |
| 2647 | These socket procedures now take bytevectors as arguments, instead of |
| 2648 | strings. There is some deprecated string support, however. |
| 2649 | |
| 2650 | ** New GNU procedures: `setaffinity' and `getaffinity'. |
| 2651 | |
| 2652 | See "Processes" in the manual, for more information. |
| 2653 | |
| 2654 | ** New procedures: `compose', `negate', and `const' |
| 2655 | |
| 2656 | See "Higher-Order Functions" in the manual, for more information. |
| 2657 | |
| 2658 | ** New procedure in `(oops goops)': `method-formals' |
| 2659 | |
| 2660 | ** New procedures in (ice-9 session): `add-value-help-handler!', |
| 2661 | `remove-value-help-handler!', `add-name-help-handler!' |
| 2662 | `remove-name-help-handler!', `procedure-arguments' |
| 2663 | |
| 2664 | The value and name help handlers provide some minimal extensibility to |
| 2665 | the help interface. Guile-lib's `(texinfo reflection)' uses them, for |
| 2666 | example, to make stexinfo help documentation available. See those |
| 2667 | procedures' docstrings for more information. |
| 2668 | |
| 2669 | `procedure-arguments' describes the arguments that a procedure can take, |
| 2670 | combining arity and formals. For example: |
| 2671 | |
| 2672 | (procedure-arguments resolve-interface) |
| 2673 | => ((required . (name)) (rest . args)) |
| 2674 | |
| 2675 | Additionally, `module-commentary' is now publically exported from |
| 2676 | `(ice-9 session). |
| 2677 | |
| 2678 | ** Removed: `procedure->memoizing-macro', `procedure->syntax' |
| 2679 | |
| 2680 | These procedures created primitive fexprs for the old evaluator, and are |
| 2681 | no longer supported. If you feel that you need these functions, you |
| 2682 | probably need to write your own metacircular evaluator (which will |
| 2683 | probably be as fast as Guile's, anyway). |
| 2684 | |
| 2685 | ** New language: ECMAScript |
| 2686 | |
| 2687 | Guile now ships with one other high-level language supported, |
| 2688 | ECMAScript. The goal is to support all of version 3.1 of the standard, |
| 2689 | but not all of the libraries are there yet. This support is not yet |
| 2690 | documented; ask on the mailing list if you are interested. |
| 2691 | |
| 2692 | ** New language: Brainfuck |
| 2693 | |
| 2694 | Brainfuck is a toy language that closely models Turing machines. Guile's |
| 2695 | brainfuck compiler is meant to be an example of implementing other |
| 2696 | languages. See the manual for details, or |
| 2697 | http://en.wikipedia.org/wiki/Brainfuck for more information about the |
| 2698 | Brainfuck language itself. |
| 2699 | |
| 2700 | ** New language: Elisp |
| 2701 | |
| 2702 | Guile now has an experimental Emacs Lisp compiler and runtime. You can |
| 2703 | now switch to Elisp at the repl: `,language elisp'. All kudos to Daniel |
| 2704 | Kraft and Brian Templeton, and all bugs to bug-guile@gnu.org. |
| 2705 | |
| 2706 | ** Better documentation infrastructure for macros |
| 2707 | |
| 2708 | It is now possible to introspect on the type of a macro, e.g. |
| 2709 | syntax-rules, identifier-syntax, etc, and extract information about that |
| 2710 | macro, such as the syntax-rules patterns or the defmacro arguments. |
| 2711 | `(texinfo reflection)' takes advantage of this to give better macro |
| 2712 | documentation. |
| 2713 | |
| 2714 | ** Support for arbitrary procedure metadata |
| 2715 | |
| 2716 | Building on its support for docstrings, Guile now supports multiple |
| 2717 | docstrings, adding them to the tail of a compiled procedure's |
| 2718 | properties. For example: |
| 2719 | |
| 2720 | (define (foo) |
| 2721 | "one" |
| 2722 | "two" |
| 2723 | 3) |
| 2724 | (procedure-properties foo) |
| 2725 | => ((name . foo) (documentation . "one") (documentation . "two")) |
| 2726 | |
| 2727 | Also, vectors of pairs are now treated as additional metadata entries: |
| 2728 | |
| 2729 | (define (bar) |
| 2730 | #((quz . #f) (docstring . "xyzzy")) |
| 2731 | 3) |
| 2732 | (procedure-properties bar) |
| 2733 | => ((name . bar) (quz . #f) (docstring . "xyzzy")) |
| 2734 | |
| 2735 | This allows arbitrary literals to be embedded as metadata in a compiled |
| 2736 | procedure. |
| 2737 | |
| 2738 | ** The psyntax expander now knows how to interpret the @ and @@ special |
| 2739 | forms. |
| 2740 | |
| 2741 | ** The psyntax expander is now hygienic with respect to modules. |
| 2742 | |
| 2743 | Free variables in a macro are scoped in the module that the macro was |
| 2744 | defined in, not in the module the macro is used in. For example, code |
| 2745 | like this works now: |
| 2746 | |
| 2747 | (define-module (foo) #:export (bar)) |
| 2748 | (define (helper x) ...) |
| 2749 | (define-syntax bar |
| 2750 | (syntax-rules () ((_ x) (helper x)))) |
| 2751 | |
| 2752 | (define-module (baz) #:use-module (foo)) |
| 2753 | (bar qux) |
| 2754 | |
| 2755 | It used to be you had to export `helper' from `(foo)' as well. |
| 2756 | Thankfully, this has been fixed. |
| 2757 | |
| 2758 | ** Support for version information in Guile's `module' form |
| 2759 | |
| 2760 | Guile modules now have a `#:version' field. See "R6RS Version |
| 2761 | References", "General Information about Modules", "Using Guile Modules", |
| 2762 | and "Creating Guile Modules" in the manual for more information. |
| 2763 | |
| 2764 | ** Support for renaming bindings on module export |
| 2765 | |
| 2766 | Wherever Guile accepts a symbol as an argument to specify a binding to |
| 2767 | export, it now also accepts a pair of symbols, indicating that a binding |
| 2768 | should be renamed on export. See "Creating Guile Modules" in the manual |
| 2769 | for more information. |
| 2770 | |
| 2771 | ** New procedure: `module-export-all!' |
| 2772 | |
| 2773 | This procedure exports all current and future bindings from a module. |
| 2774 | Use as `(module-export-all! (current-module))'. |
| 2775 | |
| 2776 | ** New procedure `reload-module', and `,reload' REPL command |
| 2777 | |
| 2778 | See "Module System Reflection" and "Module Commands" in the manual, for |
| 2779 | more information. |
| 2780 | |
| 2781 | ** `eval-case' has been deprecated, and replaced by `eval-when'. |
| 2782 | |
| 2783 | The semantics of `eval-when' are easier to understand. See "Eval When" |
| 2784 | in the manual, for more information. |
| 2785 | |
| 2786 | ** Guile is now more strict about prohibiting definitions in expression |
| 2787 | contexts. |
| 2788 | |
| 2789 | Although previous versions of Guile accepted it, the following |
| 2790 | expression is not valid, in R5RS or R6RS: |
| 2791 | |
| 2792 | (if test (define foo 'bar) (define foo 'baz)) |
| 2793 | |
| 2794 | In this specific case, it would be better to do: |
| 2795 | |
| 2796 | (define foo (if test 'bar 'baz)) |
| 2797 | |
| 2798 | It is possible to circumvent this restriction with e.g. |
| 2799 | `(module-define! (current-module) 'foo 'baz)'. Contact the list if you |
| 2800 | have any questions. |
| 2801 | |
| 2802 | ** Support for `letrec*' |
| 2803 | |
| 2804 | Guile now supports `letrec*', a recursive lexical binding operator in |
| 2805 | which the identifiers are bound in order. See "Local Bindings" in the |
| 2806 | manual, for more details. |
| 2807 | |
| 2808 | ** Internal definitions now expand to `letrec*' |
| 2809 | |
| 2810 | Following the R6RS, internal definitions now expand to letrec* instead |
| 2811 | of letrec. The following program is invalid for R5RS, but valid for |
| 2812 | R6RS: |
| 2813 | |
| 2814 | (define (foo) |
| 2815 | (define bar 10) |
| 2816 | (define baz (+ bar 20)) |
| 2817 | baz) |
| 2818 | |
| 2819 | ;; R5RS and Guile <= 1.8: |
| 2820 | (foo) => Unbound variable: bar |
| 2821 | ;; R6RS and Guile >= 2.0: |
| 2822 | (foo) => 30 |
| 2823 | |
| 2824 | This change should not affect correct R5RS programs, or programs written |
| 2825 | in earlier Guile dialects. |
| 2826 | |
| 2827 | ** Macro expansion produces structures instead of s-expressions |
| 2828 | |
| 2829 | In the olden days, macroexpanding an s-expression would yield another |
| 2830 | s-expression. Though the lexical variables were renamed, expansions of |
| 2831 | core forms like `if' and `begin' were still non-hygienic, as they relied |
| 2832 | on the toplevel definitions of `if' et al being the conventional ones. |
| 2833 | |
| 2834 | The solution is to expand to structures instead of s-expressions. There |
| 2835 | is an `if' structure, a `begin' structure, a `toplevel-ref' structure, |
| 2836 | etc. The expander already did this for compilation, producing Tree-IL |
| 2837 | directly; it has been changed now to do so when expanding for the |
| 2838 | evaluator as well. |
| 2839 | |
| 2840 | ** Defmacros must now produce valid Scheme expressions. |
| 2841 | |
| 2842 | It used to be that defmacros could unquote in Scheme values, as a way of |
| 2843 | supporting partial evaluation, and avoiding some hygiene issues. For |
| 2844 | example: |
| 2845 | |
| 2846 | (define (helper x) ...) |
| 2847 | (define-macro (foo bar) |
| 2848 | `(,helper ,bar)) |
| 2849 | |
| 2850 | Assuming this macro is in the `(baz)' module, the direct translation of |
| 2851 | this code would be: |
| 2852 | |
| 2853 | (define (helper x) ...) |
| 2854 | (define-macro (foo bar) |
| 2855 | `((@@ (baz) helper) ,bar)) |
| 2856 | |
| 2857 | Of course, one could just use a hygienic macro instead: |
| 2858 | |
| 2859 | (define-syntax foo |
| 2860 | (syntax-rules () |
| 2861 | ((_ bar) (helper bar)))) |
| 2862 | |
| 2863 | ** Guile's psyntax now supports docstrings and internal definitions. |
| 2864 | |
| 2865 | The following Scheme is not strictly legal: |
| 2866 | |
| 2867 | (define (foo) |
| 2868 | "bar" |
| 2869 | (define (baz) ...) |
| 2870 | (baz)) |
| 2871 | |
| 2872 | However its intent is fairly clear. Guile interprets "bar" to be the |
| 2873 | docstring of `foo', and the definition of `baz' is still in definition |
| 2874 | context. |
| 2875 | |
| 2876 | ** Support for settable identifier syntax |
| 2877 | |
| 2878 | Following the R6RS, "variable transformers" are settable |
| 2879 | identifier-syntax. See "Identifier macros" in the manual, for more |
| 2880 | information. |
| 2881 | |
| 2882 | ** syntax-case treats `_' as a placeholder |
| 2883 | |
| 2884 | Following R6RS, a `_' in a syntax-rules or syntax-case pattern matches |
| 2885 | anything, and binds no pattern variables. Unlike the R6RS, Guile also |
| 2886 | permits `_' to be in the literals list for a pattern. |
| 2887 | |
| 2888 | ** Macros need to be defined before their first use. |
| 2889 | |
| 2890 | It used to be that with lazy memoization, this might work: |
| 2891 | |
| 2892 | (define (foo x) |
| 2893 | (ref x)) |
| 2894 | (define-macro (ref x) x) |
| 2895 | (foo 1) => 1 |
| 2896 | |
| 2897 | But now, the body of `foo' is interpreted to mean a call to the toplevel |
| 2898 | `ref' function, instead of a macro expansion. The solution is to define |
| 2899 | macros before code that uses them. |
| 2900 | |
| 2901 | ** Functions needed by macros at expand-time need to be present at |
| 2902 | expand-time. |
| 2903 | |
| 2904 | For example, this code will work at the REPL: |
| 2905 | |
| 2906 | (define (double-helper x) (* x x)) |
| 2907 | (define-macro (double-literal x) (double-helper x)) |
| 2908 | (double-literal 2) => 4 |
| 2909 | |
| 2910 | But it will not work when a file is compiled, because the definition of |
| 2911 | `double-helper' is not present at expand-time. The solution is to wrap |
| 2912 | the definition of `double-helper' in `eval-when': |
| 2913 | |
| 2914 | (eval-when (load compile eval) |
| 2915 | (define (double-helper x) (* x x))) |
| 2916 | (define-macro (double-literal x) (double-helper x)) |
| 2917 | (double-literal 2) => 4 |
| 2918 | |
| 2919 | See the documentation for eval-when for more information. |
| 2920 | |
| 2921 | ** `macroexpand' produces structures, not S-expressions. |
| 2922 | |
| 2923 | Given the need to maintain referential transparency, both lexically and |
| 2924 | modular, the result of expanding Scheme expressions is no longer itself |
| 2925 | an s-expression. If you want a human-readable approximation of the |
| 2926 | result of `macroexpand', call `tree-il->scheme' from `(language |
| 2927 | tree-il)'. |
| 2928 | |
| 2929 | ** Removed function: `macroexpand-1' |
| 2930 | |
| 2931 | It is unclear how to implement `macroexpand-1' with syntax-case, though |
| 2932 | PLT Scheme does prove that it is possible. |
| 2933 | |
| 2934 | ** New reader macros: #' #` #, #,@ |
| 2935 | |
| 2936 | These macros translate, respectively, to `syntax', `quasisyntax', |
| 2937 | `unsyntax', and `unsyntax-splicing'. See the R6RS for more information. |
| 2938 | These reader macros may be overridden by `read-hash-extend'. |
| 2939 | |
| 2940 | ** Incompatible change to #' |
| 2941 | |
| 2942 | Guile did have a #' hash-extension, by default, which just returned the |
| 2943 | subsequent datum: #'foo => foo. In the unlikely event that anyone |
| 2944 | actually used this, this behavior may be reinstated via the |
| 2945 | `read-hash-extend' mechanism. |
| 2946 | |
| 2947 | ** `unquote' and `unquote-splicing' accept multiple expressions |
| 2948 | |
| 2949 | As per the R6RS, these syntax operators can now accept any number of |
| 2950 | expressions to unquote. |
| 2951 | |
| 2952 | ** Scheme expresssions may be commented out with #; |
| 2953 | |
| 2954 | #; comments out an entire expression. See SRFI-62 or the R6RS for more |
| 2955 | information. |
| 2956 | |
| 2957 | ** Prompts: Delimited, composable continuations |
| 2958 | |
| 2959 | Guile now has prompts as part of its primitive language. See "Prompts" |
| 2960 | in the manual, for more information. |
| 2961 | |
| 2962 | Expressions entered in at the REPL, or from the command line, are |
| 2963 | surrounded by a prompt with the default prompt tag. |
| 2964 | |
| 2965 | ** `make-stack' with a tail-called procedural narrowing argument no longer |
| 2966 | works (with compiled procedures) |
| 2967 | |
| 2968 | It used to be the case that a captured stack could be narrowed to select |
| 2969 | calls only up to or from a certain procedure, even if that procedure |
| 2970 | already tail-called another procedure. This was because the debug |
| 2971 | information from the original procedure was kept on the stack. |
| 2972 | |
| 2973 | Now with the new compiler, the stack only contains active frames from |
| 2974 | the current continuation. A narrow to a procedure that is not in the |
| 2975 | stack will result in an empty stack. To fix this, narrow to a procedure |
| 2976 | that is active in the current continuation, or narrow to a specific |
| 2977 | number of stack frames. |
| 2978 | |
| 2979 | ** Backtraces through compiled procedures only show procedures that are |
| 2980 | active in the current continuation |
| 2981 | |
| 2982 | Similarly to the previous issue, backtraces in compiled code may be |
| 2983 | different from backtraces in interpreted code. There are no semantic |
| 2984 | differences, however. Please mail bug-guile@gnu.org if you see any |
| 2985 | deficiencies with Guile's backtraces. |
| 2986 | |
| 2987 | ** `positions' reader option enabled by default |
| 2988 | |
| 2989 | This change allows primitive-load without --auto-compile to also |
| 2990 | propagate source information through the expander, for better errors and |
| 2991 | to let macros know their source locations. The compiler was already |
| 2992 | turning it on anyway. |
| 2993 | |
| 2994 | ** New macro: `current-source-location' |
| 2995 | |
| 2996 | The macro returns the current source location (to be documented). |
| 2997 | |
| 2998 | ** syntax-rules and syntax-case macros now propagate source information |
| 2999 | through to the expanded code |
| 3000 | |
| 3001 | This should result in better backtraces. |
| 3002 | |
| 3003 | ** The currying behavior of `define' has been removed. |
| 3004 | |
| 3005 | Before, `(define ((f a) b) (* a b))' would translate to |
| 3006 | |
| 3007 | (define f (lambda (a) (lambda (b) (* a b)))) |
| 3008 | |
| 3009 | Now a syntax error is signaled, as this syntax is not supported by |
| 3010 | default. Use the `(ice-9 curried-definitions)' module to get back the |
| 3011 | old behavior. |
| 3012 | |
| 3013 | ** New procedure, `define!' |
| 3014 | |
| 3015 | `define!' is a procedure that takes two arguments, a symbol and a value, |
| 3016 | and binds the value to the symbol in the current module. It's useful to |
| 3017 | programmatically make definitions in the current module, and is slightly |
| 3018 | less verbose than `module-define!'. |
| 3019 | |
| 3020 | ** All modules have names now |
| 3021 | |
| 3022 | Before, you could have anonymous modules: modules without names. Now, |
| 3023 | because of hygiene and macros, all modules have names. If a module was |
| 3024 | created without a name, the first time `module-name' is called on it, a |
| 3025 | fresh name will be lazily generated for it. |
| 3026 | |
| 3027 | ** The module namespace is now separate from the value namespace |
| 3028 | |
| 3029 | It was a little-known implementation detail of Guile's module system |
| 3030 | that it was built on a single hierarchical namespace of values -- that |
| 3031 | if there was a module named `(foo bar)', then in the module named |
| 3032 | `(foo)' there was a binding from `bar' to the `(foo bar)' module. |
| 3033 | |
| 3034 | This was a neat trick, but presented a number of problems. One problem |
| 3035 | was that the bindings in a module were not apparent from the module |
| 3036 | itself; perhaps the `(foo)' module had a private binding for `bar', and |
| 3037 | then an external contributor defined `(foo bar)'. In the end there can |
| 3038 | be only one binding, so one of the two will see the wrong thing, and |
| 3039 | produce an obtuse error of unclear provenance. |
| 3040 | |
| 3041 | Also, the public interface of a module was also bound in the value |
| 3042 | namespace, as `%module-public-interface'. This was a hack from the early |
| 3043 | days of Guile's modules. |
| 3044 | |
| 3045 | Both of these warts have been fixed by the addition of fields in the |
| 3046 | `module' data type. Access to modules and their interfaces from the |
| 3047 | value namespace has been deprecated, and all accessors use the new |
| 3048 | record accessors appropriately. |
| 3049 | |
| 3050 | When Guile is built with support for deprecated code, as is the default, |
| 3051 | the value namespace is still searched for modules and public interfaces, |
| 3052 | and a deprecation warning is raised as appropriate. |
| 3053 | |
| 3054 | Finally, to support lazy loading of modules as one used to be able to do |
| 3055 | with module binder procedures, Guile now has submodule binders, called |
| 3056 | if a given submodule is not found. See boot-9.scm for more information. |
| 3057 | |
| 3058 | ** New procedures: module-ref-submodule, module-define-submodule, |
| 3059 | nested-ref-module, nested-define-module!, local-ref-module, |
| 3060 | local-define-module |
| 3061 | |
| 3062 | These new accessors are like their bare variants, but operate on |
| 3063 | namespaces instead of values. |
| 3064 | |
| 3065 | ** The (app modules) module tree is officially deprecated |
| 3066 | |
| 3067 | It used to be that one could access a module named `(foo bar)' via |
| 3068 | `(nested-ref the-root-module '(app modules foo bar))'. The `(app |
| 3069 | modules)' bit was a never-used and never-documented abstraction, and has |
| 3070 | been deprecated. See the following mail for a full discussion: |
| 3071 | |
| 3072 | http://lists.gnu.org/archive/html/guile-devel/2010-04/msg00168.html |
| 3073 | |
| 3074 | The `%app' binding is also deprecated. |
| 3075 | |
| 3076 | ** `module-filename' field and accessor |
| 3077 | |
| 3078 | Modules now record the file in which they are defined. This field may be |
| 3079 | accessed with the new `module-filename' procedure. |
| 3080 | |
| 3081 | ** Modules load within a known environment |
| 3082 | |
| 3083 | It takes a few procedure calls to define a module, and those procedure |
| 3084 | calls need to be in scope. Now we ensure that the current module when |
| 3085 | loading a module is one that has the needed bindings, instead of relying |
| 3086 | on chance. |
| 3087 | |
| 3088 | ** `load' is a macro (!) that resolves paths relative to source file dir |
| 3089 | |
| 3090 | The familiar Schem `load' procedure is now a macro that captures the |
| 3091 | name of the source file being expanded, and dispatches to the new |
| 3092 | `load-in-vicinity'. Referencing `load' by bare name returns a closure |
| 3093 | that embeds the current source file name. |
| 3094 | |
| 3095 | This fix allows `load' of relative paths to be resolved with respect to |
| 3096 | the location of the file that calls `load'. |
| 3097 | |
| 3098 | ** Many syntax errors have different texts now |
| 3099 | |
| 3100 | Syntax errors still throw to the `syntax-error' key, but the arguments |
| 3101 | are often different now. Perhaps in the future, Guile will switch to |
| 3102 | using standard SRFI-35 conditions. |
| 3103 | |
| 3104 | ** Returning multiple values to compiled code will silently truncate the |
| 3105 | values to the expected number |
| 3106 | |
| 3107 | For example, the interpreter would raise an error evaluating the form, |
| 3108 | `(+ (values 1 2) (values 3 4))', because it would see the operands as |
| 3109 | being two compound "values" objects, to which `+' does not apply. |
| 3110 | |
| 3111 | The compiler, on the other hand, receives multiple values on the stack, |
| 3112 | not as a compound object. Given that it must check the number of values |
| 3113 | anyway, if too many values are provided for a continuation, it chooses |
| 3114 | to truncate those values, effectively evaluating `(+ 1 3)' instead. |
| 3115 | |
| 3116 | The idea is that the semantics that the compiler implements is more |
| 3117 | intuitive, and the use of the interpreter will fade out with time. |
| 3118 | This behavior is allowed both by the R5RS and the R6RS. |
| 3119 | |
| 3120 | ** Multiple values in compiled code are not represented by compound |
| 3121 | objects |
| 3122 | |
| 3123 | This change may manifest itself in the following situation: |
| 3124 | |
| 3125 | (let ((val (foo))) (do-something) val) |
| 3126 | |
| 3127 | In the interpreter, if `foo' returns multiple values, multiple values |
| 3128 | are produced from the `let' expression. In the compiler, those values |
| 3129 | are truncated to the first value, and that first value is returned. In |
| 3130 | the compiler, if `foo' returns no values, an error will be raised, while |
| 3131 | the interpreter would proceed. |
| 3132 | |
| 3133 | Both of these behaviors are allowed by R5RS and R6RS. The compiler's |
| 3134 | behavior is more correct, however. If you wish to preserve a potentially |
| 3135 | multiply-valued return, you will need to set up a multiple-value |
| 3136 | continuation, using `call-with-values'. |
| 3137 | |
| 3138 | ** Defmacros are now implemented in terms of syntax-case. |
| 3139 | |
| 3140 | The practical ramification of this is that the `defmacro?' predicate has |
| 3141 | been removed, along with `defmacro-transformer', `macro-table', |
| 3142 | `xformer-table', `assert-defmacro?!', `set-defmacro-transformer!' and |
| 3143 | `defmacro:transformer'. This is because defmacros are simply macros. If |
| 3144 | any of these procedures provided useful facilities to you, we encourage |
| 3145 | you to contact the Guile developers. |
| 3146 | |
| 3147 | ** Hygienic macros documented as the primary syntactic extension mechanism. |
| 3148 | |
| 3149 | The macro documentation was finally fleshed out with some documentation |
| 3150 | on `syntax-rules' and `syntax-case' macros, and other parts of the macro |
| 3151 | expansion process. See "Macros" in the manual, for details. |
| 3152 | |
| 3153 | ** psyntax is now the default expander |
| 3154 | |
| 3155 | Scheme code is now expanded by default by the psyntax hygienic macro |
| 3156 | expander. Expansion is performed completely before compilation or |
| 3157 | interpretation. |
| 3158 | |
| 3159 | Notably, syntax errors will be signalled before interpretation begins. |
| 3160 | In the past, many syntax errors were only detected at runtime if the |
| 3161 | code in question was memoized. |
| 3162 | |
| 3163 | As part of its expansion, psyntax renames all lexically-bound |
| 3164 | identifiers. Original identifier names are preserved and given to the |
| 3165 | compiler, but the interpreter will see the renamed variables, e.g., |
| 3166 | `x432' instead of `x'. |
| 3167 | |
| 3168 | Note that the psyntax that Guile uses is a fork, as Guile already had |
| 3169 | modules before incompatible modules were added to psyntax -- about 10 |
| 3170 | years ago! Thus there are surely a number of bugs that have been fixed |
| 3171 | in psyntax since then. If you find one, please notify bug-guile@gnu.org. |
| 3172 | |
| 3173 | ** syntax-rules and syntax-case are available by default. |
| 3174 | |
| 3175 | There is no longer any need to import the `(ice-9 syncase)' module |
| 3176 | (which is now deprecated). The expander may be invoked directly via |
| 3177 | `macroexpand', though it is normally searched for via the current module |
| 3178 | transformer. |
| 3179 | |
| 3180 | Also, the helper routines for syntax-case are available in the default |
| 3181 | environment as well: `syntax->datum', `datum->syntax', |
| 3182 | `bound-identifier=?', `free-identifier=?', `generate-temporaries', |
| 3183 | `identifier?', and `syntax-violation'. See the R6RS for documentation. |
| 3184 | |
| 3185 | ** Tail patterns in syntax-case |
| 3186 | |
| 3187 | Guile has pulled in some more recent changes from the psyntax portable |
| 3188 | syntax expander, to implement support for "tail patterns". Such patterns |
| 3189 | are supported by syntax-rules and syntax-case. This allows a syntax-case |
| 3190 | match clause to have ellipses, then a pattern at the end. For example: |
| 3191 | |
| 3192 | (define-syntax case |
| 3193 | (syntax-rules (else) |
| 3194 | ((_ val match-clause ... (else e e* ...)) |
| 3195 | [...]))) |
| 3196 | |
| 3197 | Note how there is MATCH-CLAUSE, which is ellipsized, then there is a |
| 3198 | tail pattern for the else clause. Thanks to Andreas Rottmann for the |
| 3199 | patch, and Kent Dybvig for the code. |
| 3200 | |
| 3201 | ** Lexical bindings introduced by hygienic macros may not be referenced |
| 3202 | by nonhygienic macros. |
| 3203 | |
| 3204 | If a lexical binding is introduced by a hygienic macro, it may not be |
| 3205 | referenced by a nonhygienic macro. For example, this works: |
| 3206 | |
| 3207 | (let () |
| 3208 | (define-macro (bind-x val body) |
| 3209 | `(let ((x ,val)) ,body)) |
| 3210 | (define-macro (ref x) |
| 3211 | x) |
| 3212 | (bind-x 10 (ref x))) |
| 3213 | |
| 3214 | But this does not: |
| 3215 | |
| 3216 | (let () |
| 3217 | (define-syntax bind-x |
| 3218 | (syntax-rules () |
| 3219 | ((_ val body) (let ((x val)) body)))) |
| 3220 | (define-macro (ref x) |
| 3221 | x) |
| 3222 | (bind-x 10 (ref x))) |
| 3223 | |
| 3224 | It is not normal to run into this situation with existing code. However, |
| 3225 | if you have defmacros that expand to hygienic macros, it is possible to |
| 3226 | run into situations like this. For example, if you have a defmacro that |
| 3227 | generates a `while' expression, the `break' bound by the `while' may not |
| 3228 | be visible within other parts of your defmacro. The solution is to port |
| 3229 | from defmacros to syntax-rules or syntax-case. |
| 3230 | |
| 3231 | ** Macros may no longer be referenced as first-class values. |
| 3232 | |
| 3233 | In the past, you could evaluate e.g. `if', and get its macro value. Now, |
| 3234 | expanding this form raises a syntax error. |
| 3235 | |
| 3236 | Macros still /exist/ as first-class values, but they must be |
| 3237 | /referenced/ via the module system, e.g. `(module-ref (current-module) |
| 3238 | 'if)'. |
| 3239 | |
| 3240 | ** Macros may now have docstrings. |
| 3241 | |
| 3242 | `object-documentation' from `(ice-9 documentation)' may be used to |
| 3243 | retrieve the docstring, once you have a macro value -- but see the above |
| 3244 | note about first-class macros. Docstrings are associated with the syntax |
| 3245 | transformer procedures. |
| 3246 | |
| 3247 | ** `case-lambda' is now available in the default environment. |
| 3248 | |
| 3249 | The binding in the default environment is equivalent to the one from the |
| 3250 | `(srfi srfi-16)' module. Use the srfi-16 module explicitly if you wish |
| 3251 | to maintain compatibility with Guile 1.8 and earlier. |
| 3252 | |
| 3253 | ** Procedures may now have more than one arity. |
| 3254 | |
| 3255 | This can be the case, for example, in case-lambda procedures. The |
| 3256 | arities of compiled procedures may be accessed via procedures from the |
| 3257 | `(system vm program)' module; see "Compiled Procedures", "Optional |
| 3258 | Arguments", and "Case-lambda" in the manual. |
| 3259 | |
| 3260 | ** Deprecate arity access via (procedure-properties proc 'arity) |
| 3261 | |
| 3262 | Instead of accessing a procedure's arity as a property, use the new |
| 3263 | `procedure-minimum-arity' function, which gives the most permissive |
| 3264 | arity that the function has, in the same format as the old arity |
| 3265 | accessor. |
| 3266 | |
| 3267 | ** `lambda*' and `define*' are now available in the default environment |
| 3268 | |
| 3269 | As with `case-lambda', `(ice-9 optargs)' continues to be supported, for |
| 3270 | compatibility purposes. No semantic change has been made (we hope). |
| 3271 | Optional and keyword arguments now dispatch via special VM operations, |
| 3272 | without the need to cons rest arguments, making them very fast. |
| 3273 | |
| 3274 | ** New syntax: define-once |
| 3275 | |
| 3276 | `define-once' is like Lisp's `defvar': it creates a toplevel binding, |
| 3277 | but only if one does not exist already. |
| 3278 | |
| 3279 | ** New function, `truncated-print', with `format' support |
| 3280 | |
| 3281 | `(ice-9 pretty-print)' now exports `truncated-print', a printer that |
| 3282 | will ensure that the output stays within a certain width, truncating the |
| 3283 | output in what is hopefully an intelligent manner. See the manual for |
| 3284 | more details. |
| 3285 | |
| 3286 | There is a new `format' specifier, `~@y', for doing a truncated |
| 3287 | print (as opposed to `~y', which does a pretty-print). See the `format' |
| 3288 | documentation for more details. |
| 3289 | |
| 3290 | ** Better pretty-printing |
| 3291 | |
| 3292 | Indentation recognizes more special forms, like `syntax-case', and read |
| 3293 | macros like `quote' are printed better. |
| 3294 | |
| 3295 | ** Passing a number as the destination of `format' is deprecated |
| 3296 | |
| 3297 | The `format' procedure in `(ice-9 format)' now emits a deprecation |
| 3298 | warning if a number is passed as its first argument. |
| 3299 | |
| 3300 | Also, it used to be that you could omit passing a port to `format', in |
| 3301 | some cases. This still works, but has been formally deprecated. |
| 3302 | |
| 3303 | ** SRFI-4 vectors reimplemented in terms of R6RS bytevectors |
| 3304 | |
| 3305 | Guile now implements SRFI-4 vectors using bytevectors. Often when you |
| 3306 | have a numeric vector, you end up wanting to write its bytes somewhere, |
| 3307 | or have access to the underlying bytes, or read in bytes from somewhere |
| 3308 | else. Bytevectors are very good at this sort of thing. But the SRFI-4 |
| 3309 | APIs are nicer to use when doing number-crunching, because they are |
| 3310 | addressed by element and not by byte. |
| 3311 | |
| 3312 | So as a compromise, Guile allows all bytevector functions to operate on |
| 3313 | numeric vectors. They address the underlying bytes in the native |
| 3314 | endianness, as one would expect. |
| 3315 | |
| 3316 | Following the same reasoning, that it's just bytes underneath, Guile |
| 3317 | also allows uniform vectors of a given type to be accessed as if they |
| 3318 | were of any type. One can fill a u32vector, and access its elements with |
| 3319 | u8vector-ref. One can use f64vector-ref on bytevectors. It's all the |
| 3320 | same to Guile. |
| 3321 | |
| 3322 | In this way, uniform numeric vectors may be written to and read from |
| 3323 | input/output ports using the procedures that operate on bytevectors. |
| 3324 | |
| 3325 | Calls to SRFI-4 accessors (ref and set functions) from Scheme are now |
| 3326 | inlined to the VM instructions for bytevector access. |
| 3327 | |
| 3328 | See "SRFI-4" in the manual, for more information. |
| 3329 | |
| 3330 | ** Nonstandard SRFI-4 procedures now available from `(srfi srfi-4 gnu)' |
| 3331 | |
| 3332 | Guile's `(srfi srfi-4)' now only exports those srfi-4 procedures that |
| 3333 | are part of the standard. Complex uniform vectors and the |
| 3334 | `any->FOOvector' family are now available only from `(srfi srfi-4 gnu)'. |
| 3335 | |
| 3336 | Guile's default environment imports `(srfi srfi-4)', and probably should |
| 3337 | import `(srfi srfi-4 gnu)' as well. |
| 3338 | |
| 3339 | See "SRFI-4 Extensions" in the manual, for more information. |
| 3340 | |
| 3341 | ** New syntax: include-from-path. |
| 3342 | |
| 3343 | `include-from-path' is like `include', except it looks for its file in |
| 3344 | the load path. It can be used to compile other files into a file. |
| 3345 | |
| 3346 | ** New syntax: quasisyntax. |
| 3347 | |
| 3348 | `quasisyntax' is to `syntax' as `quasiquote' is to `quote'. See the R6RS |
| 3349 | documentation for more information. Thanks to Andre van Tonder for the |
| 3350 | implementation. |
| 3351 | |
| 3352 | ** `*unspecified*' is identifier syntax |
| 3353 | |
| 3354 | `*unspecified*' is no longer a variable, so it is optimized properly by |
| 3355 | the compiler, and is not `set!'-able. |
| 3356 | |
| 3357 | ** Changes and bugfixes in numerics code |
| 3358 | |
| 3359 | *** Added six new sets of fast quotient and remainder operators |
| 3360 | |
| 3361 | Added six new sets of fast quotient and remainder operator pairs with |
| 3362 | different semantics than the R5RS operators. They support not only |
| 3363 | integers, but all reals, including exact rationals and inexact |
| 3364 | floating point numbers. |
| 3365 | |
| 3366 | These procedures accept two real numbers N and D, where the divisor D |
| 3367 | must be non-zero. Each set of operators computes an integer quotient |
| 3368 | Q and a real remainder R such that N = Q*D + R and |R| < |D|. They |
| 3369 | differ only in how N/D is rounded to produce Q. |
| 3370 | |
| 3371 | `euclidean-quotient' returns the integer Q and `euclidean-remainder' |
| 3372 | returns the real R such that N = Q*D + R and 0 <= R < |D|. `euclidean/' |
| 3373 | returns both Q and R, and is more efficient than computing each |
| 3374 | separately. Note that when D > 0, `euclidean-quotient' returns |
| 3375 | floor(N/D), and when D < 0 it returns ceiling(N/D). |
| 3376 | |
| 3377 | `centered-quotient', `centered-remainder', and `centered/' are similar |
| 3378 | except that the range of remainders is -abs(D/2) <= R < abs(D/2), and |
| 3379 | `centered-quotient' rounds N/D to the nearest integer. Note that these |
| 3380 | operators are equivalent to the R6RS integer division operators `div', |
| 3381 | `mod', `div-and-mod', `div0', `mod0', and `div0-and-mod0'. |
| 3382 | |
| 3383 | `floor-quotient' and `floor-remainder' compute Q and R, respectively, |
| 3384 | where Q has been rounded toward negative infinity. `floor/' returns |
| 3385 | both Q and R, and is more efficient than computing each separately. |
| 3386 | Note that when applied to integers, `floor-remainder' is equivalent to |
| 3387 | the R5RS integer-only `modulo' operator. `ceiling-quotient', |
| 3388 | `ceiling-remainder', and `ceiling/' are similar except that Q is |
| 3389 | rounded toward positive infinity. |
| 3390 | |
| 3391 | For `truncate-quotient', `truncate-remainder', and `truncate/', Q is |
| 3392 | rounded toward zero. Note that when applied to integers, |
| 3393 | `truncate-quotient' and `truncate-remainder' are equivalent to the |
| 3394 | R5RS integer-only operators `quotient' and `remainder'. |
| 3395 | |
| 3396 | For `round-quotient', `round-remainder', and `round/', Q is rounded to |
| 3397 | the nearest integer, with ties going to the nearest even integer. |
| 3398 | |
| 3399 | *** Complex number changes |
| 3400 | |
| 3401 | Guile is now able to represent non-real complex numbers whose |
| 3402 | imaginary part is an _inexact_ zero (0.0 or -0.0), per R6RS. |
| 3403 | Previously, such numbers were immediately changed into inexact reals. |
| 3404 | |
| 3405 | (real? 0.0+0.0i) now returns #f, per R6RS, although (zero? 0.0+0.0i) |
| 3406 | still returns #t, per R6RS. (= 0 0.0+0.0i) and (= 0.0 0.0+0.0i) are |
| 3407 | #t, but the same comparisons using `eqv?' or `equal?' are #f. |
| 3408 | |
| 3409 | Like other non-real numbers, these complex numbers with inexact zero |
| 3410 | imaginary part will raise exceptions is passed to procedures requiring |
| 3411 | reals, such as `<', `>', `<=', `>=', `min', `max', `positive?', |
| 3412 | `negative?', `inf?', `nan?', `finite?', etc. |
| 3413 | |
| 3414 | **** `make-rectangular' changes |
| 3415 | |
| 3416 | scm_make_rectangular `make-rectangular' now returns a real number only |
| 3417 | if the imaginary part is an _exact_ 0. Previously, it would return a |
| 3418 | real number if the imaginary part was an inexact zero. |
| 3419 | |
| 3420 | scm_c_make_rectangular now always returns a non-real complex number, |
| 3421 | even if the imaginary part is zero. Previously, it would return a |
| 3422 | real number if the imaginary part was zero. |
| 3423 | |
| 3424 | **** `make-polar' changes |
| 3425 | |
| 3426 | scm_make_polar `make-polar' now returns a real number only if the |
| 3427 | angle or magnitude is an _exact_ 0. If the magnitude is an exact 0, |
| 3428 | it now returns an exact 0. Previously, it would return a real |
| 3429 | number if the imaginary part was an inexact zero. |
| 3430 | |
| 3431 | scm_c_make_polar now always returns a non-real complex number, even if |
| 3432 | the imaginary part is 0.0. Previously, it would return a real number |
| 3433 | if the imaginary part was 0.0. |
| 3434 | |
| 3435 | **** `imag-part' changes |
| 3436 | |
| 3437 | scm_imag_part `imag-part' now returns an exact 0 if applied to an |
| 3438 | inexact real number. Previously it returned an inexact zero in this |
| 3439 | case. |
| 3440 | |
| 3441 | *** `eqv?' and `equal?' now compare numbers equivalently |
| 3442 | |
| 3443 | scm_equal_p `equal?' now behaves equivalently to scm_eqv_p `eqv?' for |
| 3444 | numeric values, per R5RS. Previously, equal? worked differently, |
| 3445 | e.g. `(equal? 0.0 -0.0)' returned #t but `(eqv? 0.0 -0.0)' returned #f, |
| 3446 | and `(equal? +nan.0 +nan.0)' returned #f but `(eqv? +nan.0 +nan.0)' |
| 3447 | returned #t. |
| 3448 | |
| 3449 | *** `(equal? +nan.0 +nan.0)' now returns #t |
| 3450 | |
| 3451 | Previously, `(equal? +nan.0 +nan.0)' returned #f, although |
| 3452 | `(let ((x +nan.0)) (equal? x x))' and `(eqv? +nan.0 +nan.0)' |
| 3453 | both returned #t. R5RS requires that `equal?' behave like |
| 3454 | `eqv?' when comparing numbers. |
| 3455 | |
| 3456 | *** Change in handling products `*' involving exact 0 |
| 3457 | |
| 3458 | scm_product `*' now handles exact 0 differently. A product containing |
| 3459 | an exact 0 now returns an exact 0 if and only if the other arguments |
| 3460 | are all exact. An inexact zero is returned if and only if the other |
| 3461 | arguments are all finite but not all exact. If an infinite or NaN |
| 3462 | value is present, a NaN value is returned. Previously, any product |
| 3463 | containing an exact 0 yielded an exact 0, regardless of the other |
| 3464 | arguments. |
| 3465 | |
| 3466 | *** `expt' and `integer-expt' changes when the base is 0 |
| 3467 | |
| 3468 | While `(expt 0 0)' is still 1, and `(expt 0 N)' for N > 0 is still |
| 3469 | zero, `(expt 0 N)' for N < 0 is now a NaN value, and likewise for |
| 3470 | integer-expt. This is more correct, and conforming to R6RS, but seems |
| 3471 | to be incompatible with R5RS, which would return 0 for all non-zero |
| 3472 | values of N. |
| 3473 | |
| 3474 | *** `expt' and `integer-expt' are more generic, less strict |
| 3475 | |
| 3476 | When raising to an exact non-negative integer exponent, `expt' and |
| 3477 | `integer-expt' are now able to exponentiate any object that can be |
| 3478 | multiplied using `*'. They can also raise an object to an exact |
| 3479 | negative integer power if its reciprocal can be taken using `/'. |
| 3480 | In order to allow this, the type of the first argument is no longer |
| 3481 | checked when raising to an exact integer power. If the exponent is 0 |
| 3482 | or 1, the first parameter is not manipulated at all, and need not |
| 3483 | even support multiplication. |
| 3484 | |
| 3485 | *** Infinities are no longer integers, nor rationals |
| 3486 | |
| 3487 | scm_integer_p `integer?' and scm_rational_p `rational?' now return #f |
| 3488 | for infinities, per R6RS. Previously they returned #t for real |
| 3489 | infinities. The real infinities and NaNs are still considered real by |
| 3490 | scm_real `real?' however, per R6RS. |
| 3491 | |
| 3492 | *** NaNs are no longer rationals |
| 3493 | |
| 3494 | scm_rational_p `rational?' now returns #f for NaN values, per R6RS. |
| 3495 | Previously it returned #t for real NaN values. They are still |
| 3496 | considered real by scm_real `real?' however, per R6RS. |
| 3497 | |
| 3498 | *** `inf?' and `nan?' now throw exceptions for non-reals |
| 3499 | |
| 3500 | The domain of `inf?' and `nan?' is the real numbers. Guile now signals |
| 3501 | an error when a non-real number or non-number is passed to these |
| 3502 | procedures. (Note that NaNs _are_ considered numbers by scheme, despite |
| 3503 | their name). |
| 3504 | |
| 3505 | *** `rationalize' bugfixes and changes |
| 3506 | |
| 3507 | Fixed bugs in scm_rationalize `rationalize'. Previously, it returned |
| 3508 | exact integers unmodified, although that was incorrect if the epsilon |
| 3509 | was at least 1 or inexact, e.g. (rationalize 4 1) should return 3 per |
| 3510 | R5RS and R6RS, but previously it returned 4. It also now handles |
| 3511 | cases involving infinities and NaNs properly, per R6RS. |
| 3512 | |
| 3513 | *** Trigonometric functions now return exact numbers in some cases |
| 3514 | |
| 3515 | scm_sin `sin', scm_cos `cos', scm_tan `tan', scm_asin `asin', scm_acos |
| 3516 | `acos', scm_atan `atan', scm_sinh `sinh', scm_cosh `cosh', scm_tanh |
| 3517 | `tanh', scm_sys_asinh `asinh', scm_sys_acosh `acosh', and |
| 3518 | scm_sys_atanh `atanh' now return exact results in some cases. |
| 3519 | |
| 3520 | *** New procedure: `finite?' |
| 3521 | |
| 3522 | Add scm_finite_p `finite?' from R6RS to guile core, which returns #t |
| 3523 | if and only if its argument is neither infinite nor a NaN. Note that |
| 3524 | this is not the same as (not (inf? x)) or (not (infinite? x)), since |
| 3525 | NaNs are neither finite nor infinite. |
| 3526 | |
| 3527 | *** Improved exactness handling for complex number parsing |
| 3528 | |
| 3529 | When parsing non-real complex numbers, exactness specifiers are now |
| 3530 | applied to each component, as is done in PLT Scheme. For complex |
| 3531 | numbers written in rectangular form, exactness specifiers are applied |
| 3532 | to the real and imaginary parts before calling scm_make_rectangular. |
| 3533 | For complex numbers written in polar form, exactness specifiers are |
| 3534 | applied to the magnitude and angle before calling scm_make_polar. |
| 3535 | |
| 3536 | Previously, exactness specifiers were applied to the number as a whole |
| 3537 | _after_ calling scm_make_rectangular or scm_make_polar. |
| 3538 | |
| 3539 | For example, (string->number "#i5.0+0i") now does the equivalent of: |
| 3540 | |
| 3541 | (make-rectangular (exact->inexact 5.0) (exact->inexact 0)) |
| 3542 | |
| 3543 | which yields 5.0+0.0i. Previously it did the equivalent of: |
| 3544 | |
| 3545 | (exact->inexact (make-rectangular 5.0 0)) |
| 3546 | |
| 3547 | which yielded 5.0. |
| 3548 | |
| 3549 | ** Unicode characters |
| 3550 | |
| 3551 | Unicode characters may be entered in octal format via e.g. `#\454', or |
| 3552 | created via (integer->char 300). A hex external representation will |
| 3553 | probably be introduced at some point. |
| 3554 | |
| 3555 | ** Unicode strings |
| 3556 | |
| 3557 | Internally, strings are now represented either in the `latin-1' |
| 3558 | encoding, one byte per character, or in UTF-32, with four bytes per |
| 3559 | character. Strings manage their own allocation, switching if needed. |
| 3560 | |
| 3561 | Extended characters may be written in a literal string using the |
| 3562 | hexadecimal escapes `\xXX', `\uXXXX', or `\UXXXXXX', for 8-bit, 16-bit, |
| 3563 | or 24-bit codepoints, respectively, or entered directly in the native |
| 3564 | encoding of the port on which the string is read. |
| 3565 | |
| 3566 | ** Unicode symbols |
| 3567 | |
| 3568 | One may now use U+03BB (GREEK SMALL LETTER LAMBDA) as an identifier. |
| 3569 | |
| 3570 | ** Support for non-ASCII source code files |
| 3571 | |
| 3572 | The default reader now handles source code files for some of the |
| 3573 | non-ASCII character encodings, such as UTF-8. A non-ASCII source file |
| 3574 | should have an encoding declaration near the top of the file. Also, |
| 3575 | there is a new function, `file-encoding', that scans a port for a coding |
| 3576 | declaration. See the section of the manual entitled, "Character Encoding |
| 3577 | of Source Files". |
| 3578 | |
| 3579 | The pre-1.9.3 reader handled 8-bit clean but otherwise unspecified source |
| 3580 | code. This use is now discouraged. Binary input and output is |
| 3581 | currently supported by opening ports in the ISO-8859-1 locale. |
| 3582 | |
| 3583 | ** Source files default to UTF-8. |
| 3584 | |
| 3585 | If source files do not specify their encoding via a `coding:' block, |
| 3586 | the default encoding is UTF-8, instead of being taken from the current |
| 3587 | locale. |
| 3588 | |
| 3589 | ** Interactive Guile installs the current locale. |
| 3590 | |
| 3591 | Instead of leaving the user in the "C" locale, running the Guile REPL |
| 3592 | installs the current locale. [FIXME xref?] |
| 3593 | |
| 3594 | ** Support for locale transcoding when reading from and writing to ports |
| 3595 | |
| 3596 | Ports now have an associated character encoding, and port read and write |
| 3597 | operations do conversion to and from locales automatically. Ports also |
| 3598 | have an associated strategy for how to deal with locale conversion |
| 3599 | failures. |
| 3600 | |
| 3601 | See the documentation in the manual for the four new support functions, |
| 3602 | `set-port-encoding!', `port-encoding', `set-port-conversion-strategy!', |
| 3603 | and `port-conversion-strategy'. |
| 3604 | |
| 3605 | ** String and SRFI-13 functions can operate on Unicode strings |
| 3606 | |
| 3607 | ** Unicode support for SRFI-14 character sets |
| 3608 | |
| 3609 | The default character sets are no longer locale dependent and contain |
| 3610 | characters from the whole Unicode range. There is a new predefined |
| 3611 | character set, `char-set:designated', which contains all assigned |
| 3612 | Unicode characters. There is a new debugging function, `%char-set-dump'. |
| 3613 | |
| 3614 | ** Character functions operate on Unicode characters |
| 3615 | |
| 3616 | `char-upcase' and `char-downcase' use default Unicode casing rules. |
| 3617 | Character comparisons such as `char<?' and `char-ci<?' now sort based on |
| 3618 | Unicode code points. |
| 3619 | |
| 3620 | ** Global variables `scm_charnames' and `scm_charnums' are removed |
| 3621 | |
| 3622 | These variables contained the names of control characters and were |
| 3623 | used when writing characters. While these were global, they were |
| 3624 | never intended to be public API. They have been replaced with private |
| 3625 | functions. |
| 3626 | |
| 3627 | ** EBCDIC support is removed |
| 3628 | |
| 3629 | There was an EBCDIC compile flag that altered some of the character |
| 3630 | processing. It appeared that full EBCDIC support was never completed |
| 3631 | and was unmaintained. |
| 3632 | |
| 3633 | ** Compile-time warnings |
| 3634 | |
| 3635 | Guile can warn about potentially unbound free variables. Pass the |
| 3636 | -Wunbound-variable on the `guile-tools compile' command line, or add |
| 3637 | `#:warnings '(unbound-variable)' to your `compile' or `compile-file' |
| 3638 | invocation. Warnings are also enabled by default for expressions entered |
| 3639 | at the REPL. |
| 3640 | |
| 3641 | Guile can also warn when you pass the wrong number of arguments to a |
| 3642 | procedure, with -Warity-mismatch, or `arity-mismatch' in the |
| 3643 | `#:warnings' as above. |
| 3644 | |
| 3645 | Other warnings include `-Wunused-variable' and `-Wunused-toplevel', to |
| 3646 | warn about unused local or global (top-level) variables, and `-Wformat', |
| 3647 | to check for various errors related to the `format' procedure. |
| 3648 | |
| 3649 | ** A new `memoize-symbol' evaluator trap has been added. |
| 3650 | |
| 3651 | This trap can be used for efficiently implementing a Scheme code |
| 3652 | coverage. |
| 3653 | |
| 3654 | ** Duplicate bindings among used modules are resolved lazily. |
| 3655 | |
| 3656 | This slightly improves program startup times. |
| 3657 | |
| 3658 | ** New thread cancellation and thread cleanup API |
| 3659 | |
| 3660 | See `cancel-thread', `set-thread-cleanup!', and `thread-cleanup'. |
| 3661 | |
| 3662 | ** New threads are in `(guile-user)' by default, not `(guile)' |
| 3663 | |
| 3664 | It used to be that a new thread entering Guile would do so in the |
| 3665 | `(guile)' module, unless this was the first time Guile was initialized, |
| 3666 | in which case it was `(guile-user)'. This has been fixed to have all |
| 3667 | new threads unknown to Guile default to `(guile-user)'. |
| 3668 | |
| 3669 | ** New helpers: `print-exception', `set-exception-printer!' |
| 3670 | |
| 3671 | These functions implement an extensible exception printer. Guile |
| 3672 | registers printers for all of the exceptions it throws. Users may add |
| 3673 | their own printers. There is also `scm_print_exception', for use by C |
| 3674 | programs. Pleasantly, this allows SRFI-35 and R6RS exceptions to be |
| 3675 | printed appropriately. |
| 3676 | |
| 3677 | ** GOOPS dispatch in scheme |
| 3678 | |
| 3679 | As an implementation detail, GOOPS dispatch is no longer implemented by |
| 3680 | special evaluator bytecodes, but rather directly via a Scheme function |
| 3681 | associated with an applicable struct. There is some VM support for the |
| 3682 | underlying primitives, like `class-of'. |
| 3683 | |
| 3684 | This change will in the future allow users to customize generic function |
| 3685 | dispatch without incurring a performance penalty, and allow us to |
| 3686 | implement method combinations. |
| 3687 | |
| 3688 | ** Applicable struct support |
| 3689 | |
| 3690 | One may now make structs from Scheme that may be applied as procedures. |
| 3691 | To do so, make a struct whose vtable is `<applicable-struct-vtable>'. |
| 3692 | That struct will be the vtable of your applicable structs; instances of |
| 3693 | that new struct are assumed to have the procedure in their first slot. |
| 3694 | `<applicable-struct-vtable>' is like Common Lisp's |
| 3695 | `funcallable-standard-class'. Likewise there is |
| 3696 | `<applicable-struct-with-setter-vtable>', which looks for the setter in |
| 3697 | the second slot. This needs to be better documented. |
| 3698 | |
| 3699 | ** GOOPS cleanups. |
| 3700 | |
| 3701 | GOOPS had a number of concepts that were relevant to the days of Tcl, |
| 3702 | but not any more: operators and entities, mainly. These objects were |
| 3703 | never documented, and it is unlikely that they were ever used. Operators |
| 3704 | were a kind of generic specific to the Tcl support. Entities were |
| 3705 | replaced by applicable structs, mentioned above. |
| 3706 | |
| 3707 | ** New struct slot allocation: "hidden" |
| 3708 | |
| 3709 | A hidden slot is readable and writable, but will not be initialized by a |
| 3710 | call to make-struct. For example in your layout you would say "ph" |
| 3711 | instead of "pw". Hidden slots are useful for adding new slots to a |
| 3712 | vtable without breaking existing invocations to make-struct. |
| 3713 | |
| 3714 | ** eqv? not a generic |
| 3715 | |
| 3716 | One used to be able to extend `eqv?' as a primitive-generic, but no |
| 3717 | more. Because `eqv?' is in the expansion of `case' (via `memv'), which |
| 3718 | should be able to compile to static dispatch tables, it doesn't make |
| 3719 | sense to allow extensions that would subvert this optimization. |
| 3720 | |
| 3721 | ** `inet-ntop' and `inet-pton' are always available. |
| 3722 | |
| 3723 | Guile now use a portable implementation of `inet_pton'/`inet_ntop', so |
| 3724 | there is no more need to use `inet-aton'/`inet-ntoa'. The latter |
| 3725 | functions are deprecated. |
| 3726 | |
| 3727 | ** `getopt-long' parsing errors throw to `quit', not `misc-error' |
| 3728 | |
| 3729 | This change should inhibit backtraces on argument parsing errors. |
| 3730 | `getopt-long' has been modified to print out the error that it throws |
| 3731 | itself. |
| 3732 | |
| 3733 | ** New primitive: `tmpfile'. |
| 3734 | |
| 3735 | See "File System" in the manual. |
| 3736 | |
| 3737 | ** Random generator state may be serialized to a datum |
| 3738 | |
| 3739 | `random-state->datum' will serialize a random state to a datum, which |
| 3740 | may be written out, read back in later, and revivified using |
| 3741 | `datum->random-state'. See "Random" in the manual, for more details. |
| 3742 | |
| 3743 | ** Fix random number generator on 64-bit platforms |
| 3744 | |
| 3745 | There was a nasty bug on 64-bit platforms in which asking for a random |
| 3746 | integer with a range between 2**32 and 2**64 caused a segfault. After |
| 3747 | many embarrassing iterations, this was fixed. |
| 3748 | |
| 3749 | ** Fast bit operations. |
| 3750 | |
| 3751 | The bit-twiddling operations `ash', `logand', `logior', and `logxor' now |
| 3752 | have dedicated bytecodes. Guile is not just for symbolic computation, |
| 3753 | it's for number crunching too. |
| 3754 | |
| 3755 | ** Faster SRFI-9 record access |
| 3756 | |
| 3757 | SRFI-9 records are now implemented directly on top of Guile's structs, |
| 3758 | and their accessors are defined in such a way that normal call-sites |
| 3759 | inline to special VM opcodes, while still allowing for the general case |
| 3760 | (e.g. passing a record accessor to `apply'). |
| 3761 | |
| 3762 | ** R6RS block comment support |
| 3763 | |
| 3764 | Guile now supports R6RS nested block comments. The start of a comment is |
| 3765 | marked with `#|', and the end with `|#'. |
| 3766 | |
| 3767 | ** `guile-2' cond-expand feature |
| 3768 | |
| 3769 | To test if your code is running under Guile 2.0 (or its alpha releases), |
| 3770 | test for the `guile-2' cond-expand feature. Like this: |
| 3771 | |
| 3772 | (cond-expand (guile-2 (eval-when (compile) |
| 3773 | ;; This must be evaluated at compile time. |
| 3774 | (fluid-set! current-reader my-reader))) |
| 3775 | (guile |
| 3776 | ;; Earlier versions of Guile do not have a |
| 3777 | ;; separate compilation phase. |
| 3778 | (fluid-set! current-reader my-reader))) |
| 3779 | |
| 3780 | ** New global variables: %load-compiled-path, %load-compiled-extensions |
| 3781 | |
| 3782 | These are analogous to %load-path and %load-extensions. |
| 3783 | |
| 3784 | ** New fluid: `%file-port-name-canonicalization' |
| 3785 | |
| 3786 | This fluid parameterizes the file names that are associated with file |
| 3787 | ports. If %file-port-name-canonicalization is 'absolute, then file names |
| 3788 | are canonicalized to be absolute paths. If it is 'relative, then the |
| 3789 | name is canonicalized, but any prefix corresponding to a member of |
| 3790 | `%load-path' is stripped off. Otherwise the names are passed through |
| 3791 | unchanged. |
| 3792 | |
| 3793 | In addition, the `compile-file' and `compile-and-load' procedures bind |
| 3794 | %file-port-name-canonicalization to their `#:canonicalization' keyword |
| 3795 | argument, which defaults to 'relative. In this way, one might compile |
| 3796 | "../module/ice-9/boot-9.scm", but the path that gets residualized into |
| 3797 | the .go is "ice-9/boot-9.scm". |
| 3798 | |
| 3799 | ** New procedure, `make-promise' |
| 3800 | |
| 3801 | `(make-promise (lambda () foo))' is equivalent to `(delay foo)'. |
| 3802 | |
| 3803 | ** `defined?' may accept a module as its second argument |
| 3804 | |
| 3805 | Previously it only accepted internal structures from the evaluator. |
| 3806 | |
| 3807 | ** New entry into %guile-build-info: `ccachedir' |
| 3808 | |
| 3809 | ** Fix bug in `module-bound?'. |
| 3810 | |
| 3811 | `module-bound?' was returning true if a module did have a local |
| 3812 | variable, but one that was unbound, but another imported module bound |
| 3813 | the variable. This was an error, and was fixed. |
| 3814 | |
| 3815 | ** `(ice-9 syncase)' has been deprecated. |
| 3816 | |
| 3817 | As syntax-case is available by default, importing `(ice-9 syncase)' has |
| 3818 | no effect, and will trigger a deprecation warning. |
| 3819 | |
| 3820 | ** New readline history functions |
| 3821 | |
| 3822 | The (ice-9 readline) module now provides add-history, read-history, |
| 3823 | write-history and clear-history, which wrap the corresponding GNU |
| 3824 | History library functions. |
| 3825 | |
| 3826 | ** Removed deprecated uniform array procedures: |
| 3827 | dimensions->uniform-array, list->uniform-array, array-prototype |
| 3828 | |
| 3829 | Instead, use make-typed-array, list->typed-array, or array-type, |
| 3830 | respectively. |
| 3831 | |
| 3832 | ** Deprecate the old `scm-style-repl' |
| 3833 | |
| 3834 | The following bindings from boot-9 are now found in `(ice-9 |
| 3835 | scm-style-repl)': `scm-style-repl', `error-catching-loop', |
| 3836 | `error-catching-repl', `bad-throw', `scm-repl-silent' |
| 3837 | `assert-repl-silence', `repl-print-unspecified', |
| 3838 | `assert-repl-print-unspecified', `scm-repl-verbose', |
| 3839 | `assert-repl-verbosity', `scm-repl-prompt', `set-repl-prompt!', `repl', |
| 3840 | `default-pre-unwind-handler', `handle-system-error', |
| 3841 | |
| 3842 | The following bindings have been deprecated, with no replacement: |
| 3843 | `pre-unwind-handler-dispatch'. |
| 3844 | |
| 3845 | The following bindings have been totally removed: |
| 3846 | `before-signal-stack'. |
| 3847 | |
| 3848 | Deprecated forwarding shims have been installed so that users that |
| 3849 | expect these bindings in the main namespace will still work, but receive |
| 3850 | a deprecation warning. |
| 3851 | |
| 3852 | ** `set-batch-mode?!' replaced by `ensure-batch-mode!' |
| 3853 | |
| 3854 | "Batch mode" is a flag used to tell a program that it is not running |
| 3855 | interactively. One usually turns it on after a fork. It may not be |
| 3856 | turned off. `ensure-batch-mode!' deprecates the old `set-batch-mode?!', |
| 3857 | because it is a better interface, as it can only turn on batch mode, not |
| 3858 | turn it off. |
| 3859 | |
| 3860 | ** Deprecate `save-stack', `the-last-stack' |
| 3861 | |
| 3862 | It used to be that the way to debug programs in Guile was to capture the |
| 3863 | stack at the time of error, drop back to the REPL, then debug that |
| 3864 | stack. But this approach didn't compose, was tricky to get right in the |
| 3865 | presence of threads, and was not very powerful. |
| 3866 | |
| 3867 | So `save-stack', `stack-saved?', and `the-last-stack' have been moved to |
| 3868 | `(ice-9 save-stack)', with deprecated bindings left in the root module. |
| 3869 | |
| 3870 | ** `top-repl' has its own module |
| 3871 | |
| 3872 | The `top-repl' binding, called with Guile is run interactively, is now |
| 3873 | is its own module, `(ice-9 top-repl)'. A deprecated forwarding shim was |
| 3874 | left in the default environment. |
| 3875 | |
| 3876 | ** `display-error' takes a frame |
| 3877 | |
| 3878 | The `display-error' / `scm_display_error' helper now takes a frame as an |
| 3879 | argument instead of a stack. Stacks are still supported in deprecated |
| 3880 | builds. Additionally, `display-error' will again source location |
| 3881 | information for the error. |
| 3882 | |
| 3883 | ** No more `(ice-9 debug)' |
| 3884 | |
| 3885 | This module had some debugging helpers that are no longer applicable to |
| 3886 | the current debugging model. Importing this module will produce a |
| 3887 | deprecation warning. Users should contact bug-guile for support. |
| 3888 | |
| 3889 | ** Remove obsolete debug-options |
| 3890 | |
| 3891 | Removed `breakpoints', `trace', `procnames', `indent', `frames', |
| 3892 | `maxdepth', and `debug' debug-options. |
| 3893 | |
| 3894 | ** `backtrace' debug option on by default |
| 3895 | |
| 3896 | Given that Guile 2.0 can always give you a backtrace, backtraces are now |
| 3897 | on by default. |
| 3898 | |
| 3899 | ** `turn-on-debugging' deprecated |
| 3900 | |
| 3901 | ** Remove obsolete print-options |
| 3902 | |
| 3903 | The `source' and `closure-hook' print options are obsolete, and have |
| 3904 | been removed. |
| 3905 | |
| 3906 | ** Remove obsolete read-options |
| 3907 | |
| 3908 | The "elisp-strings" and "elisp-vectors" read options were unused and |
| 3909 | obsolete, so they have been removed. |
| 3910 | |
| 3911 | ** Remove eval-options and trap-options |
| 3912 | |
| 3913 | Eval-options and trap-options are obsolete with the new VM and |
| 3914 | evaluator. |
| 3915 | |
| 3916 | ** Remove (ice-9 debugger) and (ice-9 debugging) |
| 3917 | |
| 3918 | See "Traps" and "Interactive Debugging" in the manual, for information |
| 3919 | on their replacements. |
| 3920 | |
| 3921 | ** Remove the GDS Emacs integration |
| 3922 | |
| 3923 | See "Using Guile in Emacs" in the manual, for info on how we think you |
| 3924 | should use Guile with Emacs. |
| 3925 | |
| 3926 | ** Deprecated: `lazy-catch' |
| 3927 | |
| 3928 | `lazy-catch' was a form that captured the stack at the point of a |
| 3929 | `throw', but the dynamic state at the point of the `catch'. It was a bit |
| 3930 | crazy. Please change to use `catch', possibly with a throw-handler, or |
| 3931 | `with-throw-handler'. |
| 3932 | |
| 3933 | ** Deprecated: primitive properties |
| 3934 | |
| 3935 | The `primitive-make-property', `primitive-property-set!', |
| 3936 | `primitive-property-ref', and `primitive-property-del!' procedures were |
| 3937 | crufty and only used to implement object properties, which has a new, |
| 3938 | threadsafe implementation. Use object properties or weak hash tables |
| 3939 | instead. |
| 3940 | |
| 3941 | ** Deprecated `@bind' syntax |
| 3942 | |
| 3943 | `@bind' was part of an older implementation of the Emacs Lisp language, |
| 3944 | and is no longer used. |
| 3945 | |
| 3946 | ** Miscellaneous other deprecations |
| 3947 | |
| 3948 | `cuserid' has been deprecated, as it only returns 8 bytes of a user's |
| 3949 | login. Use `(passwd:name (getpwuid (geteuid)))' instead. |
| 3950 | |
| 3951 | Additionally, the procedures `apply-to-args', `has-suffix?', `scheme-file-suffix' |
| 3952 | `get-option', `for-next-option', `display-usage-report', |
| 3953 | `transform-usage-lambda', `collect', and `set-batch-mode?!' have all |
| 3954 | been deprecated. |
| 3955 | |
| 3956 | ** Add support for unbound fluids |
| 3957 | |
| 3958 | See `make-unbound-fluid', `fluid-unset!', and `fluid-bound?' in the |
| 3959 | manual. |
| 3960 | |
| 3961 | ** Add `variable-unset!' |
| 3962 | |
| 3963 | See "Variables" in the manual, for more details. |
| 3964 | |
| 3965 | ** Last but not least, the `λ' macro can be used in lieu of `lambda' |
| 3966 | |
| 3967 | * Changes to the C interface |
| 3968 | |
| 3969 | ** Guile now uses libgc, the Boehm-Demers-Weiser garbage collector |
| 3970 | |
| 3971 | The semantics of `scm_gc_malloc ()' have been changed, in a |
| 3972 | backward-compatible way. A new allocation routine, |
| 3973 | `scm_gc_malloc_pointerless ()', was added. |
| 3974 | |
| 3975 | Libgc is a conservative GC, which we hope will make interaction with C |
| 3976 | code easier and less error-prone. |
| 3977 | |
| 3978 | ** New procedures: `scm_to_stringn', `scm_from_stringn' |
| 3979 | ** New procedures: scm_{to,from}_{utf8,latin1}_symbol{n,} |
| 3980 | ** New procedures: scm_{to,from}_{utf8,utf32,latin1}_string{n,} |
| 3981 | |
| 3982 | These new procedures convert to and from string representations in |
| 3983 | particular encodings. |
| 3984 | |
| 3985 | Users should continue to use locale encoding for user input, user |
| 3986 | output, or interacting with the C library. |
| 3987 | |
| 3988 | Use the Latin-1 functions for ASCII, and for literals in source code. |
| 3989 | |
| 3990 | Use UTF-8 functions for interaction with modern libraries which deal in |
| 3991 | UTF-8, and UTF-32 for interaction with utf32-using libraries. |
| 3992 | |
| 3993 | Otherwise, use scm_to_stringn or scm_from_stringn with a specific |
| 3994 | encoding. |
| 3995 | |
| 3996 | ** New type definitions for `scm_t_intptr' and friends. |
| 3997 | |
| 3998 | `SCM_T_UINTPTR_MAX', `SCM_T_INTPTR_MIN', `SCM_T_INTPTR_MAX', |
| 3999 | `SIZEOF_SCM_T_BITS', `scm_t_intptr' and `scm_t_uintptr' are now |
| 4000 | available to C. Have fun! |
| 4001 | |
| 4002 | ** The GH interface (deprecated in version 1.6, 2001) was removed. |
| 4003 | |
| 4004 | ** Internal `scm_i_' functions now have "hidden" linkage with GCC/ELF |
| 4005 | |
| 4006 | This makes these internal functions technically not callable from |
| 4007 | application code. |
| 4008 | |
| 4009 | ** Functions for handling `scm_option' now no longer require an argument |
| 4010 | indicating length of the `scm_t_option' array. |
| 4011 | |
| 4012 | ** Procedures-with-setters are now implemented using applicable structs |
| 4013 | |
| 4014 | From a user's perspective this doesn't mean very much. But if, for some |
| 4015 | odd reason, you used the SCM_PROCEDURE_WITH_SETTER_P, SCM_PROCEDURE, or |
| 4016 | SCM_SETTER macros, know that they're deprecated now. Also, scm_tc7_pws |
| 4017 | is gone. |
| 4018 | |
| 4019 | ** Remove old evaluator closures |
| 4020 | |
| 4021 | There used to be ranges of typecodes allocated to interpreted data |
| 4022 | structures, but that it no longer the case, given that interpreted |
| 4023 | procedure are now just regular VM closures. As a result, there is a |
| 4024 | newly free tc3, and a number of removed macros. See the ChangeLog for |
| 4025 | details. |
| 4026 | |
| 4027 | ** Primitive procedures are now VM trampoline procedures |
| 4028 | |
| 4029 | It used to be that there were something like 12 different typecodes |
| 4030 | allocated to primitive procedures, each with its own calling convention. |
| 4031 | Now there is only one, the gsubr. This may affect user code if you were |
| 4032 | defining a procedure using scm_c_make_subr rather scm_c_make_gsubr. The |
| 4033 | solution is to switch to use scm_c_make_gsubr. This solution works well |
| 4034 | both with the old 1.8 and with the current 1.9 branch. |
| 4035 | |
| 4036 | Guile's old evaluator used to have special cases for applying "gsubrs", |
| 4037 | primitive procedures with specified numbers of required, optional, and |
| 4038 | rest arguments. Now, however, Guile represents gsubrs as normal VM |
| 4039 | procedures, with appropriate bytecode to parse out the correct number of |
| 4040 | arguments, including optional and rest arguments, and then with a |
| 4041 | special bytecode to apply the gsubr. |
| 4042 | |
| 4043 | This allows primitive procedures to appear on the VM stack, allowing |
| 4044 | them to be accurately counted in profiles. Also they now have more |
| 4045 | debugging information attached to them -- their number of arguments, for |
| 4046 | example. In addition, the VM can completely inline the application |
| 4047 | mechanics, allowing for faster primitive calls. |
| 4048 | |
| 4049 | However there are some changes on the C level. There is no more |
| 4050 | `scm_tc7_gsubr' or `scm_tcs_subrs' typecode for primitive procedures, as |
| 4051 | they are just VM procedures. Likewise the macros `SCM_GSUBR_TYPE', |
| 4052 | `SCM_GSUBR_MAKTYPE', `SCM_GSUBR_REQ', `SCM_GSUBR_OPT', and |
| 4053 | `SCM_GSUBR_REST' are gone, as are `SCM_SUBR_META_INFO', `SCM_SUBR_PROPS' |
| 4054 | `SCM_SET_SUBR_GENERIC_LOC', and `SCM_SUBR_ARITY_TO_TYPE'. |
| 4055 | |
| 4056 | Perhaps more significantly, `scm_c_make_subr', |
| 4057 | `scm_c_make_subr_with_generic', `scm_c_define_subr', and |
| 4058 | `scm_c_define_subr_with_generic'. They all operated on subr typecodes, |
| 4059 | and there are no more subr typecodes. Use the scm_c_make_gsubr family |
| 4060 | instead. |
| 4061 | |
| 4062 | Normal users of gsubrs should not be affected, though, as the |
| 4063 | scm_c_make_gsubr family still is the correct way to create primitive |
| 4064 | procedures. |
| 4065 | |
| 4066 | ** Remove deprecated array C interfaces |
| 4067 | |
| 4068 | Removed the deprecated array functions `scm_i_arrayp', |
| 4069 | `scm_i_array_ndim', `scm_i_array_mem', `scm_i_array_v', |
| 4070 | `scm_i_array_base', `scm_i_array_dims', and the deprecated macros |
| 4071 | `SCM_ARRAYP', `SCM_ARRAY_NDIM', `SCM_ARRAY_CONTP', `SCM_ARRAY_MEM', |
| 4072 | `SCM_ARRAY_V', `SCM_ARRAY_BASE', and `SCM_ARRAY_DIMS'. |
| 4073 | |
| 4074 | ** Remove unused snarf macros |
| 4075 | |
| 4076 | `SCM_DEFINE1', `SCM_PRIMITIVE_GENERIC_1', `SCM_PROC1, and `SCM_GPROC1' |
| 4077 | are no more. Use SCM_DEFINE or SCM_PRIMITIVE_GENERIC instead. |
| 4078 | |
| 4079 | ** New functions: `scm_call_n', `scm_c_run_hookn' |
| 4080 | |
| 4081 | `scm_call_n' applies to apply a function to an array of arguments. |
| 4082 | `scm_c_run_hookn' runs a hook with an array of arguments. |
| 4083 | |
| 4084 | ** Some SMOB types changed to have static typecodes |
| 4085 | |
| 4086 | Fluids, dynamic states, and hash tables used to be SMOB objects, but now |
| 4087 | they have statically allocated tc7 typecodes. |
| 4088 | |
| 4089 | ** Preparations for changing SMOB representation |
| 4090 | |
| 4091 | If things go right, we'll be changing the SMOB representation soon. To |
| 4092 | that end, we did a lot of cleanups to calls to e.g. SCM_CELL_WORD_2(x) when |
| 4093 | the code meant SCM_SMOB_DATA_2(x); user code will need similar changes |
| 4094 | in the future. Code accessing SMOBs using SCM_CELL macros was never |
| 4095 | correct, but until now things still worked. Users should be aware of |
| 4096 | such changes. |
| 4097 | |
| 4098 | ** Changed invocation mechanics of applicable SMOBs |
| 4099 | |
| 4100 | Guile's old evaluator used to have special cases for applying SMOB |
| 4101 | objects. Now, with the VM, when Guile sees a SMOB, it looks up a VM |
| 4102 | trampoline procedure for it, and use the normal mechanics to apply the |
| 4103 | trampoline. This simplifies procedure application in the normal, |
| 4104 | non-SMOB case. |
| 4105 | |
| 4106 | The upshot is that the mechanics used to apply a SMOB are different from |
| 4107 | 1.8. Descriptors no longer have `apply_0', `apply_1', `apply_2', and |
| 4108 | `apply_3' functions, and the macros SCM_SMOB_APPLY_0 and friends are now |
| 4109 | deprecated. Just use the scm_call_0 family of procedures. |
| 4110 | |
| 4111 | ** Removed support shlibs for SRFIs 1, 4, 13, 14, and 60 |
| 4112 | |
| 4113 | Though these SRFI support libraries did expose API, they encoded a |
| 4114 | strange version string into their library names. That version was never |
| 4115 | programmatically exported, so there was no way people could use the |
| 4116 | libs. |
| 4117 | |
| 4118 | This was a fortunate oversight, as it allows us to remove the need for |
| 4119 | extra, needless shared libraries --- the C support code for SRFIs 4, 13, |
| 4120 | and 14 was already in core --- and allow us to incrementally return the |
| 4121 | SRFI implementation to Scheme. |
| 4122 | |
| 4123 | ** New C function: scm_module_public_interface |
| 4124 | |
| 4125 | This procedure corresponds to Scheme's `module-public-interface'. |
| 4126 | |
| 4127 | ** Undeprecate `scm_the_root_module ()' |
| 4128 | |
| 4129 | It's useful to be able to get the root module from C without doing a |
| 4130 | full module lookup. |
| 4131 | |
| 4132 | ** Inline vector allocation |
| 4133 | |
| 4134 | Instead of having vectors point out into the heap for their data, their |
| 4135 | data is now allocated inline to the vector object itself. The same is |
| 4136 | true for bytevectors, by default, though there is an indirection |
| 4137 | available which should allow for making a bytevector from an existing |
| 4138 | memory region. |
| 4139 | |
| 4140 | ** New struct constructors that don't involve making lists |
| 4141 | |
| 4142 | `scm_c_make_struct' and `scm_c_make_structv' are new varargs and array |
| 4143 | constructors, respectively, for structs. You might find them useful. |
| 4144 | |
| 4145 | ** Stack refactor |
| 4146 | |
| 4147 | In Guile 1.8, there were debugging frames on the C stack. Now there is |
| 4148 | no more need to explicitly mark the stack in this way, because Guile has |
| 4149 | a VM stack that it knows how to walk, which simplifies the C API |
| 4150 | considerably. See the ChangeLog for details; the relevant interface is |
| 4151 | in libguile/stacks.h. The Scheme API has not been changed significantly. |
| 4152 | |
| 4153 | ** Removal of Guile's primitive object system. |
| 4154 | |
| 4155 | There were a number of pieces in `objects.[ch]' that tried to be a |
| 4156 | minimal object system, but were never documented, and were quickly |
| 4157 | obseleted by GOOPS' merge into Guile proper. So `scm_make_class_object', |
| 4158 | `scm_make_subclass_object', `scm_metaclass_standard', and like symbols |
| 4159 | from objects.h are no more. In the very unlikely case in which these |
| 4160 | were useful to you, we urge you to contact guile-devel. |
| 4161 | |
| 4162 | ** No future. |
| 4163 | |
| 4164 | Actually the future is still in the state that it was, is, and ever |
| 4165 | shall be, Amen, except that `futures.c' and `futures.h' are no longer a |
| 4166 | part of it. These files were experimental, never compiled, and would be |
| 4167 | better implemented in Scheme anyway. In the future, that is. |
| 4168 | |
| 4169 | ** Deprecate trampolines |
| 4170 | |
| 4171 | There used to be C functions `scm_trampoline_0', `scm_trampoline_1', and |
| 4172 | so on. The point was to do some precomputation on the type of the |
| 4173 | procedure, then return a specialized "call" procedure. However this |
| 4174 | optimization wasn't actually an optimization, so it is now deprecated. |
| 4175 | Just use `scm_call_0', etc instead. |
| 4176 | |
| 4177 | ** Deprecated `scm_badargsp' |
| 4178 | |
| 4179 | This function is unused in Guile, but was part of its API. |
| 4180 | |
| 4181 | ** Better support for Lisp `nil'. |
| 4182 | |
| 4183 | The bit representation of `nil' has been tweaked so that it is now very |
| 4184 | efficient to check e.g. if a value is equal to Scheme's end-of-list or |
| 4185 | Lisp's nil. Additionally there are a heap of new, specific predicates |
| 4186 | like scm_is_null_or_nil. |
| 4187 | |
| 4188 | ** Better integration of Lisp `nil'. |
| 4189 | |
| 4190 | `scm_is_boolean', `scm_is_false', and `scm_is_null' all return true now |
| 4191 | for Lisp's `nil'. This shouldn't affect any Scheme code at this point, |
| 4192 | but when we start to integrate more with Emacs, it is possible that we |
| 4193 | break code that assumes that, for example, `(not x)' implies that `x' is |
| 4194 | `eq?' to `#f'. This is not a common assumption. Refactoring affected |
| 4195 | code to rely on properties instead of identities will improve code |
| 4196 | correctness. See "Nil" in the manual, for more details. |
| 4197 | |
| 4198 | ** Support for static allocation of strings, symbols, and subrs. |
| 4199 | |
| 4200 | Calls to snarfing CPP macros like SCM_DEFINE macro will now allocate |
| 4201 | much of their associated data as static variables, reducing Guile's |
| 4202 | memory footprint. |
| 4203 | |
| 4204 | ** `scm_stat' has an additional argument, `exception_on_error' |
| 4205 | ** `scm_primitive_load_path' has an additional argument `exception_on_not_found' |
| 4206 | |
| 4207 | ** `scm_set_port_seek' and `scm_set_port_truncate' use the `scm_t_off' type |
| 4208 | |
| 4209 | Previously they would use the `off_t' type, which is fragile since its |
| 4210 | definition depends on the application's value for `_FILE_OFFSET_BITS'. |
| 4211 | |
| 4212 | ** The `long_long' C type, deprecated in 1.8, has been removed |
| 4213 | |
| 4214 | ** Removed deprecated uniform array procedures: scm_make_uve, |
| 4215 | scm_array_prototype, scm_list_to_uniform_array, |
| 4216 | scm_dimensions_to_uniform_array, scm_make_ra, scm_shap2ra, scm_cvref, |
| 4217 | scm_ra_set_contp, scm_aind, scm_raprin1 |
| 4218 | |
| 4219 | These functions have been deprecated since early 2005. |
| 4220 | |
| 4221 | * Changes to the distribution |
| 4222 | |
| 4223 | ** Guile's license is now LGPLv3+ |
| 4224 | |
| 4225 | In other words the GNU Lesser General Public License, version 3 or |
| 4226 | later (at the discretion of each person that chooses to redistribute |
| 4227 | part of Guile). |
| 4228 | |
| 4229 | ** AM_SILENT_RULES |
| 4230 | |
| 4231 | Guile's build is visually quieter, due to the use of Automake 1.11's |
| 4232 | AM_SILENT_RULES. Build as `make V=1' to see all of the output. |
| 4233 | |
| 4234 | ** GOOPS documentation folded into Guile reference manual |
| 4235 | |
| 4236 | GOOPS, Guile's object system, used to be documented in separate manuals. |
| 4237 | This content is now included in Guile's manual directly. |
| 4238 | |
| 4239 | ** `guile-config' will be deprecated in favor of `pkg-config' |
| 4240 | |
| 4241 | `guile-config' has been rewritten to get its information from |
| 4242 | `pkg-config', so this should be a transparent change. Note however that |
| 4243 | guile.m4 has yet to be modified to call pkg-config instead of |
| 4244 | guile-config. |
| 4245 | |
| 4246 | ** Guile now provides `guile-2.0.pc' instead of `guile-1.8.pc' |
| 4247 | |
| 4248 | Programs that use `pkg-config' to find Guile or one of its Autoconf |
| 4249 | macros should now require `guile-2.0' instead of `guile-1.8'. |
| 4250 | |
| 4251 | ** New installation directory: $(pkglibdir)/1.9/ccache |
| 4252 | |
| 4253 | If $(libdir) is /usr/lib, for example, Guile will install its .go files |
| 4254 | to /usr/lib/guile/1.9/ccache. These files are architecture-specific. |
| 4255 | |
| 4256 | ** Parallel installability fixes |
| 4257 | |
| 4258 | Guile now installs its header files to a effective-version-specific |
| 4259 | directory, and includes the effective version (e.g. 2.0) in the library |
| 4260 | name (e.g. libguile-2.0.so). |
| 4261 | |
| 4262 | This change should be transparent to users, who should detect Guile via |
| 4263 | the guile.m4 macro, or the guile-2.0.pc pkg-config file. It will allow |
| 4264 | parallel installs for multiple versions of Guile development |
| 4265 | environments. |
| 4266 | |
| 4267 | ** Dynamically loadable extensions may be placed in a Guile-specific path |
| 4268 | |
| 4269 | Before, Guile only searched the system library paths for extensions |
| 4270 | (e.g. /usr/lib), which meant that the names of Guile extensions had to |
| 4271 | be globally unique. Installing them to a Guile-specific extensions |
| 4272 | directory is cleaner. Use `pkg-config --variable=extensiondir |
| 4273 | guile-2.0' to get the location of the extensions directory. |
| 4274 | |
| 4275 | ** User Scheme code may be placed in a version-specific path |
| 4276 | |
| 4277 | Before, there was only one way to install user Scheme code to a |
| 4278 | version-specific Guile directory: install to Guile's own path, |
| 4279 | e.g. /usr/share/guile/2.0. The site directory, |
| 4280 | e.g. /usr/share/guile/site, was unversioned. This has been changed to |
| 4281 | add a version-specific site directory, e.g. /usr/share/guile/site/2.0, |
| 4282 | searched before the global site directory. |
| 4283 | |
| 4284 | ** New dependency: libgc |
| 4285 | |
| 4286 | See http://www.hpl.hp.com/personal/Hans_Boehm/gc/, for more information. |
| 4287 | |
| 4288 | ** New dependency: GNU libunistring |
| 4289 | |
| 4290 | See http://www.gnu.org/software/libunistring/, for more information. Our |
| 4291 | Unicode support uses routines from libunistring. |
| 4292 | |
| 4293 | ** New dependency: libffi |
| 4294 | |
| 4295 | See http://sourceware.org/libffi/, for more information. |
| 4296 | |
| 4297 | |
| 4298 | \f |
| 4299 | Changes in 1.8.8 (since 1.8.7) |
| 4300 | |
| 4301 | * Bugs fixed |
| 4302 | |
| 4303 | ** Fix possible buffer overruns when parsing numbers |
| 4304 | ** Avoid clash with system setjmp/longjmp on IA64 |
| 4305 | ** Fix `wrong type arg' exceptions with IPv6 addresses |
| 4306 | |
| 4307 | \f |
| 4308 | Changes in 1.8.7 (since 1.8.6) |
| 4309 | |
| 4310 | * New modules (see the manual for details) |
| 4311 | |
| 4312 | ** `(srfi srfi-98)', an interface to access environment variables |
| 4313 | |
| 4314 | * Bugs fixed |
| 4315 | |
| 4316 | ** Fix compilation with `--disable-deprecated' |
| 4317 | ** Fix %fast-slot-ref/set!, to avoid possible segmentation fault |
| 4318 | ** Fix MinGW build problem caused by HAVE_STRUCT_TIMESPEC confusion |
| 4319 | ** Fix build problem when scm_t_timespec is different from struct timespec |
| 4320 | ** Fix build when compiled with -Wundef -Werror |
| 4321 | ** More build fixes for `alphaev56-dec-osf5.1b' (Tru64) |
| 4322 | ** Build fixes for `powerpc-ibm-aix5.3.0.0' (AIX 5.3) |
| 4323 | ** With GCC, always compile with `-mieee' on `alpha*' and `sh*' |
| 4324 | ** Better diagnose broken `(strftime "%z" ...)' in `time.test' (bug #24130) |
| 4325 | ** Fix parsing of SRFI-88/postfix keywords longer than 128 characters |
| 4326 | ** Fix reading of complex numbers where both parts are inexact decimals |
| 4327 | |
| 4328 | ** Allow @ macro to work with (ice-9 syncase) |
| 4329 | |
| 4330 | Previously, use of the @ macro in a module whose code is being |
| 4331 | transformed by (ice-9 syncase) would cause an "Invalid syntax" error. |
| 4332 | Now it works as you would expect (giving the value of the specified |
| 4333 | module binding). |
| 4334 | |
| 4335 | ** Have `scm_take_locale_symbol ()' return an interned symbol (bug #25865) |
| 4336 | |
| 4337 | \f |
| 4338 | Changes in 1.8.6 (since 1.8.5) |
| 4339 | |
| 4340 | * New features (see the manual for details) |
| 4341 | |
| 4342 | ** New convenience function `scm_c_symbol_length ()' |
| 4343 | |
| 4344 | ** Single stepping through code from Emacs |
| 4345 | |
| 4346 | When you use GDS to evaluate Scheme code from Emacs, you can now use |
| 4347 | `C-u' to indicate that you want to single step through that code. See |
| 4348 | `Evaluating Scheme Code' in the manual for more details. |
| 4349 | |
| 4350 | ** New "guile(1)" man page! |
| 4351 | |
| 4352 | * Changes to the distribution |
| 4353 | |
| 4354 | ** Automake's `AM_MAINTAINER_MODE' is no longer used |
| 4355 | |
| 4356 | Thus, the `--enable-maintainer-mode' configure option is no longer |
| 4357 | available: Guile is now always configured in "maintainer mode". |
| 4358 | |
| 4359 | ** `ChangeLog' files are no longer updated |
| 4360 | |
| 4361 | Instead, changes are detailed in the version control system's logs. See |
| 4362 | the top-level `ChangeLog' files for details. |
| 4363 | |
| 4364 | |
| 4365 | * Bugs fixed |
| 4366 | |
| 4367 | ** `symbol->string' now returns a read-only string, as per R5RS |
| 4368 | ** Fix incorrect handling of the FLAGS argument of `fold-matches' |
| 4369 | ** `guile-config link' now prints `-L$libdir' before `-lguile' |
| 4370 | ** Fix memory corruption involving GOOPS' `class-redefinition' |
| 4371 | ** Fix possible deadlock in `mutex-lock' |
| 4372 | ** Fix build issue on Tru64 and ia64-hp-hpux11.23 (`SCM_UNPACK' macro) |
| 4373 | ** Fix build issue on mips, mipsel, powerpc and ia64 (stack direction) |
| 4374 | ** Fix build issue on hppa2.0w-hp-hpux11.11 (`dirent64' and `readdir64_r') |
| 4375 | ** Fix build issue on i386-unknown-freebsd7.0 ("break strict-aliasing rules") |
| 4376 | ** Fix misleading output from `(help rationalize)' |
| 4377 | ** Fix build failure on Debian hppa architecture (bad stack growth detection) |
| 4378 | ** Fix `gcd' when called with a single, negative argument. |
| 4379 | ** Fix `Stack overflow' errors seen when building on some platforms |
| 4380 | ** Fix bug when `scm_with_guile ()' was called several times from the |
| 4381 | same thread |
| 4382 | ** The handler of SRFI-34 `with-exception-handler' is now invoked in the |
| 4383 | dynamic environment of the call to `raise' |
| 4384 | ** Fix potential deadlock in `make-struct' |
| 4385 | ** Fix compilation problem with libltdl from Libtool 2.2.x |
| 4386 | ** Fix sloppy bound checking in `string-{ref,set!}' with the empty string |
| 4387 | |
| 4388 | \f |
| 4389 | Changes in 1.8.5 (since 1.8.4) |
| 4390 | |
| 4391 | * Infrastructure changes |
| 4392 | |
| 4393 | ** Guile repository switched from CVS to Git |
| 4394 | |
| 4395 | The new repository can be accessed using |
| 4396 | "git-clone git://git.sv.gnu.org/guile.git", or can be browsed on-line at |
| 4397 | http://git.sv.gnu.org/gitweb/?p=guile.git . See `README' for details. |
| 4398 | |
| 4399 | ** Add support for `pkg-config' |
| 4400 | |
| 4401 | See "Autoconf Support" in the manual for details. |
| 4402 | |
| 4403 | * New modules (see the manual for details) |
| 4404 | |
| 4405 | ** `(srfi srfi-88)' |
| 4406 | |
| 4407 | * New features (see the manual for details) |
| 4408 | |
| 4409 | ** New `postfix' read option, for SRFI-88 keyword syntax |
| 4410 | ** Some I/O primitives have been inlined, which improves I/O performance |
| 4411 | ** New object-based traps infrastructure |
| 4412 | |
| 4413 | This is a GOOPS-based infrastructure that builds on Guile's low-level |
| 4414 | evaluator trap calls and facilitates the development of debugging |
| 4415 | features like single-stepping, breakpoints, tracing and profiling. |
| 4416 | See the `Traps' node of the manual for details. |
| 4417 | |
| 4418 | ** New support for working on Guile code from within Emacs |
| 4419 | |
| 4420 | Guile now incorporates the `GDS' library (previously distributed |
| 4421 | separately) for working on Guile code from within Emacs. See the |
| 4422 | `Using Guile In Emacs' node of the manual for details. |
| 4423 | |
| 4424 | * Bugs fixed |
| 4425 | |
| 4426 | ** `scm_add_slot ()' no longer segfaults (fixes bug #22369) |
| 4427 | ** Fixed `(ice-9 match)' for patterns like `((_ ...) ...)' |
| 4428 | |
| 4429 | Previously, expressions like `(match '((foo) (bar)) (((_ ...) ...) #t))' |
| 4430 | would trigger an unbound variable error for `match:andmap'. |
| 4431 | |
| 4432 | ** `(oop goops describe)' now properly provides the `describe' feature |
| 4433 | ** Fixed `args-fold' from `(srfi srfi-37)' |
| 4434 | |
| 4435 | Previously, parsing short option names of argument-less options would |
| 4436 | lead to a stack overflow. |
| 4437 | |
| 4438 | ** `(srfi srfi-35)' is now visible through `cond-expand' |
| 4439 | ** Fixed type-checking for the second argument of `eval' |
| 4440 | ** Fixed type-checking for SRFI-1 `partition' |
| 4441 | ** Fixed `struct-ref' and `struct-set!' on "light structs" |
| 4442 | ** Honor struct field access rights in GOOPS |
| 4443 | ** Changed the storage strategy of source properties, which fixes a deadlock |
| 4444 | ** Allow compilation of Guile-using programs in C99 mode with GCC 4.3 and later |
| 4445 | ** Fixed build issue for GNU/Linux on IA64 |
| 4446 | ** Fixed build issues on NetBSD 1.6 |
| 4447 | ** Fixed build issue on Solaris 2.10 x86_64 |
| 4448 | ** Fixed build issue with DEC/Compaq/HP's compiler |
| 4449 | ** Fixed `scm_from_complex_double' build issue on FreeBSD |
| 4450 | ** Fixed `alloca' build issue on FreeBSD 6 |
| 4451 | ** Removed use of non-portable makefile constructs |
| 4452 | ** Fixed shadowing of libc's <random.h> on Tru64, which broke compilation |
| 4453 | ** Make sure all tests honor `$TMPDIR' |
| 4454 | |
| 4455 | \f |
| 4456 | Changes in 1.8.4 (since 1.8.3) |
| 4457 | |
| 4458 | * Bugs fixed |
| 4459 | |
| 4460 | ** CR (ASCII 0x0d) is (again) recognized as a token delimiter by the reader |
| 4461 | ** Fixed a segmentation fault which occurred when displaying the |
| 4462 | backtrace of a stack with a promise object (made by `delay') in it. |
| 4463 | ** Make `accept' leave guile mode while blocking |
| 4464 | ** `scm_c_read ()' and `scm_c_write ()' now type-check their port argument |
| 4465 | ** Fixed a build problem on AIX (use of func_data identifier) |
| 4466 | ** Fixed a segmentation fault which occurred when hashx-ref or hashx-set! was |
| 4467 | called with an associator proc that returns neither a pair nor #f. |
| 4468 | ** Secondary threads now always return a valid module for (current-module). |
| 4469 | ** Avoid MacOS build problems caused by incorrect combination of "64" |
| 4470 | system and library calls. |
| 4471 | ** `guile-snarf' now honors `$TMPDIR' |
| 4472 | ** `guile-config compile' now reports CPPFLAGS used at compile-time |
| 4473 | ** Fixed build with Sun Studio (Solaris 9) |
| 4474 | ** Fixed wrong-type-arg errors when creating zero length SRFI-4 |
| 4475 | uniform vectors on AIX. |
| 4476 | ** Fixed a deadlock that occurs upon GC with multiple threads. |
| 4477 | ** Fixed compile problem with GCC on Solaris and AIX (use of _Complex_I) |
| 4478 | ** Fixed autotool-derived build problems on AIX 6.1. |
| 4479 | ** Fixed NetBSD/alpha support |
| 4480 | ** Fixed MacOS build problem caused by use of rl_get_keymap(_name) |
| 4481 | |
| 4482 | * New modules (see the manual for details) |
| 4483 | |
| 4484 | ** `(srfi srfi-69)' |
| 4485 | |
| 4486 | * Documentation fixes and improvements |
| 4487 | |
| 4488 | ** Removed premature breakpoint documentation |
| 4489 | |
| 4490 | The features described are not available in the series of 1.8.x |
| 4491 | releases, so the documentation was misleading and has been removed. |
| 4492 | |
| 4493 | ** More about Guile's default *random-state* variable |
| 4494 | |
| 4495 | ** GOOPS: more about how to use `next-method' |
| 4496 | |
| 4497 | * Changes to the distribution |
| 4498 | |
| 4499 | ** Corrected a few files that referred incorrectly to the old GPL + special exception licence |
| 4500 | |
| 4501 | In fact Guile since 1.8.0 has been licensed with the GNU Lesser |
| 4502 | General Public License, and the few incorrect files have now been |
| 4503 | fixed to agree with the rest of the Guile distribution. |
| 4504 | |
| 4505 | ** Removed unnecessary extra copies of COPYING* |
| 4506 | |
| 4507 | The distribution now contains a single COPYING.LESSER at its top level. |
| 4508 | |
| 4509 | \f |
| 4510 | Changes in 1.8.3 (since 1.8.2) |
| 4511 | |
| 4512 | * New modules (see the manual for details) |
| 4513 | |
| 4514 | ** `(srfi srfi-35)' |
| 4515 | ** `(srfi srfi-37)' |
| 4516 | |
| 4517 | * Bugs fixed |
| 4518 | |
| 4519 | ** The `(ice-9 slib)' module now works as expected |
| 4520 | ** Expressions like "(set! 'x #t)" no longer yield a crash |
| 4521 | ** Warnings about duplicate bindings now go to stderr |
| 4522 | ** A memory leak in `make-socket-address' was fixed |
| 4523 | ** Alignment issues (e.g., on SPARC) in network routines were fixed |
| 4524 | ** A threading issue that showed up at least on NetBSD was fixed |
| 4525 | ** Build problems on Solaris and IRIX fixed |
| 4526 | |
| 4527 | * Implementation improvements |
| 4528 | |
| 4529 | ** The reader is now faster, which reduces startup time |
| 4530 | ** Procedures returned by `record-accessor' and `record-modifier' are faster |
| 4531 | |
| 4532 | \f |
| 4533 | Changes in 1.8.2 (since 1.8.1): |
| 4534 | |
| 4535 | * New procedures (see the manual for details) |
| 4536 | |
| 4537 | ** set-program-arguments |
| 4538 | ** make-vtable |
| 4539 | |
| 4540 | * Incompatible changes |
| 4541 | |
| 4542 | ** The body of a top-level `define' no longer sees the binding being created |
| 4543 | |
| 4544 | In a top-level `define', the binding being created is no longer visible |
| 4545 | from the `define' body. This breaks code like |
| 4546 | "(define foo (begin (set! foo 1) (+ foo 1)))", where `foo' is now |
| 4547 | unbound in the body. However, such code was not R5RS-compliant anyway, |
| 4548 | per Section 5.2.1. |
| 4549 | |
| 4550 | * Bugs fixed |
| 4551 | |
| 4552 | ** Fractions were not `equal?' if stored in unreduced form. |
| 4553 | (A subtle problem, since printing a value reduced it, making it work.) |
| 4554 | ** srfi-60 `copy-bit' failed on 64-bit systems |
| 4555 | ** "guile --use-srfi" option at the REPL can replace core functions |
| 4556 | (Programs run with that option were ok, but in the interactive REPL |
| 4557 | the core bindings got priority, preventing SRFI replacements or |
| 4558 | extensions.) |
| 4559 | ** `regexp-exec' doesn't abort() on #\nul in the input or bad flags arg |
| 4560 | ** `kill' on mingw throws an error for a PID other than oneself |
| 4561 | ** Procedure names are attached to procedure-with-setters |
| 4562 | ** Array read syntax works with negative lower bound |
| 4563 | ** `array-in-bounds?' fix if an array has different lower bounds on each index |
| 4564 | ** `*' returns exact 0 for "(* inexact 0)" |
| 4565 | This follows what it always did for "(* 0 inexact)". |
| 4566 | ** SRFI-19: Value returned by `(current-time time-process)' was incorrect |
| 4567 | ** SRFI-19: `date->julian-day' did not account for timezone offset |
| 4568 | ** `ttyname' no longer crashes when passed a non-tty argument |
| 4569 | ** `inet-ntop' no longer crashes on SPARC when passed an `AF_INET' address |
| 4570 | ** Small memory leaks have been fixed in `make-fluid' and `add-history' |
| 4571 | ** GOOPS: Fixed a bug in `method-more-specific?' |
| 4572 | ** Build problems on Solaris fixed |
| 4573 | ** Build problems on HP-UX IA64 fixed |
| 4574 | ** Build problems on MinGW fixed |
| 4575 | |
| 4576 | \f |
| 4577 | Changes in 1.8.1 (since 1.8.0): |
| 4578 | |
| 4579 | * LFS functions are now used to access 64-bit files on 32-bit systems. |
| 4580 | |
| 4581 | * New procedures (see the manual for details) |
| 4582 | |
| 4583 | ** primitive-_exit - [Scheme] the-root-module |
| 4584 | ** scm_primitive__exit - [C] |
| 4585 | ** make-completion-function - [Scheme] (ice-9 readline) |
| 4586 | ** scm_c_locale_stringn_to_number - [C] |
| 4587 | ** scm_srfi1_append_reverse [C] |
| 4588 | ** scm_srfi1_append_reverse_x [C] |
| 4589 | ** scm_log - [C] |
| 4590 | ** scm_log10 - [C] |
| 4591 | ** scm_exp - [C] |
| 4592 | ** scm_sqrt - [C] |
| 4593 | |
| 4594 | * Bugs fixed |
| 4595 | |
| 4596 | ** Build problems have been fixed on MacOS, SunOS, and QNX. |
| 4597 | |
| 4598 | ** `strftime' fix sign of %z timezone offset. |
| 4599 | |
| 4600 | ** A one-dimensional array can now be 'equal?' to a vector. |
| 4601 | |
| 4602 | ** Structures, records, and SRFI-9 records can now be compared with `equal?'. |
| 4603 | |
| 4604 | ** SRFI-14 standard char sets are recomputed upon a successful `setlocale'. |
| 4605 | |
| 4606 | ** `record-accessor' and `record-modifier' now have strict type checks. |
| 4607 | |
| 4608 | Record accessor and modifier procedures now throw an error if the |
| 4609 | record type of the record they're given is not the type expected. |
| 4610 | (Previously accessors returned #f and modifiers silently did nothing). |
| 4611 | |
| 4612 | ** It is now OK to use both autoload and use-modules on a given module. |
| 4613 | |
| 4614 | ** `apply' checks the number of arguments more carefully on "0 or 1" funcs. |
| 4615 | |
| 4616 | Previously there was no checking on primatives like make-vector that |
| 4617 | accept "one or two" arguments. Now there is. |
| 4618 | |
| 4619 | ** The srfi-1 assoc function now calls its equality predicate properly. |
| 4620 | |
| 4621 | Previously srfi-1 assoc would call the equality predicate with the key |
| 4622 | last. According to the SRFI, the key should be first. |
| 4623 | |
| 4624 | ** A bug in n-par-for-each and n-for-each-par-map has been fixed. |
| 4625 | |
| 4626 | ** The array-set! procedure no longer segfaults when given a bit vector. |
| 4627 | |
| 4628 | ** Bugs in make-shared-array have been fixed. |
| 4629 | |
| 4630 | ** string<? and friends now follow char<? etc order on 8-bit chars. |
| 4631 | |
| 4632 | ** The format procedure now handles inf and nan values for ~f correctly. |
| 4633 | |
| 4634 | ** exact->inexact should no longer overflow when given certain large fractions. |
| 4635 | |
| 4636 | ** srfi-9 accessor and modifier procedures now have strict record type checks. |
| 4637 | |
| 4638 | This matches the srfi-9 specification. |
| 4639 | |
| 4640 | ** (ice-9 ftw) procedures won't ignore different files with same inode number. |
| 4641 | |
| 4642 | Previously the (ice-9 ftw) procedures would ignore any file that had |
| 4643 | the same inode number as a file they had already seen, even if that |
| 4644 | file was on a different device. |
| 4645 | |
| 4646 | \f |
| 4647 | Changes in 1.8.0 (changes since the 1.6.x series): |
| 4648 | |
| 4649 | * Changes to the distribution |
| 4650 | |
| 4651 | ** Guile is now licensed with the GNU Lesser General Public License. |
| 4652 | |
| 4653 | ** The manual is now licensed with the GNU Free Documentation License. |
| 4654 | |
| 4655 | ** Guile now requires GNU MP (http://swox.com/gmp). |
| 4656 | |
| 4657 | Guile now uses the GNU MP library for arbitrary precision arithmetic. |
| 4658 | |
| 4659 | ** Guile now has separate private and public configuration headers. |
| 4660 | |
| 4661 | That is, things like HAVE_STRING_H no longer leak from Guile's |
| 4662 | headers. |
| 4663 | |
| 4664 | ** Guile now provides and uses an "effective" version number. |
| 4665 | |
| 4666 | Guile now provides scm_effective_version and effective-version |
| 4667 | functions which return the "effective" version number. This is just |
| 4668 | the normal full version string without the final micro-version number, |
| 4669 | so the current effective-version is "1.8". The effective version |
| 4670 | should remain unchanged during a stable series, and should be used for |
| 4671 | items like the versioned share directory name |
| 4672 | i.e. /usr/share/guile/1.8. |
| 4673 | |
| 4674 | Providing an unchanging version number during a stable release for |
| 4675 | things like the versioned share directory can be particularly |
| 4676 | important for Guile "add-on" packages, since it provides a directory |
| 4677 | that they can install to that won't be changed out from under them |
| 4678 | with each micro release during a stable series. |
| 4679 | |
| 4680 | ** Thread implementation has changed. |
| 4681 | |
| 4682 | When you configure "--with-threads=null", you will get the usual |
| 4683 | threading API (call-with-new-thread, make-mutex, etc), but you can't |
| 4684 | actually create new threads. Also, "--with-threads=no" is now |
| 4685 | equivalent to "--with-threads=null". This means that the thread API |
| 4686 | is always present, although you might not be able to create new |
| 4687 | threads. |
| 4688 | |
| 4689 | When you configure "--with-threads=pthreads" or "--with-threads=yes", |
| 4690 | you will get threads that are implemented with the portable POSIX |
| 4691 | threads. These threads can run concurrently (unlike the previous |
| 4692 | "coop" thread implementation), but need to cooperate for things like |
| 4693 | the GC. |
| 4694 | |
| 4695 | The default is "pthreads", unless your platform doesn't have pthreads, |
| 4696 | in which case "null" threads are used. |
| 4697 | |
| 4698 | See the manual for details, nodes "Initialization", "Multi-Threading", |
| 4699 | "Blocking", and others. |
| 4700 | |
| 4701 | ** There is the new notion of 'discouraged' features. |
| 4702 | |
| 4703 | This is a milder form of deprecation. |
| 4704 | |
| 4705 | Things that are discouraged should not be used in new code, but it is |
| 4706 | OK to leave them in old code for now. When a discouraged feature is |
| 4707 | used, no warning message is printed like there is for 'deprecated' |
| 4708 | features. Also, things that are merely discouraged are nevertheless |
| 4709 | implemented efficiently, while deprecated features can be very slow. |
| 4710 | |
| 4711 | You can omit discouraged features from libguile by configuring it with |
| 4712 | the '--disable-discouraged' option. |
| 4713 | |
| 4714 | ** Deprecation warnings can be controlled at run-time. |
| 4715 | |
| 4716 | (debug-enable 'warn-deprecated) switches them on and (debug-disable |
| 4717 | 'warn-deprecated) switches them off. |
| 4718 | |
| 4719 | ** Support for SRFI 61, extended cond syntax for multiple values has |
| 4720 | been added. |
| 4721 | |
| 4722 | This SRFI is always available. |
| 4723 | |
| 4724 | ** Support for require-extension, SRFI-55, has been added. |
| 4725 | |
| 4726 | The SRFI-55 special form `require-extension' has been added. It is |
| 4727 | available at startup, and provides a portable way to load Scheme |
| 4728 | extensions. SRFI-55 only requires support for one type of extension, |
| 4729 | "srfi"; so a set of SRFIs may be loaded via (require-extension (srfi 1 |
| 4730 | 13 14)). |
| 4731 | |
| 4732 | ** New module (srfi srfi-26) provides support for `cut' and `cute'. |
| 4733 | |
| 4734 | The (srfi srfi-26) module is an implementation of SRFI-26 which |
| 4735 | provides the `cut' and `cute' syntax. These may be used to specialize |
| 4736 | parameters without currying. |
| 4737 | |
| 4738 | ** New module (srfi srfi-31) |
| 4739 | |
| 4740 | This is an implementation of SRFI-31 which provides a special form |
| 4741 | `rec' for recursive evaluation. |
| 4742 | |
| 4743 | ** The modules (srfi srfi-13), (srfi srfi-14) and (srfi srfi-4) have |
| 4744 | been merged with the core, making their functionality always |
| 4745 | available. |
| 4746 | |
| 4747 | The modules are still available, tho, and you could use them together |
| 4748 | with a renaming import, for example. |
| 4749 | |
| 4750 | ** Guile no longer includes its own version of libltdl. |
| 4751 | |
| 4752 | The official version is good enough now. |
| 4753 | |
| 4754 | ** The --enable-htmldoc option has been removed from 'configure'. |
| 4755 | |
| 4756 | Support for translating the documentation into HTML is now always |
| 4757 | provided. Use 'make html'. |
| 4758 | |
| 4759 | ** New module (ice-9 serialize): |
| 4760 | |
| 4761 | (serialize FORM1 ...) and (parallelize FORM1 ...) are useful when you |
| 4762 | don't trust the thread safety of most of your program, but where you |
| 4763 | have some section(s) of code which you consider can run in parallel to |
| 4764 | other sections. See ice-9/serialize.scm for more information. |
| 4765 | |
| 4766 | ** The configure option '--disable-arrays' has been removed. |
| 4767 | |
| 4768 | Support for arrays and uniform numeric arrays is now always included |
| 4769 | in Guile. |
| 4770 | |
| 4771 | * Changes to the stand-alone interpreter |
| 4772 | |
| 4773 | ** New command line option `-L'. |
| 4774 | |
| 4775 | This option adds a directory to the front of the load path. |
| 4776 | |
| 4777 | ** New command line option `--no-debug'. |
| 4778 | |
| 4779 | Specifying `--no-debug' on the command line will keep the debugging |
| 4780 | evaluator turned off, even for interactive sessions. |
| 4781 | |
| 4782 | ** User-init file ~/.guile is now loaded with the debugging evaluator. |
| 4783 | |
| 4784 | Previously, the normal evaluator would have been used. Using the |
| 4785 | debugging evaluator gives better error messages. |
| 4786 | |
| 4787 | ** The '-e' option now 'read's its argument. |
| 4788 | |
| 4789 | This is to allow the new '(@ MODULE-NAME VARIABLE-NAME)' construct to |
| 4790 | be used with '-e'. For example, you can now write a script like |
| 4791 | |
| 4792 | #! /bin/sh |
| 4793 | exec guile -e '(@ (demo) main)' -s "$0" "$@" |
| 4794 | !# |
| 4795 | |
| 4796 | (define-module (demo) |
| 4797 | :export (main)) |
| 4798 | |
| 4799 | (define (main args) |
| 4800 | (format #t "Demo: ~a~%" args)) |
| 4801 | |
| 4802 | |
| 4803 | * Changes to Scheme functions and syntax |
| 4804 | |
| 4805 | ** Guardians have changed back to their original semantics |
| 4806 | |
| 4807 | Guardians now behave like described in the paper by Dybvig et al. In |
| 4808 | particular, they no longer make guarantees about the order in which |
| 4809 | they return objects, and they can no longer be greedy. |
| 4810 | |
| 4811 | They no longer drop cyclic data structures. |
| 4812 | |
| 4813 | The C function scm_make_guardian has been changed incompatibly and no |
| 4814 | longer takes the 'greedy_p' argument. |
| 4815 | |
| 4816 | ** New function hashx-remove! |
| 4817 | |
| 4818 | This function completes the set of 'hashx' functions. |
| 4819 | |
| 4820 | ** The concept of dynamic roots has been factored into continuation |
| 4821 | barriers and dynamic states. |
| 4822 | |
| 4823 | Each thread has a current dynamic state that carries the values of the |
| 4824 | fluids. You can create and copy dynamic states and use them as the |
| 4825 | second argument for 'eval'. See "Fluids and Dynamic States" in the |
| 4826 | manual. |
| 4827 | |
| 4828 | To restrict the influence that captured continuations can have on the |
| 4829 | control flow, you can errect continuation barriers. See "Continuation |
| 4830 | Barriers" in the manual. |
| 4831 | |
| 4832 | The function call-with-dynamic-root now essentially temporarily |
| 4833 | installs a new dynamic state and errects a continuation barrier. |
| 4834 | |
| 4835 | ** The default load path no longer includes "." at the end. |
| 4836 | |
| 4837 | Automatically loading modules from the current directory should not |
| 4838 | happen by default. If you want to allow it in a more controlled |
| 4839 | manner, set the environment variable GUILE_LOAD_PATH or the Scheme |
| 4840 | variable %load-path. |
| 4841 | |
| 4842 | ** The uniform vector and array support has been overhauled. |
| 4843 | |
| 4844 | It now complies with SRFI-4 and the weird prototype based uniform |
| 4845 | array creation has been deprecated. See the manual for more details. |
| 4846 | |
| 4847 | Some non-compatible changes have been made: |
| 4848 | - characters can no longer be stored into byte arrays. |
| 4849 | - strings and bit vectors are no longer considered to be uniform numeric |
| 4850 | vectors. |
| 4851 | - array-rank throws an error for non-arrays instead of returning zero. |
| 4852 | - array-ref does no longer accept non-arrays when no indices are given. |
| 4853 | |
| 4854 | There is the new notion of 'generalized vectors' and corresponding |
| 4855 | procedures like 'generalized-vector-ref'. Generalized vectors include |
| 4856 | strings, bitvectors, ordinary vectors, and uniform numeric vectors. |
| 4857 | |
| 4858 | Arrays use generalized vectors as their storage, so that you still |
| 4859 | have arrays of characters, bits, etc. However, uniform-array-read! |
| 4860 | and uniform-array-write can no longer read/write strings and |
| 4861 | bitvectors. |
| 4862 | |
| 4863 | ** There is now support for copy-on-write substrings, mutation-sharing |
| 4864 | substrings and read-only strings. |
| 4865 | |
| 4866 | Three new procedures are related to this: substring/shared, |
| 4867 | substring/copy, and substring/read-only. See the manual for more |
| 4868 | information. |
| 4869 | |
| 4870 | ** Backtraces will now highlight the value that caused the error. |
| 4871 | |
| 4872 | By default, these values are enclosed in "{...}", such as in this |
| 4873 | example: |
| 4874 | |
| 4875 | guile> (car 'a) |
| 4876 | |
| 4877 | Backtrace: |
| 4878 | In current input: |
| 4879 | 1: 0* [car {a}] |
| 4880 | |
| 4881 | <unnamed port>:1:1: In procedure car in expression (car (quote a)): |
| 4882 | <unnamed port>:1:1: Wrong type (expecting pair): a |
| 4883 | ABORT: (wrong-type-arg) |
| 4884 | |
| 4885 | The prefix and suffix used for highlighting can be set via the two new |
| 4886 | printer options 'highlight-prefix' and 'highlight-suffix'. For |
| 4887 | example, putting this into ~/.guile will output the bad value in bold |
| 4888 | on an ANSI terminal: |
| 4889 | |
| 4890 | (print-set! highlight-prefix "\x1b[1m") |
| 4891 | (print-set! highlight-suffix "\x1b[22m") |
| 4892 | |
| 4893 | |
| 4894 | ** 'gettext' support for internationalization has been added. |
| 4895 | |
| 4896 | See the manual for details. |
| 4897 | |
| 4898 | ** New syntax '@' and '@@': |
| 4899 | |
| 4900 | You can now directly refer to variables exported from a module by |
| 4901 | writing |
| 4902 | |
| 4903 | (@ MODULE-NAME VARIABLE-NAME) |
| 4904 | |
| 4905 | For example (@ (ice-9 pretty-print) pretty-print) will directly access |
| 4906 | the pretty-print variable exported from the (ice-9 pretty-print) |
| 4907 | module. You don't need to 'use' that module first. You can also use |
| 4908 | '@' as a target of 'set!', as in (set! (@ mod var) val). |
| 4909 | |
| 4910 | The related syntax (@@ MODULE-NAME VARIABLE-NAME) works just like '@', |
| 4911 | but it can also access variables that have not been exported. It is |
| 4912 | intended only for kluges and temporary fixes and for debugging, not |
| 4913 | for ordinary code. |
| 4914 | |
| 4915 | ** Keyword syntax has been made more disciplined. |
| 4916 | |
| 4917 | Previously, the name of a keyword was read as a 'token' but printed as |
| 4918 | a symbol. Now, it is read as a general Scheme datum which must be a |
| 4919 | symbol. |
| 4920 | |
| 4921 | Previously: |
| 4922 | |
| 4923 | guile> #:12 |
| 4924 | #:#{12}# |
| 4925 | guile> #:#{12}# |
| 4926 | #:#{\#{12}\#}# |
| 4927 | guile> #:(a b c) |
| 4928 | #:#{}# |
| 4929 | ERROR: In expression (a b c): |
| 4930 | Unbound variable: a |
| 4931 | guile> #: foo |
| 4932 | #:#{}# |
| 4933 | ERROR: Unbound variable: foo |
| 4934 | |
| 4935 | Now: |
| 4936 | |
| 4937 | guile> #:12 |
| 4938 | ERROR: Wrong type (expecting symbol): 12 |
| 4939 | guile> #:#{12}# |
| 4940 | #:#{12}# |
| 4941 | guile> #:(a b c) |
| 4942 | ERROR: Wrong type (expecting symbol): (a b c) |
| 4943 | guile> #: foo |
| 4944 | #:foo |
| 4945 | |
| 4946 | ** The printing of symbols that might look like keywords can be |
| 4947 | controlled. |
| 4948 | |
| 4949 | The new printer option 'quote-keywordish-symbols' controls how symbols |
| 4950 | are printed that have a colon as their first or last character. The |
| 4951 | default now is to only quote a symbol with #{...}# when the read |
| 4952 | option 'keywords' is not '#f'. Thus: |
| 4953 | |
| 4954 | guile> (define foo (string->symbol ":foo")) |
| 4955 | guile> (read-set! keywords #f) |
| 4956 | guile> foo |
| 4957 | :foo |
| 4958 | guile> (read-set! keywords 'prefix) |
| 4959 | guile> foo |
| 4960 | #{:foo}# |
| 4961 | guile> (print-set! quote-keywordish-symbols #f) |
| 4962 | guile> foo |
| 4963 | :foo |
| 4964 | |
| 4965 | ** 'while' now provides 'break' and 'continue' |
| 4966 | |
| 4967 | break and continue were previously bound in a while loop, but not |
| 4968 | documented, and continue didn't quite work properly. The undocumented |
| 4969 | parameter to break which gave a return value for the while has been |
| 4970 | dropped. |
| 4971 | |
| 4972 | ** 'call-with-current-continuation' is now also available under the name |
| 4973 | 'call/cc'. |
| 4974 | |
| 4975 | ** The module system now checks for duplicate bindings. |
| 4976 | |
| 4977 | The module system now can check for name conflicts among imported |
| 4978 | bindings. |
| 4979 | |
| 4980 | The behavior can be controlled by specifying one or more 'duplicates' |
| 4981 | handlers. For example, to make Guile return an error for every name |
| 4982 | collision, write: |
| 4983 | |
| 4984 | (define-module (foo) |
| 4985 | :use-module (bar) |
| 4986 | :use-module (baz) |
| 4987 | :duplicates check) |
| 4988 | |
| 4989 | The new default behavior of the module system when a name collision |
| 4990 | has been detected is to |
| 4991 | |
| 4992 | 1. Give priority to bindings marked as a replacement. |
| 4993 | 2. Issue a warning (different warning if overriding core binding). |
| 4994 | 3. Give priority to the last encountered binding (this corresponds to |
| 4995 | the old behavior). |
| 4996 | |
| 4997 | If you want the old behavior back without replacements or warnings you |
| 4998 | can add the line: |
| 4999 | |
| 5000 | (default-duplicate-binding-handler 'last) |
| 5001 | |
| 5002 | to your .guile init file. |
| 5003 | |
| 5004 | ** New define-module option: :replace |
| 5005 | |
| 5006 | :replace works as :export, but, in addition, marks the binding as a |
| 5007 | replacement. |
| 5008 | |
| 5009 | A typical example is `format' in (ice-9 format) which is a replacement |
| 5010 | for the core binding `format'. |
| 5011 | |
| 5012 | ** Adding prefixes to imported bindings in the module system |
| 5013 | |
| 5014 | There is now a new :use-module option :prefix. It can be used to add |
| 5015 | a prefix to all imported bindings. |
| 5016 | |
| 5017 | (define-module (foo) |
| 5018 | :use-module ((bar) :prefix bar:)) |
| 5019 | |
| 5020 | will import all bindings exported from bar, but rename them by adding |
| 5021 | the prefix `bar:'. |
| 5022 | |
| 5023 | ** Conflicting generic functions can be automatically merged. |
| 5024 | |
| 5025 | When two imported bindings conflict and they are both generic |
| 5026 | functions, the two functions can now be merged automatically. This is |
| 5027 | activated with the 'duplicates' handler 'merge-generics'. |
| 5028 | |
| 5029 | ** New function: effective-version |
| 5030 | |
| 5031 | Returns the "effective" version number. This is just the normal full |
| 5032 | version string without the final micro-version number. See "Changes |
| 5033 | to the distribution" above. |
| 5034 | |
| 5035 | ** New threading functions: parallel, letpar, par-map, and friends |
| 5036 | |
| 5037 | These are convenient ways to run calculations in parallel in new |
| 5038 | threads. See "Parallel forms" in the manual for details. |
| 5039 | |
| 5040 | ** New function 'try-mutex'. |
| 5041 | |
| 5042 | This function will attempt to lock a mutex but will return immediately |
| 5043 | instead of blocking and indicate failure. |
| 5044 | |
| 5045 | ** Waiting on a condition variable can have a timeout. |
| 5046 | |
| 5047 | The function 'wait-condition-variable' now takes a third, optional |
| 5048 | argument that specifies the point in time where the waiting should be |
| 5049 | aborted. |
| 5050 | |
| 5051 | ** New function 'broadcast-condition-variable'. |
| 5052 | |
| 5053 | ** New functions 'all-threads' and 'current-thread'. |
| 5054 | |
| 5055 | ** Signals and system asyncs work better with threads. |
| 5056 | |
| 5057 | The function 'sigaction' now takes a fourth, optional, argument that |
| 5058 | specifies the thread that the handler should run in. When the |
| 5059 | argument is omitted, the handler will run in the thread that called |
| 5060 | 'sigaction'. |
| 5061 | |
| 5062 | Likewise, 'system-async-mark' takes a second, optional, argument that |
| 5063 | specifies the thread that the async should run in. When it is |
| 5064 | omitted, the async will run in the thread that called |
| 5065 | 'system-async-mark'. |
| 5066 | |
| 5067 | C code can use the new functions scm_sigaction_for_thread and |
| 5068 | scm_system_async_mark_for_thread to pass the new thread argument. |
| 5069 | |
| 5070 | When a thread blocks on a mutex, a condition variable or is waiting |
| 5071 | for IO to be possible, it will still execute system asyncs. This can |
| 5072 | be used to interrupt such a thread by making it execute a 'throw', for |
| 5073 | example. |
| 5074 | |
| 5075 | ** The function 'system-async' is deprecated. |
| 5076 | |
| 5077 | You can now pass any zero-argument procedure to 'system-async-mark'. |
| 5078 | The function 'system-async' will just return its argument unchanged |
| 5079 | now. |
| 5080 | |
| 5081 | ** New functions 'call-with-blocked-asyncs' and |
| 5082 | 'call-with-unblocked-asyncs' |
| 5083 | |
| 5084 | The expression (call-with-blocked-asyncs PROC) will call PROC and will |
| 5085 | block execution of system asyncs for the current thread by one level |
| 5086 | while PROC runs. Likewise, call-with-unblocked-asyncs will call a |
| 5087 | procedure and will unblock the execution of system asyncs by one |
| 5088 | level for the current thread. |
| 5089 | |
| 5090 | Only system asyncs are affected by these functions. |
| 5091 | |
| 5092 | ** The functions 'mask-signals' and 'unmask-signals' are deprecated. |
| 5093 | |
| 5094 | Use 'call-with-blocked-asyncs' or 'call-with-unblocked-asyncs' |
| 5095 | instead. Those functions are easier to use correctly and can be |
| 5096 | nested. |
| 5097 | |
| 5098 | ** New function 'unsetenv'. |
| 5099 | |
| 5100 | ** New macro 'define-syntax-public'. |
| 5101 | |
| 5102 | It works like 'define-syntax' and also exports the defined macro (but |
| 5103 | only on top-level). |
| 5104 | |
| 5105 | ** There is support for Infinity and NaNs. |
| 5106 | |
| 5107 | Following PLT Scheme, Guile can now work with infinite numbers, and |
| 5108 | 'not-a-numbers'. |
| 5109 | |
| 5110 | There is new syntax for numbers: "+inf.0" (infinity), "-inf.0" |
| 5111 | (negative infinity), "+nan.0" (not-a-number), and "-nan.0" (same as |
| 5112 | "+nan.0"). These numbers are inexact and have no exact counterpart. |
| 5113 | |
| 5114 | Dividing by an inexact zero returns +inf.0 or -inf.0, depending on the |
| 5115 | sign of the dividend. The infinities are integers, and they answer #t |
| 5116 | for both 'even?' and 'odd?'. The +nan.0 value is not an integer and is |
| 5117 | not '=' to itself, but '+nan.0' is 'eqv?' to itself. |
| 5118 | |
| 5119 | For example |
| 5120 | |
| 5121 | (/ 1 0.0) |
| 5122 | => +inf.0 |
| 5123 | |
| 5124 | (/ 0 0.0) |
| 5125 | => +nan.0 |
| 5126 | |
| 5127 | (/ 0) |
| 5128 | ERROR: Numerical overflow |
| 5129 | |
| 5130 | Two new predicates 'inf?' and 'nan?' can be used to test for the |
| 5131 | special values. |
| 5132 | |
| 5133 | ** Inexact zero can have a sign. |
| 5134 | |
| 5135 | Guile can now distinguish between plus and minus inexact zero, if your |
| 5136 | platform supports this, too. The two zeros are equal according to |
| 5137 | '=', but not according to 'eqv?'. For example |
| 5138 | |
| 5139 | (- 0.0) |
| 5140 | => -0.0 |
| 5141 | |
| 5142 | (= 0.0 (- 0.0)) |
| 5143 | => #t |
| 5144 | |
| 5145 | (eqv? 0.0 (- 0.0)) |
| 5146 | => #f |
| 5147 | |
| 5148 | ** Guile now has exact rationals. |
| 5149 | |
| 5150 | Guile can now represent fractions such as 1/3 exactly. Computing with |
| 5151 | them is also done exactly, of course: |
| 5152 | |
| 5153 | (* 1/3 3/2) |
| 5154 | => 1/2 |
| 5155 | |
| 5156 | ** 'floor', 'ceiling', 'round' and 'truncate' now return exact numbers |
| 5157 | for exact arguments. |
| 5158 | |
| 5159 | For example: (floor 2) now returns an exact 2 where in the past it |
| 5160 | returned an inexact 2.0. Likewise, (floor 5/4) returns an exact 1. |
| 5161 | |
| 5162 | ** inexact->exact no longer returns only integers. |
| 5163 | |
| 5164 | Without exact rationals, the closest exact number was always an |
| 5165 | integer, but now inexact->exact returns the fraction that is exactly |
| 5166 | equal to a floating point number. For example: |
| 5167 | |
| 5168 | (inexact->exact 1.234) |
| 5169 | => 694680242521899/562949953421312 |
| 5170 | |
| 5171 | When you want the old behavior, use 'round' explicitly: |
| 5172 | |
| 5173 | (inexact->exact (round 1.234)) |
| 5174 | => 1 |
| 5175 | |
| 5176 | ** New function 'rationalize'. |
| 5177 | |
| 5178 | This function finds a simple fraction that is close to a given real |
| 5179 | number. For example (and compare with inexact->exact above): |
| 5180 | |
| 5181 | (rationalize (inexact->exact 1.234) 1/2000) |
| 5182 | => 58/47 |
| 5183 | |
| 5184 | Note that, as required by R5RS, rationalize returns only then an exact |
| 5185 | result when both its arguments are exact. |
| 5186 | |
| 5187 | ** 'odd?' and 'even?' work also for inexact integers. |
| 5188 | |
| 5189 | Previously, (odd? 1.0) would signal an error since only exact integers |
| 5190 | were recognized as integers. Now (odd? 1.0) returns #t, (odd? 2.0) |
| 5191 | returns #f and (odd? 1.5) signals an error. |
| 5192 | |
| 5193 | ** Guile now has uninterned symbols. |
| 5194 | |
| 5195 | The new function 'make-symbol' will return an uninterned symbol. This |
| 5196 | is a symbol that is unique and is guaranteed to remain unique. |
| 5197 | However, uninterned symbols can not yet be read back in. |
| 5198 | |
| 5199 | Use the new function 'symbol-interned?' to check whether a symbol is |
| 5200 | interned or not. |
| 5201 | |
| 5202 | ** pretty-print has more options. |
| 5203 | |
| 5204 | The function pretty-print from the (ice-9 pretty-print) module can now |
| 5205 | also be invoked with keyword arguments that control things like |
| 5206 | maximum output width. See the manual for details. |
| 5207 | |
| 5208 | ** Variables have no longer a special behavior for `equal?'. |
| 5209 | |
| 5210 | Previously, comparing two variables with `equal?' would recursivly |
| 5211 | compare their values. This is no longer done. Variables are now only |
| 5212 | `equal?' if they are `eq?'. |
| 5213 | |
| 5214 | ** `(begin)' is now valid. |
| 5215 | |
| 5216 | You can now use an empty `begin' form. It will yield #<unspecified> |
| 5217 | when evaluated and simply be ignored in a definition context. |
| 5218 | |
| 5219 | ** Deprecated: procedure->macro |
| 5220 | |
| 5221 | Change your code to use 'define-macro' or r5rs macros. Also, be aware |
| 5222 | that macro expansion will not be done during evaluation, but prior to |
| 5223 | evaluation. |
| 5224 | |
| 5225 | ** Soft ports now allow a `char-ready?' procedure |
| 5226 | |
| 5227 | The vector argument to `make-soft-port' can now have a length of |
| 5228 | either 5 or 6. (Previously the length had to be 5.) The optional 6th |
| 5229 | element is interpreted as an `input-waiting' thunk -- i.e. a thunk |
| 5230 | that returns the number of characters that can be read immediately |
| 5231 | without the soft port blocking. |
| 5232 | |
| 5233 | ** Deprecated: undefine |
| 5234 | |
| 5235 | There is no replacement for undefine. |
| 5236 | |
| 5237 | ** The functions make-keyword-from-dash-symbol and keyword-dash-symbol |
| 5238 | have been discouraged. |
| 5239 | |
| 5240 | They are relics from a time where a keyword like #:foo was used |
| 5241 | directly as a Tcl option "-foo" and thus keywords were internally |
| 5242 | stored as a symbol with a starting dash. We now store a symbol |
| 5243 | without the dash. |
| 5244 | |
| 5245 | Use symbol->keyword and keyword->symbol instead. |
| 5246 | |
| 5247 | ** The `cheap' debug option is now obsolete |
| 5248 | |
| 5249 | Evaluator trap calls are now unconditionally "cheap" - in other words, |
| 5250 | they pass a debug object to the trap handler rather than a full |
| 5251 | continuation. The trap handler code can capture a full continuation |
| 5252 | by using `call-with-current-continuation' in the usual way, if it so |
| 5253 | desires. |
| 5254 | |
| 5255 | The `cheap' option is retained for now so as not to break existing |
| 5256 | code which gets or sets it, but setting it now has no effect. It will |
| 5257 | be removed in the next major Guile release. |
| 5258 | |
| 5259 | ** Evaluator trap calls now support `tweaking' |
| 5260 | |
| 5261 | `Tweaking' means that the trap handler code can modify the Scheme |
| 5262 | expression that is about to be evaluated (in the case of an |
| 5263 | enter-frame trap) or the value that is being returned (in the case of |
| 5264 | an exit-frame trap). The trap handler code indicates that it wants to |
| 5265 | do this by returning a pair whose car is the symbol 'instead and whose |
| 5266 | cdr is the modified expression or return value. |
| 5267 | |
| 5268 | * Changes to the C interface |
| 5269 | |
| 5270 | ** The functions scm_hash_fn_remove_x and scm_hashx_remove_x no longer |
| 5271 | take a 'delete' function argument. |
| 5272 | |
| 5273 | This argument makes no sense since the delete function is used to |
| 5274 | remove a pair from an alist, and this must not be configurable. |
| 5275 | |
| 5276 | This is an incompatible change. |
| 5277 | |
| 5278 | ** The GH interface is now subject to the deprecation mechanism |
| 5279 | |
| 5280 | The GH interface has been deprecated for quite some time but now it is |
| 5281 | actually removed from Guile when it is configured with |
| 5282 | --disable-deprecated. |
| 5283 | |
| 5284 | See the manual "Transitioning away from GH" for more information. |
| 5285 | |
| 5286 | ** A new family of functions for converting between C values and |
| 5287 | Scheme values has been added. |
| 5288 | |
| 5289 | These functions follow a common naming scheme and are designed to be |
| 5290 | easier to use, thread-safe and more future-proof than the older |
| 5291 | alternatives. |
| 5292 | |
| 5293 | - int scm_is_* (...) |
| 5294 | |
| 5295 | These are predicates that return a C boolean: 1 or 0. Instead of |
| 5296 | SCM_NFALSEP, you can now use scm_is_true, for example. |
| 5297 | |
| 5298 | - <type> scm_to_<type> (SCM val, ...) |
| 5299 | |
| 5300 | These are functions that convert a Scheme value into an appropriate |
| 5301 | C value. For example, you can use scm_to_int to safely convert from |
| 5302 | a SCM to an int. |
| 5303 | |
| 5304 | - SCM scm_from_<type> (<type> val, ...) |
| 5305 | |
| 5306 | These functions convert from a C type to a SCM value; for example, |
| 5307 | scm_from_int for ints. |
| 5308 | |
| 5309 | There is a huge number of these functions, for numbers, strings, |
| 5310 | symbols, vectors, etc. They are documented in the reference manual in |
| 5311 | the API section together with the types that they apply to. |
| 5312 | |
| 5313 | ** New functions for dealing with complex numbers in C have been added. |
| 5314 | |
| 5315 | The new functions are scm_c_make_rectangular, scm_c_make_polar, |
| 5316 | scm_c_real_part, scm_c_imag_part, scm_c_magnitude and scm_c_angle. |
| 5317 | They work like scm_make_rectangular etc but take or return doubles |
| 5318 | directly. |
| 5319 | |
| 5320 | ** The function scm_make_complex has been discouraged. |
| 5321 | |
| 5322 | Use scm_c_make_rectangular instead. |
| 5323 | |
| 5324 | ** The INUM macros have been deprecated. |
| 5325 | |
| 5326 | A lot of code uses these macros to do general integer conversions, |
| 5327 | although the macros only work correctly with fixnums. Use the |
| 5328 | following alternatives. |
| 5329 | |
| 5330 | SCM_INUMP -> scm_is_integer or similar |
| 5331 | SCM_NINUMP -> !scm_is_integer or similar |
| 5332 | SCM_MAKINUM -> scm_from_int or similar |
| 5333 | SCM_INUM -> scm_to_int or similar |
| 5334 | |
| 5335 | SCM_VALIDATE_INUM_* -> Do not use these; scm_to_int, etc. will |
| 5336 | do the validating for you. |
| 5337 | |
| 5338 | ** The scm_num2<type> and scm_<type>2num functions and scm_make_real |
| 5339 | have been discouraged. |
| 5340 | |
| 5341 | Use the newer scm_to_<type> and scm_from_<type> functions instead for |
| 5342 | new code. The functions have been discouraged since they don't fit |
| 5343 | the naming scheme. |
| 5344 | |
| 5345 | ** The 'boolean' macros SCM_FALSEP etc have been discouraged. |
| 5346 | |
| 5347 | They have strange names, especially SCM_NFALSEP, and SCM_BOOLP |
| 5348 | evaluates its argument twice. Use scm_is_true, etc. instead for new |
| 5349 | code. |
| 5350 | |
| 5351 | ** The macro SCM_EQ_P has been discouraged. |
| 5352 | |
| 5353 | Use scm_is_eq for new code, which fits better into the naming |
| 5354 | conventions. |
| 5355 | |
| 5356 | ** The macros SCM_CONSP, SCM_NCONSP, SCM_NULLP, and SCM_NNULLP have |
| 5357 | been discouraged. |
| 5358 | |
| 5359 | Use the function scm_is_pair or scm_is_null instead. |
| 5360 | |
| 5361 | ** The functions scm_round and scm_truncate have been deprecated and |
| 5362 | are now available as scm_c_round and scm_c_truncate, respectively. |
| 5363 | |
| 5364 | These functions occupy the names that scm_round_number and |
| 5365 | scm_truncate_number should have. |
| 5366 | |
| 5367 | ** The functions scm_c_string2str, scm_c_substring2str, and |
| 5368 | scm_c_symbol2str have been deprecated. |
| 5369 | |
| 5370 | Use scm_to_locale_stringbuf or similar instead, maybe together with |
| 5371 | scm_substring. |
| 5372 | |
| 5373 | ** New functions scm_c_make_string, scm_c_string_length, |
| 5374 | scm_c_string_ref, scm_c_string_set_x, scm_c_substring, |
| 5375 | scm_c_substring_shared, scm_c_substring_copy. |
| 5376 | |
| 5377 | These are like scm_make_string, scm_length, etc. but are slightly |
| 5378 | easier to use from C. |
| 5379 | |
| 5380 | ** The macros SCM_STRINGP, SCM_STRING_CHARS, SCM_STRING_LENGTH, |
| 5381 | SCM_SYMBOL_CHARS, and SCM_SYMBOL_LENGTH have been deprecated. |
| 5382 | |
| 5383 | They export too many assumptions about the implementation of strings |
| 5384 | and symbols that are no longer true in the presence of |
| 5385 | mutation-sharing substrings and when Guile switches to some form of |
| 5386 | Unicode. |
| 5387 | |
| 5388 | When working with strings, it is often best to use the normal string |
| 5389 | functions provided by Guile, such as scm_c_string_ref, |
| 5390 | scm_c_string_set_x, scm_string_append, etc. Be sure to look in the |
| 5391 | manual since many more such functions are now provided than |
| 5392 | previously. |
| 5393 | |
| 5394 | When you want to convert a SCM string to a C string, use the |
| 5395 | scm_to_locale_string function or similar instead. For symbols, use |
| 5396 | scm_symbol_to_string and then work with that string. Because of the |
| 5397 | new string representation, scm_symbol_to_string does not need to copy |
| 5398 | and is thus quite efficient. |
| 5399 | |
| 5400 | ** Some string, symbol and keyword functions have been discouraged. |
| 5401 | |
| 5402 | They don't fit into the uniform naming scheme and are not explicit |
| 5403 | about the character encoding. |
| 5404 | |
| 5405 | Replace according to the following table: |
| 5406 | |
| 5407 | scm_allocate_string -> scm_c_make_string |
| 5408 | scm_take_str -> scm_take_locale_stringn |
| 5409 | scm_take0str -> scm_take_locale_string |
| 5410 | scm_mem2string -> scm_from_locale_stringn |
| 5411 | scm_str2string -> scm_from_locale_string |
| 5412 | scm_makfrom0str -> scm_from_locale_string |
| 5413 | scm_mem2symbol -> scm_from_locale_symboln |
| 5414 | scm_mem2uninterned_symbol -> scm_from_locale_stringn + scm_make_symbol |
| 5415 | scm_str2symbol -> scm_from_locale_symbol |
| 5416 | |
| 5417 | SCM_SYMBOL_HASH -> scm_hashq |
| 5418 | SCM_SYMBOL_INTERNED_P -> scm_symbol_interned_p |
| 5419 | |
| 5420 | scm_c_make_keyword -> scm_from_locale_keyword |
| 5421 | |
| 5422 | ** The functions scm_keyword_to_symbol and sym_symbol_to_keyword are |
| 5423 | now also available to C code. |
| 5424 | |
| 5425 | ** SCM_KEYWORDP and SCM_KEYWORDSYM have been deprecated. |
| 5426 | |
| 5427 | Use scm_is_keyword and scm_keyword_to_symbol instead, but note that |
| 5428 | the latter returns the true name of the keyword, not the 'dash name', |
| 5429 | as SCM_KEYWORDSYM used to do. |
| 5430 | |
| 5431 | ** A new way to access arrays in a thread-safe and efficient way has |
| 5432 | been added. |
| 5433 | |
| 5434 | See the manual, node "Accessing Arrays From C". |
| 5435 | |
| 5436 | ** The old uniform vector and bitvector implementations have been |
| 5437 | unceremoniously removed. |
| 5438 | |
| 5439 | This implementation exposed the details of the tagging system of |
| 5440 | Guile. Use the new C API explained in the manual in node "Uniform |
| 5441 | Numeric Vectors" and "Bit Vectors", respectively. |
| 5442 | |
| 5443 | The following macros are gone: SCM_UVECTOR_BASE, SCM_SET_UVECTOR_BASE, |
| 5444 | SCM_UVECTOR_MAXLENGTH, SCM_UVECTOR_LENGTH, SCM_MAKE_UVECTOR_TAG, |
| 5445 | SCM_SET_UVECTOR_LENGTH, SCM_BITVECTOR_P, SCM_BITVECTOR_BASE, |
| 5446 | SCM_SET_BITVECTOR_BASE, SCM_BITVECTOR_MAX_LENGTH, |
| 5447 | SCM_BITVECTOR_LENGTH, SCM_MAKE_BITVECTOR_TAG, |
| 5448 | SCM_SET_BITVECTOR_LENGTH, SCM_BITVEC_REF, SCM_BITVEC_SET, |
| 5449 | SCM_BITVEC_CLR. |
| 5450 | |
| 5451 | ** The macros dealing with vectors have been deprecated. |
| 5452 | |
| 5453 | Use the new functions scm_is_vector, scm_vector_elements, |
| 5454 | scm_vector_writable_elements, etc, or scm_is_simple_vector, |
| 5455 | SCM_SIMPLE_VECTOR_REF, SCM_SIMPLE_VECTOR_SET, etc instead. See the |
| 5456 | manual for more details. |
| 5457 | |
| 5458 | Deprecated are SCM_VECTORP, SCM_VELTS, SCM_VECTOR_MAX_LENGTH, |
| 5459 | SCM_VECTOR_LENGTH, SCM_VECTOR_REF, SCM_VECTOR_SET, SCM_WRITABLE_VELTS. |
| 5460 | |
| 5461 | The following macros have been removed: SCM_VECTOR_BASE, |
| 5462 | SCM_SET_VECTOR_BASE, SCM_MAKE_VECTOR_TAG, SCM_SET_VECTOR_LENGTH, |
| 5463 | SCM_VELTS_AS_STACKITEMS, SCM_SETVELTS, SCM_GC_WRITABLE_VELTS. |
| 5464 | |
| 5465 | ** Some C functions and macros related to arrays have been deprecated. |
| 5466 | |
| 5467 | Migrate according to the following table: |
| 5468 | |
| 5469 | scm_make_uve -> scm_make_typed_array, scm_make_u8vector etc. |
| 5470 | scm_make_ra -> scm_make_array |
| 5471 | scm_shap2ra -> scm_make_array |
| 5472 | scm_cvref -> scm_c_generalized_vector_ref |
| 5473 | scm_ra_set_contp -> do not use |
| 5474 | scm_aind -> scm_array_handle_pos |
| 5475 | scm_raprin1 -> scm_display or scm_write |
| 5476 | |
| 5477 | SCM_ARRAYP -> scm_is_array |
| 5478 | SCM_ARRAY_NDIM -> scm_c_array_rank |
| 5479 | SCM_ARRAY_DIMS -> scm_array_handle_dims |
| 5480 | SCM_ARRAY_CONTP -> do not use |
| 5481 | SCM_ARRAY_MEM -> do not use |
| 5482 | SCM_ARRAY_V -> scm_array_handle_elements or similar |
| 5483 | SCM_ARRAY_BASE -> do not use |
| 5484 | |
| 5485 | ** SCM_CELL_WORD_LOC has been deprecated. |
| 5486 | |
| 5487 | Use the new macro SCM_CELL_OBJECT_LOC instead, which returns a pointer |
| 5488 | to a SCM, as opposed to a pointer to a scm_t_bits. |
| 5489 | |
| 5490 | This was done to allow the correct use of pointers into the Scheme |
| 5491 | heap. Previously, the heap words were of type scm_t_bits and local |
| 5492 | variables and function arguments were of type SCM, making it |
| 5493 | non-standards-conformant to have a pointer that can point to both. |
| 5494 | |
| 5495 | ** New macros SCM_SMOB_DATA_2, SCM_SMOB_DATA_3, etc. |
| 5496 | |
| 5497 | These macros should be used instead of SCM_CELL_WORD_2/3 to access the |
| 5498 | second and third words of double smobs. Likewise for |
| 5499 | SCM_SET_SMOB_DATA_2 and SCM_SET_SMOB_DATA_3. |
| 5500 | |
| 5501 | Also, there is SCM_SMOB_FLAGS and SCM_SET_SMOB_FLAGS that should be |
| 5502 | used to get and set the 16 exra bits in the zeroth word of a smob. |
| 5503 | |
| 5504 | And finally, there is SCM_SMOB_OBJECT and SCM_SMOB_SET_OBJECT for |
| 5505 | accesing the first immediate word of a smob as a SCM value, and there |
| 5506 | is SCM_SMOB_OBJECT_LOC for getting a pointer to the first immediate |
| 5507 | smob word. Like wise for SCM_SMOB_OBJECT_2, etc. |
| 5508 | |
| 5509 | ** New way to deal with non-local exits and re-entries. |
| 5510 | |
| 5511 | There is a new set of functions that essentially do what |
| 5512 | scm_internal_dynamic_wind does, but in a way that is more convenient |
| 5513 | for C code in some situations. Here is a quick example of how to |
| 5514 | prevent a potential memory leak: |
| 5515 | |
| 5516 | void |
| 5517 | foo () |
| 5518 | { |
| 5519 | char *mem; |
| 5520 | |
| 5521 | scm_dynwind_begin (0); |
| 5522 | |
| 5523 | mem = scm_malloc (100); |
| 5524 | scm_dynwind_unwind_handler (free, mem, SCM_F_WIND_EXPLICITLY); |
| 5525 | |
| 5526 | /* MEM would leak if BAR throws an error. |
| 5527 | SCM_DYNWIND_UNWIND_HANDLER frees it nevertheless. |
| 5528 | */ |
| 5529 | |
| 5530 | bar (); |
| 5531 | |
| 5532 | scm_dynwind_end (); |
| 5533 | |
| 5534 | /* Because of SCM_F_WIND_EXPLICITLY, MEM will be freed by |
| 5535 | SCM_DYNWIND_END as well. |
| 5536 | */ |
| 5537 | } |
| 5538 | |
| 5539 | For full documentation, see the node "Dynamic Wind" in the manual. |
| 5540 | |
| 5541 | ** New function scm_dynwind_free |
| 5542 | |
| 5543 | This function calls 'free' on a given pointer when a dynwind context |
| 5544 | is left. Thus the call to scm_dynwind_unwind_handler above could be |
| 5545 | replaced with simply scm_dynwind_free (mem). |
| 5546 | |
| 5547 | ** New functions scm_c_call_with_blocked_asyncs and |
| 5548 | scm_c_call_with_unblocked_asyncs |
| 5549 | |
| 5550 | Like scm_call_with_blocked_asyncs etc. but for C functions. |
| 5551 | |
| 5552 | ** New functions scm_dynwind_block_asyncs and scm_dynwind_unblock_asyncs |
| 5553 | |
| 5554 | In addition to scm_c_call_with_blocked_asyncs you can now also use |
| 5555 | scm_dynwind_block_asyncs in a 'dynwind context' (see above). Likewise for |
| 5556 | scm_c_call_with_unblocked_asyncs and scm_dynwind_unblock_asyncs. |
| 5557 | |
| 5558 | ** The macros SCM_DEFER_INTS, SCM_ALLOW_INTS, SCM_REDEFER_INTS, |
| 5559 | SCM_REALLOW_INTS have been deprecated. |
| 5560 | |
| 5561 | They do no longer fulfill their original role of blocking signal |
| 5562 | delivery. Depending on what you want to achieve, replace a pair of |
| 5563 | SCM_DEFER_INTS and SCM_ALLOW_INTS with a dynwind context that locks a |
| 5564 | mutex, blocks asyncs, or both. See node "Critical Sections" in the |
| 5565 | manual. |
| 5566 | |
| 5567 | ** The value 'scm_mask_ints' is no longer writable. |
| 5568 | |
| 5569 | Previously, you could set scm_mask_ints directly. This is no longer |
| 5570 | possible. Use scm_c_call_with_blocked_asyncs and |
| 5571 | scm_c_call_with_unblocked_asyncs instead. |
| 5572 | |
| 5573 | ** New way to temporarily set the current input, output or error ports |
| 5574 | |
| 5575 | C code can now use scm_dynwind_current_<foo>_port in a 'dynwind |
| 5576 | context' (see above). <foo> is one of "input", "output" or "error". |
| 5577 | |
| 5578 | ** New way to temporarily set fluids |
| 5579 | |
| 5580 | C code can now use scm_dynwind_fluid in a 'dynwind context' (see |
| 5581 | above) to temporarily set the value of a fluid. |
| 5582 | |
| 5583 | ** New types scm_t_intmax and scm_t_uintmax. |
| 5584 | |
| 5585 | On platforms that have them, these types are identical to intmax_t and |
| 5586 | uintmax_t, respectively. On other platforms, they are identical to |
| 5587 | the largest integer types that Guile knows about. |
| 5588 | |
| 5589 | ** The functions scm_unmemocopy and scm_unmemoize have been removed. |
| 5590 | |
| 5591 | You should not have used them. |
| 5592 | |
| 5593 | ** Many public #defines with generic names have been made private. |
| 5594 | |
| 5595 | #defines with generic names like HAVE_FOO or SIZEOF_FOO have been made |
| 5596 | private or renamed with a more suitable public name. |
| 5597 | |
| 5598 | ** The macro SCM_TYP16S has been deprecated. |
| 5599 | |
| 5600 | This macro is not intended for public use. |
| 5601 | |
| 5602 | ** The macro SCM_SLOPPY_INEXACTP has been deprecated. |
| 5603 | |
| 5604 | Use scm_is_true (scm_inexact_p (...)) instead. |
| 5605 | |
| 5606 | ** The macro SCM_SLOPPY_REALP has been deprecated. |
| 5607 | |
| 5608 | Use scm_is_real instead. |
| 5609 | |
| 5610 | ** The macro SCM_SLOPPY_COMPLEXP has been deprecated. |
| 5611 | |
| 5612 | Use scm_is_complex instead. |
| 5613 | |
| 5614 | ** Some preprocessor defines have been deprecated. |
| 5615 | |
| 5616 | These defines indicated whether a certain feature was present in Guile |
| 5617 | or not. Going forward, assume that the features are always present. |
| 5618 | |
| 5619 | The macros are: USE_THREADS, GUILE_ISELECT, READER_EXTENSIONS, |
| 5620 | DEBUG_EXTENSIONS, DYNAMIC_LINKING. |
| 5621 | |
| 5622 | The following macros have been removed completely: MEMOIZE_LOCALS, |
| 5623 | SCM_RECKLESS, SCM_CAUTIOUS. |
| 5624 | |
| 5625 | ** The preprocessor define STACK_DIRECTION has been deprecated. |
| 5626 | |
| 5627 | There should be no need to know about the stack direction for ordinary |
| 5628 | programs. |
| 5629 | |
| 5630 | ** New function: scm_effective_version |
| 5631 | |
| 5632 | Returns the "effective" version number. This is just the normal full |
| 5633 | version string without the final micro-version number. See "Changes |
| 5634 | to the distribution" above. |
| 5635 | |
| 5636 | ** The function scm_call_with_new_thread has a new prototype. |
| 5637 | |
| 5638 | Instead of taking a list with the thunk and handler, these two |
| 5639 | arguments are now passed directly: |
| 5640 | |
| 5641 | SCM scm_call_with_new_thread (SCM thunk, SCM handler); |
| 5642 | |
| 5643 | This is an incompatible change. |
| 5644 | |
| 5645 | ** New snarfer macro SCM_DEFINE_PUBLIC. |
| 5646 | |
| 5647 | This is like SCM_DEFINE, but also calls scm_c_export for the defined |
| 5648 | function in the init section. |
| 5649 | |
| 5650 | ** The snarfer macro SCM_SNARF_INIT is now officially supported. |
| 5651 | |
| 5652 | ** Garbage collector rewrite. |
| 5653 | |
| 5654 | The garbage collector is cleaned up a lot, and now uses lazy |
| 5655 | sweeping. This is reflected in the output of (gc-stats); since cells |
| 5656 | are being freed when they are allocated, the cells-allocated field |
| 5657 | stays roughly constant. |
| 5658 | |
| 5659 | For malloc related triggers, the behavior is changed. It uses the same |
| 5660 | heuristic as the cell-triggered collections. It may be tuned with the |
| 5661 | environment variables GUILE_MIN_YIELD_MALLOC. This is the percentage |
| 5662 | for minimum yield of malloc related triggers. The default is 40. |
| 5663 | GUILE_INIT_MALLOC_LIMIT sets the initial trigger for doing a GC. The |
| 5664 | default is 200 kb. |
| 5665 | |
| 5666 | Debugging operations for the freelist have been deprecated, along with |
| 5667 | the C variables that control garbage collection. The environment |
| 5668 | variables GUILE_MAX_SEGMENT_SIZE, GUILE_INIT_SEGMENT_SIZE_2, |
| 5669 | GUILE_INIT_SEGMENT_SIZE_1, and GUILE_MIN_YIELD_2 should be used. |
| 5670 | |
| 5671 | For understanding the memory usage of a GUILE program, the routine |
| 5672 | gc-live-object-stats returns an alist containing the number of live |
| 5673 | objects for every type. |
| 5674 | |
| 5675 | |
| 5676 | ** The function scm_definedp has been renamed to scm_defined_p |
| 5677 | |
| 5678 | The name scm_definedp is deprecated. |
| 5679 | |
| 5680 | ** The struct scm_cell type has been renamed to scm_t_cell |
| 5681 | |
| 5682 | This is in accordance to Guile's naming scheme for types. Note that |
| 5683 | the name scm_cell is now used for a function that allocates and |
| 5684 | initializes a new cell (see below). |
| 5685 | |
| 5686 | ** New functions for memory management |
| 5687 | |
| 5688 | A new set of functions for memory management has been added since the |
| 5689 | old way (scm_must_malloc, scm_must_free, etc) was error prone and |
| 5690 | indeed, Guile itself contained some long standing bugs that could |
| 5691 | cause aborts in long running programs. |
| 5692 | |
| 5693 | The new functions are more symmetrical and do not need cooperation |
| 5694 | from smob free routines, among other improvements. |
| 5695 | |
| 5696 | The new functions are scm_malloc, scm_realloc, scm_calloc, scm_strdup, |
| 5697 | scm_strndup, scm_gc_malloc, scm_gc_calloc, scm_gc_realloc, |
| 5698 | scm_gc_free, scm_gc_register_collectable_memory, and |
| 5699 | scm_gc_unregister_collectable_memory. Refer to the manual for more |
| 5700 | details and for upgrading instructions. |
| 5701 | |
| 5702 | The old functions for memory management have been deprecated. They |
| 5703 | are: scm_must_malloc, scm_must_realloc, scm_must_free, |
| 5704 | scm_must_strdup, scm_must_strndup, scm_done_malloc, scm_done_free. |
| 5705 | |
| 5706 | ** Declarations of exported features are marked with SCM_API. |
| 5707 | |
| 5708 | Every declaration of a feature that belongs to the exported Guile API |
| 5709 | has been marked by adding the macro "SCM_API" to the start of the |
| 5710 | declaration. This macro can expand into different things, the most |
| 5711 | common of which is just "extern" for Unix platforms. On Win32, it can |
| 5712 | be used to control which symbols are exported from a DLL. |
| 5713 | |
| 5714 | If you `#define SCM_IMPORT' before including <libguile.h>, SCM_API |
| 5715 | will expand into "__declspec (dllimport) extern", which is needed for |
| 5716 | linking to the Guile DLL in Windows. |
| 5717 | |
| 5718 | There are also SCM_RL_IMPORT, SCM_SRFI1314_IMPORT, and |
| 5719 | SCM_SRFI4_IMPORT, for the corresponding libraries. |
| 5720 | |
| 5721 | ** SCM_NEWCELL and SCM_NEWCELL2 have been deprecated. |
| 5722 | |
| 5723 | Use the new functions scm_cell and scm_double_cell instead. The old |
| 5724 | macros had problems because with them allocation and initialization |
| 5725 | was separated and the GC could sometimes observe half initialized |
| 5726 | cells. Only careful coding by the user of SCM_NEWCELL and |
| 5727 | SCM_NEWCELL2 could make this safe and efficient. |
| 5728 | |
| 5729 | ** CHECK_ENTRY, CHECK_APPLY and CHECK_EXIT have been deprecated. |
| 5730 | |
| 5731 | Use the variables scm_check_entry_p, scm_check_apply_p and scm_check_exit_p |
| 5732 | instead. |
| 5733 | |
| 5734 | ** SRCBRKP has been deprecated. |
| 5735 | |
| 5736 | Use scm_c_source_property_breakpoint_p instead. |
| 5737 | |
| 5738 | ** Deprecated: scm_makmacro |
| 5739 | |
| 5740 | Change your code to use either scm_makmmacro or to define macros in |
| 5741 | Scheme, using 'define-macro'. |
| 5742 | |
| 5743 | ** New function scm_c_port_for_each. |
| 5744 | |
| 5745 | This function is like scm_port_for_each but takes a pointer to a C |
| 5746 | function as the callback instead of a SCM value. |
| 5747 | |
| 5748 | ** The names scm_internal_select, scm_thread_sleep, and |
| 5749 | scm_thread_usleep have been discouraged. |
| 5750 | |
| 5751 | Use scm_std_select, scm_std_sleep, scm_std_usleep instead. |
| 5752 | |
| 5753 | ** The GC can no longer be blocked. |
| 5754 | |
| 5755 | The global flags scm_gc_heap_lock and scm_block_gc have been removed. |
| 5756 | The GC can now run (partially) concurrently with other code and thus |
| 5757 | blocking it is not well defined. |
| 5758 | |
| 5759 | ** Many definitions have been removed that were previously deprecated. |
| 5760 | |
| 5761 | scm_lisp_nil, scm_lisp_t, s_nil_ify, scm_m_nil_ify, s_t_ify, |
| 5762 | scm_m_t_ify, s_0_cond, scm_m_0_cond, s_0_ify, scm_m_0_ify, s_1_ify, |
| 5763 | scm_m_1_ify, scm_debug_newcell, scm_debug_newcell2, |
| 5764 | scm_tc16_allocated, SCM_SET_SYMBOL_HASH, SCM_IM_NIL_IFY, SCM_IM_T_IFY, |
| 5765 | SCM_IM_0_COND, SCM_IM_0_IFY, SCM_IM_1_IFY, SCM_GC_SET_ALLOCATED, |
| 5766 | scm_debug_newcell, scm_debug_newcell2, SCM_HUP_SIGNAL, SCM_INT_SIGNAL, |
| 5767 | SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, SCM_ALRM_SIGNAL, |
| 5768 | SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, SCM_ORD_SIG, |
| 5769 | SCM_NUM_SIGS, scm_top_level_lookup_closure_var, |
| 5770 | *top-level-lookup-closure*, scm_system_transformer, scm_eval_3, |
| 5771 | scm_eval2, root_module_lookup_closure, SCM_SLOPPY_STRINGP, |
| 5772 | SCM_RWSTRINGP, scm_read_only_string_p, scm_make_shared_substring, |
| 5773 | scm_tc7_substring, sym_huh, SCM_VARVCELL, SCM_UDVARIABLEP, |
| 5774 | SCM_DEFVARIABLEP, scm_mkbig, scm_big2inum, scm_adjbig, scm_normbig, |
| 5775 | scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl, SCM_FIXNUM_BIT, |
| 5776 | SCM_SETCHARS, SCM_SLOPPY_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, |
| 5777 | SCM_LENGTH_MAX, SCM_SETLENGTH, SCM_ROSTRINGP, SCM_ROLENGTH, |
| 5778 | SCM_ROCHARS, SCM_ROUCHARS, SCM_SUBSTRP, SCM_COERCE_SUBSTR, |
| 5779 | scm_sym2vcell, scm_intern, scm_intern0, scm_sysintern, scm_sysintern0, |
| 5780 | scm_sysintern0_no_module_lookup, scm_init_symbols_deprecated, |
| 5781 | scm_vector_set_length_x, scm_contregs, scm_debug_info, |
| 5782 | scm_debug_frame, SCM_DSIDEVAL, SCM_CONST_LONG, SCM_VCELL, |
| 5783 | SCM_GLOBAL_VCELL, SCM_VCELL_INIT, SCM_GLOBAL_VCELL_INIT, |
| 5784 | SCM_HUGE_LENGTH, SCM_VALIDATE_STRINGORSUBSTR, SCM_VALIDATE_ROSTRING, |
| 5785 | SCM_VALIDATE_ROSTRING_COPY, SCM_VALIDATE_NULLORROSTRING_COPY, |
| 5786 | SCM_VALIDATE_RWSTRING, DIGITS, scm_small_istr2int, scm_istr2int, |
| 5787 | scm_istr2flo, scm_istring2number, scm_istr2int, scm_istr2flo, |
| 5788 | scm_istring2number, scm_vtable_index_vcell, scm_si_vcell, SCM_ECONSP, |
| 5789 | SCM_NECONSP, SCM_GLOC_VAR, SCM_GLOC_VAL, SCM_GLOC_SET_VAL, |
| 5790 | SCM_GLOC_VAL_LOC, scm_make_gloc, scm_gloc_p, scm_tc16_variable, |
| 5791 | SCM_CHARS, SCM_LENGTH, SCM_SET_STRING_CHARS, SCM_SET_STRING_LENGTH. |
| 5792 | |
| 5793 | * Changes to bundled modules |
| 5794 | |
| 5795 | ** (ice-9 debug) |
| 5796 | |
| 5797 | Using the (ice-9 debug) module no longer automatically switches Guile |
| 5798 | to use the debugging evaluator. If you want to switch to the |
| 5799 | debugging evaluator (which is needed for backtrace information if you |
| 5800 | hit an error), please add an explicit "(debug-enable 'debug)" to your |
| 5801 | code just after the code to use (ice-9 debug). |
| 5802 | |
| 5803 | \f |
| 5804 | Changes since Guile 1.4: |
| 5805 | |
| 5806 | * Changes to the distribution |
| 5807 | |
| 5808 | ** A top-level TODO file is included. |
| 5809 | |
| 5810 | ** Guile now uses a versioning scheme similar to that of the Linux kernel. |
| 5811 | |
| 5812 | Guile now always uses three numbers to represent the version, |
| 5813 | i.e. "1.6.5". The first number, 1, is the major version number, the |
| 5814 | second number, 6, is the minor version number, and the third number, |
| 5815 | 5, is the micro version number. Changes in major version number |
| 5816 | indicate major changes in Guile. |
| 5817 | |
| 5818 | Minor version numbers that are even denote stable releases, and odd |
| 5819 | minor version numbers denote development versions (which may be |
| 5820 | unstable). The micro version number indicates a minor sub-revision of |
| 5821 | a given MAJOR.MINOR release. |
| 5822 | |
| 5823 | In keeping with the new scheme, (minor-version) and scm_minor_version |
| 5824 | no longer return everything but the major version number. They now |
| 5825 | just return the minor version number. Two new functions |
| 5826 | (micro-version) and scm_micro_version have been added to report the |
| 5827 | micro version number. |
| 5828 | |
| 5829 | In addition, ./GUILE-VERSION now defines GUILE_MICRO_VERSION. |
| 5830 | |
| 5831 | ** New preprocessor definitions are available for checking versions. |
| 5832 | |
| 5833 | version.h now #defines SCM_MAJOR_VERSION, SCM_MINOR_VERSION, and |
| 5834 | SCM_MICRO_VERSION to the appropriate integer values. |
| 5835 | |
| 5836 | ** Guile now actively warns about deprecated features. |
| 5837 | |
| 5838 | The new configure option `--enable-deprecated=LEVEL' and the |
| 5839 | environment variable GUILE_WARN_DEPRECATED control this mechanism. |
| 5840 | See INSTALL and README for more information. |
| 5841 | |
| 5842 | ** Guile is much more likely to work on 64-bit architectures. |
| 5843 | |
| 5844 | Guile now compiles and passes "make check" with only two UNRESOLVED GC |
| 5845 | cases on Alpha and ia64 based machines now. Thanks to John Goerzen |
| 5846 | for the use of a test machine, and thanks to Stefan Jahn for ia64 |
| 5847 | patches. |
| 5848 | |
| 5849 | ** New functions: setitimer and getitimer. |
| 5850 | |
| 5851 | These implement a fairly direct interface to the libc functions of the |
| 5852 | same name. |
| 5853 | |
| 5854 | ** The #. reader extension is now disabled by default. |
| 5855 | |
| 5856 | For safety reasons, #. evaluation is disabled by default. To |
| 5857 | re-enable it, set the fluid read-eval? to #t. For example: |
| 5858 | |
| 5859 | (fluid-set! read-eval? #t) |
| 5860 | |
| 5861 | but make sure you realize the potential security risks involved. With |
| 5862 | read-eval? enabled, reading a data file from an untrusted source can |
| 5863 | be dangerous. |
| 5864 | |
| 5865 | ** New SRFI modules have been added: |
| 5866 | |
| 5867 | SRFI-0 `cond-expand' is now supported in Guile, without requiring |
| 5868 | using a module. |
| 5869 | |
| 5870 | (srfi srfi-1) is a library containing many useful pair- and list-processing |
| 5871 | procedures. |
| 5872 | |
| 5873 | (srfi srfi-2) exports and-let*. |
| 5874 | |
| 5875 | (srfi srfi-4) implements homogeneous numeric vector datatypes. |
| 5876 | |
| 5877 | (srfi srfi-6) is a dummy module for now, since guile already provides |
| 5878 | all of the srfi-6 procedures by default: open-input-string, |
| 5879 | open-output-string, get-output-string. |
| 5880 | |
| 5881 | (srfi srfi-8) exports receive. |
| 5882 | |
| 5883 | (srfi srfi-9) exports define-record-type. |
| 5884 | |
| 5885 | (srfi srfi-10) exports define-reader-ctor and implements the reader |
| 5886 | extension #,(). |
| 5887 | |
| 5888 | (srfi srfi-11) exports let-values and let*-values. |
| 5889 | |
| 5890 | (srfi srfi-13) implements the SRFI String Library. |
| 5891 | |
| 5892 | (srfi srfi-14) implements the SRFI Character-Set Library. |
| 5893 | |
| 5894 | (srfi srfi-17) implements setter and getter-with-setter and redefines |
| 5895 | some accessor procedures as procedures with getters. (such as car, |
| 5896 | cdr, vector-ref etc.) |
| 5897 | |
| 5898 | (srfi srfi-19) implements the SRFI Time/Date Library. |
| 5899 | |
| 5900 | ** New scripts / "executable modules" |
| 5901 | |
| 5902 | Subdirectory "scripts" contains Scheme modules that are packaged to |
| 5903 | also be executable as scripts. At this time, these scripts are available: |
| 5904 | |
| 5905 | display-commentary |
| 5906 | doc-snarf |
| 5907 | generate-autoload |
| 5908 | punify |
| 5909 | read-scheme-source |
| 5910 | use2dot |
| 5911 | |
| 5912 | See README there for more info. |
| 5913 | |
| 5914 | These scripts can be invoked from the shell with the new program |
| 5915 | "guile-tools", which keeps track of installation directory for you. |
| 5916 | For example: |
| 5917 | |
| 5918 | $ guile-tools display-commentary srfi/*.scm |
| 5919 | |
| 5920 | guile-tools is copied to the standard $bindir on "make install". |
| 5921 | |
| 5922 | ** New module (ice-9 stack-catch): |
| 5923 | |
| 5924 | stack-catch is like catch, but saves the current state of the stack in |
| 5925 | the fluid the-last-stack. This fluid can be useful when using the |
| 5926 | debugger and when re-throwing an error. |
| 5927 | |
| 5928 | ** The module (ice-9 and-let*) has been renamed to (ice-9 and-let-star) |
| 5929 | |
| 5930 | This has been done to prevent problems on lesser operating systems |
| 5931 | that can't tolerate `*'s in file names. The exported macro continues |
| 5932 | to be named `and-let*', of course. |
| 5933 | |
| 5934 | On systems that support it, there is also a compatibility module named |
| 5935 | (ice-9 and-let*). It will go away in the next release. |
| 5936 | |
| 5937 | ** New modules (oop goops) etc.: |
| 5938 | |
| 5939 | (oop goops) |
| 5940 | (oop goops describe) |
| 5941 | (oop goops save) |
| 5942 | (oop goops active-slot) |
| 5943 | (oop goops composite-slot) |
| 5944 | |
| 5945 | The Guile Object Oriented Programming System (GOOPS) has been |
| 5946 | integrated into Guile. For further information, consult the GOOPS |
| 5947 | manual and tutorial in the `doc' directory. |
| 5948 | |
| 5949 | ** New module (ice-9 rdelim). |
| 5950 | |
| 5951 | This exports the following procedures which were previously defined |
| 5952 | in the default environment: |
| 5953 | |
| 5954 | read-line read-line! read-delimited read-delimited! %read-delimited! |
| 5955 | %read-line write-line |
| 5956 | |
| 5957 | For backwards compatibility the definitions are still imported into the |
| 5958 | default environment in this version of Guile. However you should add: |
| 5959 | |
| 5960 | (use-modules (ice-9 rdelim)) |
| 5961 | |
| 5962 | to any program which uses the definitions, since this may change in |
| 5963 | future. |
| 5964 | |
| 5965 | Alternatively, if guile-scsh is installed, the (scsh rdelim) module |
| 5966 | can be used for similar functionality. |
| 5967 | |
| 5968 | ** New module (ice-9 rw) |
| 5969 | |
| 5970 | This is a subset of the (scsh rw) module from guile-scsh. Currently |
| 5971 | it defines two procedures: |
| 5972 | |
| 5973 | *** New function: read-string!/partial str [port_or_fdes [start [end]]] |
| 5974 | |
| 5975 | Read characters from a port or file descriptor into a string STR. |
| 5976 | A port must have an underlying file descriptor -- a so-called |
| 5977 | fport. This procedure is scsh-compatible and can efficiently read |
| 5978 | large strings. |
| 5979 | |
| 5980 | *** New function: write-string/partial str [port_or_fdes [start [end]]] |
| 5981 | |
| 5982 | Write characters from a string STR to a port or file descriptor. |
| 5983 | A port must have an underlying file descriptor -- a so-called |
| 5984 | fport. This procedure is mostly compatible and can efficiently |
| 5985 | write large strings. |
| 5986 | |
| 5987 | ** New module (ice-9 match) |
| 5988 | |
| 5989 | This module includes Andrew K. Wright's pattern matcher. See |
| 5990 | ice-9/match.scm for brief description or |
| 5991 | |
| 5992 | http://www.star-lab.com/wright/code.html |
| 5993 | |
| 5994 | for complete documentation. |
| 5995 | |
| 5996 | ** New module (ice-9 buffered-input) |
| 5997 | |
| 5998 | This module provides procedures to construct an input port from an |
| 5999 | underlying source of input that reads and returns its input in chunks. |
| 6000 | The underlying input source is a Scheme procedure, specified by the |
| 6001 | caller, which the port invokes whenever it needs more input. |
| 6002 | |
| 6003 | This is useful when building an input port whose back end is Readline |
| 6004 | or a UI element such as the GtkEntry widget. |
| 6005 | |
| 6006 | ** Documentation |
| 6007 | |
| 6008 | The reference and tutorial documentation that was previously |
| 6009 | distributed separately, as `guile-doc', is now included in the core |
| 6010 | Guile distribution. The documentation consists of the following |
| 6011 | manuals. |
| 6012 | |
| 6013 | - The Guile Tutorial (guile-tut.texi) contains a tutorial introduction |
| 6014 | to using Guile. |
| 6015 | |
| 6016 | - The Guile Reference Manual (guile.texi) contains (or is intended to |
| 6017 | contain) reference documentation on all aspects of Guile. |
| 6018 | |
| 6019 | - The GOOPS Manual (goops.texi) contains both tutorial-style and |
| 6020 | reference documentation for using GOOPS, Guile's Object Oriented |
| 6021 | Programming System. |
| 6022 | |
| 6023 | - The Revised^5 Report on the Algorithmic Language Scheme |
| 6024 | (r5rs.texi). |
| 6025 | |
| 6026 | See the README file in the `doc' directory for more details. |
| 6027 | |
| 6028 | ** There are a couple of examples in the examples/ directory now. |
| 6029 | |
| 6030 | * Changes to the stand-alone interpreter |
| 6031 | |
| 6032 | ** New command line option `--use-srfi' |
| 6033 | |
| 6034 | Using this option, SRFI modules can be loaded on startup and be |
| 6035 | available right from the beginning. This makes programming portable |
| 6036 | Scheme programs easier. |
| 6037 | |
| 6038 | The option `--use-srfi' expects a comma-separated list of numbers, |
| 6039 | each representing a SRFI number to be loaded into the interpreter |
| 6040 | before starting evaluating a script file or the REPL. Additionally, |
| 6041 | the feature identifier for the loaded SRFIs is recognized by |
| 6042 | `cond-expand' when using this option. |
| 6043 | |
| 6044 | Example: |
| 6045 | $ guile --use-srfi=8,13 |
| 6046 | guile> (receive (x z) (values 1 2) (+ 1 2)) |
| 6047 | 3 |
| 6048 | guile> (string-pad "bla" 20) |
| 6049 | " bla" |
| 6050 | |
| 6051 | ** Guile now always starts up in the `(guile-user)' module. |
| 6052 | |
| 6053 | Previously, scripts executed via the `-s' option would run in the |
| 6054 | `(guile)' module and the repl would run in the `(guile-user)' module. |
| 6055 | Now every user action takes place in the `(guile-user)' module by |
| 6056 | default. |
| 6057 | |
| 6058 | * Changes to Scheme functions and syntax |
| 6059 | |
| 6060 | ** Character classifiers work for non-ASCII characters. |
| 6061 | |
| 6062 | The predicates `char-alphabetic?', `char-numeric?', |
| 6063 | `char-whitespace?', `char-lower?', `char-upper?' and `char-is-both?' |
| 6064 | no longer check whether their arguments are ASCII characters. |
| 6065 | Previously, a character would only be considered alphabetic when it |
| 6066 | was also ASCII, for example. |
| 6067 | |
| 6068 | ** Previously deprecated Scheme functions have been removed: |
| 6069 | |
| 6070 | tag - no replacement. |
| 6071 | fseek - replaced by seek. |
| 6072 | list* - replaced by cons*. |
| 6073 | |
| 6074 | ** It's now possible to create modules with controlled environments |
| 6075 | |
| 6076 | Example: |
| 6077 | |
| 6078 | (use-modules (ice-9 safe)) |
| 6079 | (define m (make-safe-module)) |
| 6080 | ;;; m will now be a module containing only a safe subset of R5RS |
| 6081 | (eval '(+ 1 2) m) --> 3 |
| 6082 | (eval 'load m) --> ERROR: Unbound variable: load |
| 6083 | |
| 6084 | ** Evaluation of "()", the empty list, is now an error. |
| 6085 | |
| 6086 | Previously, the expression "()" evaluated to the empty list. This has |
| 6087 | been changed to signal a "missing expression" error. The correct way |
| 6088 | to write the empty list as a literal constant is to use quote: "'()". |
| 6089 | |
| 6090 | ** New concept of `Guile Extensions'. |
| 6091 | |
| 6092 | A Guile Extension is just a ordinary shared library that can be linked |
| 6093 | at run-time. We found it advantageous to give this simple concept a |
| 6094 | dedicated name to distinguish the issues related to shared libraries |
| 6095 | from the issues related to the module system. |
| 6096 | |
| 6097 | *** New function: load-extension |
| 6098 | |
| 6099 | Executing (load-extension lib init) is mostly equivalent to |
| 6100 | |
| 6101 | (dynamic-call init (dynamic-link lib)) |
| 6102 | |
| 6103 | except when scm_register_extension has been called previously. |
| 6104 | Whenever appropriate, you should use `load-extension' instead of |
| 6105 | dynamic-link and dynamic-call. |
| 6106 | |
| 6107 | *** New C function: scm_c_register_extension |
| 6108 | |
| 6109 | This function registers a initialization function for use by |
| 6110 | `load-extension'. Use it when you don't want specific extensions to |
| 6111 | be loaded as shared libraries (for example on platforms that don't |
| 6112 | support dynamic linking). |
| 6113 | |
| 6114 | ** Auto-loading of compiled-code modules is deprecated. |
| 6115 | |
| 6116 | Guile used to be able to automatically find and link a shared |
| 6117 | library to satisfy requests for a module. For example, the module |
| 6118 | `(foo bar)' could be implemented by placing a shared library named |
| 6119 | "foo/libbar.so" (or with a different extension) in a directory on the |
| 6120 | load path of Guile. |
| 6121 | |
| 6122 | This has been found to be too tricky, and is no longer supported. The |
| 6123 | shared libraries are now called "extensions". You should now write a |
| 6124 | small Scheme file that calls `load-extension' to load the shared |
| 6125 | library and initialize it explicitly. |
| 6126 | |
| 6127 | The shared libraries themselves should be installed in the usual |
| 6128 | places for shared libraries, with names like "libguile-foo-bar". |
| 6129 | |
| 6130 | For example, place this into a file "foo/bar.scm" |
| 6131 | |
| 6132 | (define-module (foo bar)) |
| 6133 | |
| 6134 | (load-extension "libguile-foo-bar" "foobar_init") |
| 6135 | |
| 6136 | ** Backward incompatible change: eval EXP ENVIRONMENT-SPECIFIER |
| 6137 | |
| 6138 | `eval' is now R5RS, that is it takes two arguments. |
| 6139 | The second argument is an environment specifier, i.e. either |
| 6140 | |
| 6141 | (scheme-report-environment 5) |
| 6142 | (null-environment 5) |
| 6143 | (interaction-environment) |
| 6144 | |
| 6145 | or |
| 6146 | |
| 6147 | any module. |
| 6148 | |
| 6149 | ** The module system has been made more disciplined. |
| 6150 | |
| 6151 | The function `eval' will save and restore the current module around |
| 6152 | the evaluation of the specified expression. While this expression is |
| 6153 | evaluated, `(current-module)' will now return the right module, which |
| 6154 | is the module specified as the second argument to `eval'. |
| 6155 | |
| 6156 | A consequence of this change is that `eval' is not particularly |
| 6157 | useful when you want allow the evaluated code to change what module is |
| 6158 | designated as the current module and have this change persist from one |
| 6159 | call to `eval' to the next. The read-eval-print-loop is an example |
| 6160 | where `eval' is now inadequate. To compensate, there is a new |
| 6161 | function `primitive-eval' that does not take a module specifier and |
| 6162 | that does not save/restore the current module. You should use this |
| 6163 | function together with `set-current-module', `current-module', etc |
| 6164 | when you want to have more control over the state that is carried from |
| 6165 | one eval to the next. |
| 6166 | |
| 6167 | Additionally, it has been made sure that forms that are evaluated at |
| 6168 | the top level are always evaluated with respect to the current module. |
| 6169 | Previously, subforms of top-level forms such as `begin', `case', |
| 6170 | etc. did not respect changes to the current module although these |
| 6171 | subforms are at the top-level as well. |
| 6172 | |
| 6173 | To prevent strange behavior, the forms `define-module', |
| 6174 | `use-modules', `use-syntax', and `export' have been restricted to only |
| 6175 | work on the top level. The forms `define-public' and |
| 6176 | `defmacro-public' only export the new binding on the top level. They |
| 6177 | behave just like `define' and `defmacro', respectively, when they are |
| 6178 | used in a lexical environment. |
| 6179 | |
| 6180 | Also, `export' will no longer silently re-export bindings imported |
| 6181 | from a used module. It will emit a `deprecation' warning and will |
| 6182 | cease to perform any re-export in the next version. If you actually |
| 6183 | want to re-export bindings, use the new `re-export' in place of |
| 6184 | `export'. The new `re-export' will not make copies of variables when |
| 6185 | rexporting them, as `export' did wrongly. |
| 6186 | |
| 6187 | ** Module system now allows selection and renaming of imported bindings |
| 6188 | |
| 6189 | Previously, when using `use-modules' or the `#:use-module' clause in |
| 6190 | the `define-module' form, all the bindings (association of symbols to |
| 6191 | values) for imported modules were added to the "current module" on an |
| 6192 | as-is basis. This has been changed to allow finer control through two |
| 6193 | new facilities: selection and renaming. |
| 6194 | |
| 6195 | You can now select which of the imported module's bindings are to be |
| 6196 | visible in the current module by using the `:select' clause. This |
| 6197 | clause also can be used to rename individual bindings. For example: |
| 6198 | |
| 6199 | ;; import all bindings no questions asked |
| 6200 | (use-modules (ice-9 common-list)) |
| 6201 | |
| 6202 | ;; import four bindings, renaming two of them; |
| 6203 | ;; the current module sees: every some zonk-y zonk-n |
| 6204 | (use-modules ((ice-9 common-list) |
| 6205 | :select (every some |
| 6206 | (remove-if . zonk-y) |
| 6207 | (remove-if-not . zonk-n)))) |
| 6208 | |
| 6209 | You can also programmatically rename all selected bindings using the |
| 6210 | `:renamer' clause, which specifies a proc that takes a symbol and |
| 6211 | returns another symbol. Because it is common practice to use a prefix, |
| 6212 | we now provide the convenience procedure `symbol-prefix-proc'. For |
| 6213 | example: |
| 6214 | |
| 6215 | ;; import four bindings, renaming two of them specifically, |
| 6216 | ;; and all four w/ prefix "CL:"; |
| 6217 | ;; the current module sees: CL:every CL:some CL:zonk-y CL:zonk-n |
| 6218 | (use-modules ((ice-9 common-list) |
| 6219 | :select (every some |
| 6220 | (remove-if . zonk-y) |
| 6221 | (remove-if-not . zonk-n)) |
| 6222 | :renamer (symbol-prefix-proc 'CL:))) |
| 6223 | |
| 6224 | ;; import four bindings, renaming two of them specifically, |
| 6225 | ;; and all four by upcasing. |
| 6226 | ;; the current module sees: EVERY SOME ZONK-Y ZONK-N |
| 6227 | (define (upcase-symbol sym) |
| 6228 | (string->symbol (string-upcase (symbol->string sym)))) |
| 6229 | |
| 6230 | (use-modules ((ice-9 common-list) |
| 6231 | :select (every some |
| 6232 | (remove-if . zonk-y) |
| 6233 | (remove-if-not . zonk-n)) |
| 6234 | :renamer upcase-symbol)) |
| 6235 | |
| 6236 | Note that programmatic renaming is done *after* individual renaming. |
| 6237 | Also, the above examples show `use-modules', but the same facilities are |
| 6238 | available for the `#:use-module' clause of `define-module'. |
| 6239 | |
| 6240 | See manual for more info. |
| 6241 | |
| 6242 | ** The semantics of guardians have changed. |
| 6243 | |
| 6244 | The changes are for the most part compatible. An important criterion |
| 6245 | was to keep the typical usage of guardians as simple as before, but to |
| 6246 | make the semantics safer and (as a result) more useful. |
| 6247 | |
| 6248 | *** All objects returned from guardians are now properly alive. |
| 6249 | |
| 6250 | It is now guaranteed that any object referenced by an object returned |
| 6251 | from a guardian is alive. It's now impossible for a guardian to |
| 6252 | return a "contained" object before its "containing" object. |
| 6253 | |
| 6254 | One incompatible (but probably not very important) change resulting |
| 6255 | from this is that it is no longer possible to guard objects that |
| 6256 | indirectly reference themselves (i.e. are parts of cycles). If you do |
| 6257 | so accidentally, you'll get a warning. |
| 6258 | |
| 6259 | *** There are now two types of guardians: greedy and sharing. |
| 6260 | |
| 6261 | If you call (make-guardian #t) or just (make-guardian), you'll get a |
| 6262 | greedy guardian, and for (make-guardian #f) a sharing guardian. |
| 6263 | |
| 6264 | Greedy guardians are the default because they are more "defensive". |
| 6265 | You can only greedily guard an object once. If you guard an object |
| 6266 | more than once, once in a greedy guardian and the rest of times in |
| 6267 | sharing guardians, then it is guaranteed that the object won't be |
| 6268 | returned from sharing guardians as long as it is greedily guarded |
| 6269 | and/or alive. |
| 6270 | |
| 6271 | Guardians returned by calls to `make-guardian' can now take one more |
| 6272 | optional parameter, which says whether to throw an error in case an |
| 6273 | attempt is made to greedily guard an object that is already greedily |
| 6274 | guarded. The default is true, i.e. throw an error. If the parameter |
| 6275 | is false, the guardian invocation returns #t if guarding was |
| 6276 | successful and #f if it wasn't. |
| 6277 | |
| 6278 | Also, since greedy guarding is, in effect, a side-effecting operation |
| 6279 | on objects, a new function is introduced: `destroy-guardian!'. |
| 6280 | Invoking this function on a guardian renders it unoperative and, if |
| 6281 | the guardian is greedy, clears the "greedily guarded" property of the |
| 6282 | objects that were guarded by it, thus undoing the side effect. |
| 6283 | |
| 6284 | Note that all this hair is hardly very important, since guardian |
| 6285 | objects are usually permanent. |
| 6286 | |
| 6287 | ** Continuations created by call-with-current-continuation now accept |
| 6288 | any number of arguments, as required by R5RS. |
| 6289 | |
| 6290 | ** New function `issue-deprecation-warning' |
| 6291 | |
| 6292 | This function is used to display the deprecation messages that are |
| 6293 | controlled by GUILE_WARN_DEPRECATION as explained in the README. |
| 6294 | |
| 6295 | (define (id x) |
| 6296 | (issue-deprecation-warning "`id' is deprecated. Use `identity' instead.") |
| 6297 | (identity x)) |
| 6298 | |
| 6299 | guile> (id 1) |
| 6300 | ;; `id' is deprecated. Use `identity' instead. |
| 6301 | 1 |
| 6302 | guile> (id 1) |
| 6303 | 1 |
| 6304 | |
| 6305 | ** New syntax `begin-deprecated' |
| 6306 | |
| 6307 | When deprecated features are included (as determined by the configure |
| 6308 | option --enable-deprecated), `begin-deprecated' is identical to |
| 6309 | `begin'. When deprecated features are excluded, it always evaluates |
| 6310 | to `#f', ignoring the body forms. |
| 6311 | |
| 6312 | ** New function `make-object-property' |
| 6313 | |
| 6314 | This function returns a new `procedure with setter' P that can be used |
| 6315 | to attach a property to objects. When calling P as |
| 6316 | |
| 6317 | (set! (P obj) val) |
| 6318 | |
| 6319 | where `obj' is any kind of object, it attaches `val' to `obj' in such |
| 6320 | a way that it can be retrieved by calling P as |
| 6321 | |
| 6322 | (P obj) |
| 6323 | |
| 6324 | This function will replace procedure properties, symbol properties and |
| 6325 | source properties eventually. |
| 6326 | |
| 6327 | ** Module (ice-9 optargs) now uses keywords instead of `#&'. |
| 6328 | |
| 6329 | Instead of #&optional, #&key, etc you should now use #:optional, |
| 6330 | #:key, etc. Since #:optional is a keyword, you can write it as just |
| 6331 | :optional when (read-set! keywords 'prefix) is active. |
| 6332 | |
| 6333 | The old reader syntax `#&' is still supported, but deprecated. It |
| 6334 | will be removed in the next release. |
| 6335 | |
| 6336 | ** New define-module option: pure |
| 6337 | |
| 6338 | Tells the module system not to include any bindings from the root |
| 6339 | module. |
| 6340 | |
| 6341 | Example: |
| 6342 | |
| 6343 | (define-module (totally-empty-module) |
| 6344 | :pure) |
| 6345 | |
| 6346 | ** New define-module option: export NAME1 ... |
| 6347 | |
| 6348 | Export names NAME1 ... |
| 6349 | |
| 6350 | This option is required if you want to be able to export bindings from |
| 6351 | a module which doesn't import one of `define-public' or `export'. |
| 6352 | |
| 6353 | Example: |
| 6354 | |
| 6355 | (define-module (foo) |
| 6356 | :pure |
| 6357 | :use-module (ice-9 r5rs) |
| 6358 | :export (bar)) |
| 6359 | |
| 6360 | ;;; Note that we're pure R5RS below this point! |
| 6361 | |
| 6362 | (define (bar) |
| 6363 | ...) |
| 6364 | |
| 6365 | ** New function: object->string OBJ |
| 6366 | |
| 6367 | Return a Scheme string obtained by printing a given object. |
| 6368 | |
| 6369 | ** New function: port? X |
| 6370 | |
| 6371 | Returns a boolean indicating whether X is a port. Equivalent to |
| 6372 | `(or (input-port? X) (output-port? X))'. |
| 6373 | |
| 6374 | ** New function: file-port? |
| 6375 | |
| 6376 | Determines whether a given object is a port that is related to a file. |
| 6377 | |
| 6378 | ** New function: port-for-each proc |
| 6379 | |
| 6380 | Apply PROC to each port in the Guile port table in turn. The return |
| 6381 | value is unspecified. More specifically, PROC is applied exactly once |
| 6382 | to every port that exists in the system at the time PORT-FOR-EACH is |
| 6383 | invoked. Changes to the port table while PORT-FOR-EACH is running |
| 6384 | have no effect as far as PORT-FOR-EACH is concerned. |
| 6385 | |
| 6386 | ** New function: dup2 oldfd newfd |
| 6387 | |
| 6388 | A simple wrapper for the `dup2' system call. Copies the file |
| 6389 | descriptor OLDFD to descriptor number NEWFD, replacing the |
| 6390 | previous meaning of NEWFD. Both OLDFD and NEWFD must be integers. |
| 6391 | Unlike for dup->fdes or primitive-move->fdes, no attempt is made |
| 6392 | to move away ports which are using NEWFD. The return value is |
| 6393 | unspecified. |
| 6394 | |
| 6395 | ** New function: close-fdes fd |
| 6396 | |
| 6397 | A simple wrapper for the `close' system call. Close file |
| 6398 | descriptor FD, which must be an integer. Unlike close (*note |
| 6399 | close: Ports and File Descriptors.), the file descriptor will be |
| 6400 | closed even if a port is using it. The return value is |
| 6401 | unspecified. |
| 6402 | |
| 6403 | ** New function: crypt password salt |
| 6404 | |
| 6405 | Encrypts `password' using the standard unix password encryption |
| 6406 | algorithm. |
| 6407 | |
| 6408 | ** New function: chroot path |
| 6409 | |
| 6410 | Change the root directory of the running process to `path'. |
| 6411 | |
| 6412 | ** New functions: getlogin, cuserid |
| 6413 | |
| 6414 | Return the login name or the user name of the current effective user |
| 6415 | id, respectively. |
| 6416 | |
| 6417 | ** New functions: getpriority which who, setpriority which who prio |
| 6418 | |
| 6419 | Get or set the priority of the running process. |
| 6420 | |
| 6421 | ** New function: getpass prompt |
| 6422 | |
| 6423 | Read a password from the terminal, first displaying `prompt' and |
| 6424 | disabling echoing. |
| 6425 | |
| 6426 | ** New function: flock file operation |
| 6427 | |
| 6428 | Set/remove an advisory shared or exclusive lock on `file'. |
| 6429 | |
| 6430 | ** New functions: sethostname name, gethostname |
| 6431 | |
| 6432 | Set or get the hostname of the machine the current process is running |
| 6433 | on. |
| 6434 | |
| 6435 | ** New function: mkstemp! tmpl |
| 6436 | |
| 6437 | mkstemp creates a new unique file in the file system and returns a |
| 6438 | new buffered port open for reading and writing to the file. TMPL |
| 6439 | is a string specifying where the file should be created: it must |
| 6440 | end with `XXXXXX' and will be changed in place to return the name |
| 6441 | of the temporary file. |
| 6442 | |
| 6443 | ** New function: open-input-string string |
| 6444 | |
| 6445 | Return an input string port which delivers the characters from |
| 6446 | `string'. This procedure, together with `open-output-string' and |
| 6447 | `get-output-string' implements SRFI-6. |
| 6448 | |
| 6449 | ** New function: open-output-string |
| 6450 | |
| 6451 | Return an output string port which collects all data written to it. |
| 6452 | The data can then be retrieved by `get-output-string'. |
| 6453 | |
| 6454 | ** New function: get-output-string |
| 6455 | |
| 6456 | Return the contents of an output string port. |
| 6457 | |
| 6458 | ** New function: identity |
| 6459 | |
| 6460 | Return the argument. |
| 6461 | |
| 6462 | ** socket, connect, accept etc., now have support for IPv6. IPv6 addresses |
| 6463 | are represented in Scheme as integers with normal host byte ordering. |
| 6464 | |
| 6465 | ** New function: inet-pton family address |
| 6466 | |
| 6467 | Convert a printable string network address into an integer. Note that |
| 6468 | unlike the C version of this function, the result is an integer with |
| 6469 | normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'. |
| 6470 | e.g., |
| 6471 | |
| 6472 | (inet-pton AF_INET "127.0.0.1") => 2130706433 |
| 6473 | (inet-pton AF_INET6 "::1") => 1 |
| 6474 | |
| 6475 | ** New function: inet-ntop family address |
| 6476 | |
| 6477 | Convert an integer network address into a printable string. Note that |
| 6478 | unlike the C version of this function, the input is an integer with |
| 6479 | normal host byte ordering. FAMILY can be `AF_INET' or `AF_INET6'. |
| 6480 | e.g., |
| 6481 | |
| 6482 | (inet-ntop AF_INET 2130706433) => "127.0.0.1" |
| 6483 | (inet-ntop AF_INET6 (- (expt 2 128) 1)) => |
| 6484 | ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff |
| 6485 | |
| 6486 | ** Deprecated: id |
| 6487 | |
| 6488 | Use `identity' instead. |
| 6489 | |
| 6490 | ** Deprecated: -1+ |
| 6491 | |
| 6492 | Use `1-' instead. |
| 6493 | |
| 6494 | ** Deprecated: return-it |
| 6495 | |
| 6496 | Do without it. |
| 6497 | |
| 6498 | ** Deprecated: string-character-length |
| 6499 | |
| 6500 | Use `string-length' instead. |
| 6501 | |
| 6502 | ** Deprecated: flags |
| 6503 | |
| 6504 | Use `logior' instead. |
| 6505 | |
| 6506 | ** Deprecated: close-all-ports-except. |
| 6507 | |
| 6508 | This was intended for closing ports in a child process after a fork, |
| 6509 | but it has the undesirable side effect of flushing buffers. |
| 6510 | port-for-each is more flexible. |
| 6511 | |
| 6512 | ** The (ice-9 popen) module now attempts to set up file descriptors in |
| 6513 | the child process from the current Scheme ports, instead of using the |
| 6514 | current values of file descriptors 0, 1, and 2 in the parent process. |
| 6515 | |
| 6516 | ** Removed function: builtin-weak-bindings |
| 6517 | |
| 6518 | There is no such concept as a weak binding any more. |
| 6519 | |
| 6520 | ** Removed constants: bignum-radix, scm-line-incrementors |
| 6521 | |
| 6522 | ** define-method: New syntax mandatory. |
| 6523 | |
| 6524 | The new method syntax is now mandatory: |
| 6525 | |
| 6526 | (define-method (NAME ARG-SPEC ...) BODY ...) |
| 6527 | (define-method (NAME ARG-SPEC ... . REST-ARG) BODY ...) |
| 6528 | |
| 6529 | ARG-SPEC ::= ARG-NAME | (ARG-NAME TYPE) |
| 6530 | REST-ARG ::= ARG-NAME |
| 6531 | |
| 6532 | If you have old code using the old syntax, import |
| 6533 | (oop goops old-define-method) before (oop goops) as in: |
| 6534 | |
| 6535 | (use-modules (oop goops old-define-method) (oop goops)) |
| 6536 | |
| 6537 | ** Deprecated function: builtin-variable |
| 6538 | Removed function: builtin-bindings |
| 6539 | |
| 6540 | There is no longer a distinction between builtin or other variables. |
| 6541 | Use module system operations for all variables. |
| 6542 | |
| 6543 | ** Lazy-catch handlers are no longer allowed to return. |
| 6544 | |
| 6545 | That is, a call to `throw', `error', etc is now guaranteed to not |
| 6546 | return. |
| 6547 | |
| 6548 | ** Bugfixes for (ice-9 getopt-long) |
| 6549 | |
| 6550 | This module is now tested using test-suite/tests/getopt-long.test. |
| 6551 | The following bugs have been fixed: |
| 6552 | |
| 6553 | *** Parsing for options that are specified to have `optional' args now checks |
| 6554 | if the next element is an option instead of unconditionally taking it as the |
| 6555 | option arg. |
| 6556 | |
| 6557 | *** An error is now thrown for `--opt=val' when the option description |
| 6558 | does not specify `(value #t)' or `(value optional)'. This condition used to |
| 6559 | be accepted w/o error, contrary to the documentation. |
| 6560 | |
| 6561 | *** The error message for unrecognized options is now more informative. |
| 6562 | It used to be "not a record", an artifact of the implementation. |
| 6563 | |
| 6564 | *** The error message for `--opt' terminating the arg list (no value), when |
| 6565 | `(value #t)' is specified, is now more informative. It used to be "not enough |
| 6566 | args". |
| 6567 | |
| 6568 | *** "Clumped" single-char args now preserve trailing string, use it as arg. |
| 6569 | The expansion used to be like so: |
| 6570 | |
| 6571 | ("-abc5d" "--xyz") => ("-a" "-b" "-c" "--xyz") |
| 6572 | |
| 6573 | Note that the "5d" is dropped. Now it is like so: |
| 6574 | |
| 6575 | ("-abc5d" "--xyz") => ("-a" "-b" "-c" "5d" "--xyz") |
| 6576 | |
| 6577 | This enables single-char options to have adjoining arguments as long as their |
| 6578 | constituent characters are not potential single-char options. |
| 6579 | |
| 6580 | ** (ice-9 session) procedure `arity' now works with (ice-9 optargs) `lambda*' |
| 6581 | |
| 6582 | The `lambda*' and derivative forms in (ice-9 optargs) now set a procedure |
| 6583 | property `arglist', which can be retrieved by `arity'. The result is that |
| 6584 | `arity' can give more detailed information than before: |
| 6585 | |
| 6586 | Before: |
| 6587 | |
| 6588 | guile> (use-modules (ice-9 optargs)) |
| 6589 | guile> (define* (foo #:optional a b c) a) |
| 6590 | guile> (arity foo) |
| 6591 | 0 or more arguments in `lambda*:G0'. |
| 6592 | |
| 6593 | After: |
| 6594 | |
| 6595 | guile> (arity foo) |
| 6596 | 3 optional arguments: `a', `b' and `c'. |
| 6597 | guile> (define* (bar a b #:key c d #:allow-other-keys) a) |
| 6598 | guile> (arity bar) |
| 6599 | 2 required arguments: `a' and `b', 2 keyword arguments: `c' |
| 6600 | and `d', other keywords allowed. |
| 6601 | guile> (define* (baz a b #:optional c #:rest r) a) |
| 6602 | guile> (arity baz) |
| 6603 | 2 required arguments: `a' and `b', 1 optional argument: `c', |
| 6604 | the rest in `r'. |
| 6605 | |
| 6606 | * Changes to the C interface |
| 6607 | |
| 6608 | ** Types have been renamed from scm_*_t to scm_t_*. |
| 6609 | |
| 6610 | This has been done for POSIX sake. It reserves identifiers ending |
| 6611 | with "_t". What a concept. |
| 6612 | |
| 6613 | The old names are still available with status `deprecated'. |
| 6614 | |
| 6615 | ** scm_t_bits (former scm_bits_t) is now a unsigned type. |
| 6616 | |
| 6617 | ** Deprecated features have been removed. |
| 6618 | |
| 6619 | *** Macros removed |
| 6620 | |
| 6621 | SCM_INPORTP, SCM_OUTPORTP SCM_ICHRP, SCM_ICHR, SCM_MAKICHR |
| 6622 | SCM_SETJMPBUF SCM_NSTRINGP SCM_NRWSTRINGP SCM_NVECTORP SCM_DOUBLE_CELLP |
| 6623 | |
| 6624 | *** C Functions removed |
| 6625 | |
| 6626 | scm_sysmissing scm_tag scm_tc16_flo scm_tc_flo |
| 6627 | scm_fseek - replaced by scm_seek. |
| 6628 | gc-thunk - replaced by after-gc-hook. |
| 6629 | gh_int2scmb - replaced by gh_bool2scm. |
| 6630 | scm_tc_dblr - replaced by scm_tc16_real. |
| 6631 | scm_tc_dblc - replaced by scm_tc16_complex. |
| 6632 | scm_list_star - replaced by scm_cons_star. |
| 6633 | |
| 6634 | ** Deprecated: scm_makfromstr |
| 6635 | |
| 6636 | Use scm_mem2string instead. |
| 6637 | |
| 6638 | ** Deprecated: scm_make_shared_substring |
| 6639 | |
| 6640 | Explicit shared substrings will disappear from Guile. |
| 6641 | |
| 6642 | Instead, "normal" strings will be implemented using sharing |
| 6643 | internally, combined with a copy-on-write strategy. |
| 6644 | |
| 6645 | ** Deprecated: scm_read_only_string_p |
| 6646 | |
| 6647 | The concept of read-only strings will disappear in next release of |
| 6648 | Guile. |
| 6649 | |
| 6650 | ** Deprecated: scm_sloppy_memq, scm_sloppy_memv, scm_sloppy_member |
| 6651 | |
| 6652 | Instead, use scm_c_memq or scm_memq, scm_memv, scm_member. |
| 6653 | |
| 6654 | ** New functions: scm_call_0, scm_call_1, scm_call_2, scm_call_3 |
| 6655 | |
| 6656 | Call a procedure with the indicated number of arguments. See "Fly |
| 6657 | Evaluation" in the manual. |
| 6658 | |
| 6659 | ** New functions: scm_apply_0, scm_apply_1, scm_apply_2, scm_apply_3 |
| 6660 | |
| 6661 | Call a procedure with the indicated number of arguments and a list of |
| 6662 | further arguments. See "Fly Evaluation" in the manual. |
| 6663 | |
| 6664 | ** New functions: scm_list_1, scm_list_2, scm_list_3, scm_list_4, scm_list_5 |
| 6665 | |
| 6666 | Create a list of the given number of elements. See "List |
| 6667 | Constructors" in the manual. |
| 6668 | |
| 6669 | ** Renamed function: scm_listify has been replaced by scm_list_n. |
| 6670 | |
| 6671 | ** Deprecated macros: SCM_LIST0, SCM_LIST1, SCM_LIST2, SCM_LIST3, SCM_LIST4, |
| 6672 | SCM_LIST5, SCM_LIST6, SCM_LIST7, SCM_LIST8, SCM_LIST9. |
| 6673 | |
| 6674 | Use functions scm_list_N instead. |
| 6675 | |
| 6676 | ** New function: scm_c_read (SCM port, void *buffer, scm_sizet size) |
| 6677 | |
| 6678 | Used by an application to read arbitrary number of bytes from a port. |
| 6679 | Same semantics as libc read, except that scm_c_read only returns less |
| 6680 | than SIZE bytes if at end-of-file. |
| 6681 | |
| 6682 | Warning: Doesn't update port line and column counts! |
| 6683 | |
| 6684 | ** New function: scm_c_write (SCM port, const void *ptr, scm_sizet size) |
| 6685 | |
| 6686 | Used by an application to write arbitrary number of bytes to an SCM |
| 6687 | port. Similar semantics as libc write. However, unlike libc |
| 6688 | write, scm_c_write writes the requested number of bytes and has no |
| 6689 | return value. |
| 6690 | |
| 6691 | Warning: Doesn't update port line and column counts! |
| 6692 | |
| 6693 | ** New function: scm_init_guile () |
| 6694 | |
| 6695 | In contrast to scm_boot_guile, scm_init_guile will return normally |
| 6696 | after initializing Guile. It is not available on all systems, tho. |
| 6697 | |
| 6698 | ** New functions: scm_str2symbol, scm_mem2symbol |
| 6699 | |
| 6700 | The function scm_str2symbol takes a const char* pointing to a zero-terminated |
| 6701 | field of characters and creates a scheme symbol object from that C string. |
| 6702 | The function scm_mem2symbol takes a const char* and a number of characters and |
| 6703 | creates a symbol from the characters in that memory area. |
| 6704 | |
| 6705 | ** New functions: scm_primitive_make_property |
| 6706 | scm_primitive_property_ref |
| 6707 | scm_primitive_property_set_x |
| 6708 | scm_primitive_property_del_x |
| 6709 | |
| 6710 | These functions implement a new way to deal with object properties. |
| 6711 | See libguile/properties.c for their documentation. |
| 6712 | |
| 6713 | ** New function: scm_done_free (long size) |
| 6714 | |
| 6715 | This function is the inverse of scm_done_malloc. Use it to report the |
| 6716 | amount of smob memory you free. The previous method, which involved |
| 6717 | calling scm_done_malloc with negative argument, was somewhat |
| 6718 | unintuitive (and is still available, of course). |
| 6719 | |
| 6720 | ** New function: scm_c_memq (SCM obj, SCM list) |
| 6721 | |
| 6722 | This function provides a fast C level alternative for scm_memq for the case |
| 6723 | that the list parameter is known to be a proper list. The function is a |
| 6724 | replacement for scm_sloppy_memq, but is stricter in its requirements on its |
| 6725 | list input parameter, since for anything else but a proper list the function's |
| 6726 | behaviour is undefined - it may even crash or loop endlessly. Further, for |
| 6727 | the case that the object is not found in the list, scm_c_memq returns #f which |
| 6728 | is similar to scm_memq, but different from scm_sloppy_memq's behaviour. |
| 6729 | |
| 6730 | ** New functions: scm_remember_upto_here_1, scm_remember_upto_here_2, |
| 6731 | scm_remember_upto_here |
| 6732 | |
| 6733 | These functions replace the function scm_remember. |
| 6734 | |
| 6735 | ** Deprecated function: scm_remember |
| 6736 | |
| 6737 | Use one of the new functions scm_remember_upto_here_1, |
| 6738 | scm_remember_upto_here_2 or scm_remember_upto_here instead. |
| 6739 | |
| 6740 | ** New function: scm_allocate_string |
| 6741 | |
| 6742 | This function replaces the function scm_makstr. |
| 6743 | |
| 6744 | ** Deprecated function: scm_makstr |
| 6745 | |
| 6746 | Use the new function scm_allocate_string instead. |
| 6747 | |
| 6748 | ** New global variable scm_gc_running_p introduced. |
| 6749 | |
| 6750 | Use this variable to find out if garbage collection is being executed. Up to |
| 6751 | now applications have used scm_gc_heap_lock to test if garbage collection was |
| 6752 | running, which also works because of the fact that up to know only the garbage |
| 6753 | collector has set this variable. But, this is an implementation detail that |
| 6754 | may change. Further, scm_gc_heap_lock is not set throughout gc, thus the use |
| 6755 | of this variable is (and has been) not fully safe anyway. |
| 6756 | |
| 6757 | ** New macros: SCM_BITVECTOR_MAX_LENGTH, SCM_UVECTOR_MAX_LENGTH |
| 6758 | |
| 6759 | Use these instead of SCM_LENGTH_MAX. |
| 6760 | |
| 6761 | ** New macros: SCM_CONTINUATION_LENGTH, SCM_CCLO_LENGTH, SCM_STACK_LENGTH, |
| 6762 | SCM_STRING_LENGTH, SCM_SYMBOL_LENGTH, SCM_UVECTOR_LENGTH, |
| 6763 | SCM_BITVECTOR_LENGTH, SCM_VECTOR_LENGTH. |
| 6764 | |
| 6765 | Use these instead of SCM_LENGTH. |
| 6766 | |
| 6767 | ** New macros: SCM_SET_CONTINUATION_LENGTH, SCM_SET_STRING_LENGTH, |
| 6768 | SCM_SET_SYMBOL_LENGTH, SCM_SET_VECTOR_LENGTH, SCM_SET_UVECTOR_LENGTH, |
| 6769 | SCM_SET_BITVECTOR_LENGTH |
| 6770 | |
| 6771 | Use these instead of SCM_SETLENGTH |
| 6772 | |
| 6773 | ** New macros: SCM_STRING_CHARS, SCM_SYMBOL_CHARS, SCM_CCLO_BASE, |
| 6774 | SCM_VECTOR_BASE, SCM_UVECTOR_BASE, SCM_BITVECTOR_BASE, SCM_COMPLEX_MEM, |
| 6775 | SCM_ARRAY_MEM |
| 6776 | |
| 6777 | Use these instead of SCM_CHARS, SCM_UCHARS, SCM_ROCHARS, SCM_ROUCHARS or |
| 6778 | SCM_VELTS. |
| 6779 | |
| 6780 | ** New macros: SCM_SET_BIGNUM_BASE, SCM_SET_STRING_CHARS, |
| 6781 | SCM_SET_SYMBOL_CHARS, SCM_SET_UVECTOR_BASE, SCM_SET_BITVECTOR_BASE, |
| 6782 | SCM_SET_VECTOR_BASE |
| 6783 | |
| 6784 | Use these instead of SCM_SETCHARS. |
| 6785 | |
| 6786 | ** New macro: SCM_BITVECTOR_P |
| 6787 | |
| 6788 | ** New macro: SCM_STRING_COERCE_0TERMINATION_X |
| 6789 | |
| 6790 | Use instead of SCM_COERCE_SUBSTR. |
| 6791 | |
| 6792 | ** New macros: SCM_DIR_OPEN_P, SCM_DIR_FLAG_OPEN |
| 6793 | |
| 6794 | For directory objects, use these instead of SCM_OPDIRP and SCM_OPN. |
| 6795 | |
| 6796 | ** Deprecated macros: SCM_OUTOFRANGE, SCM_NALLOC, SCM_HUP_SIGNAL, |
| 6797 | SCM_INT_SIGNAL, SCM_FPE_SIGNAL, SCM_BUS_SIGNAL, SCM_SEGV_SIGNAL, |
| 6798 | SCM_ALRM_SIGNAL, SCM_GC_SIGNAL, SCM_TICK_SIGNAL, SCM_SIG_ORD, |
| 6799 | SCM_ORD_SIG, SCM_NUM_SIGS, SCM_SYMBOL_SLOTS, SCM_SLOTS, SCM_SLOPPY_STRINGP, |
| 6800 | SCM_VALIDATE_STRINGORSUBSTR, SCM_FREEP, SCM_NFREEP, SCM_CHARS, SCM_UCHARS, |
| 6801 | SCM_VALIDATE_ROSTRING, SCM_VALIDATE_ROSTRING_COPY, |
| 6802 | SCM_VALIDATE_NULLORROSTRING_COPY, SCM_ROLENGTH, SCM_LENGTH, SCM_HUGE_LENGTH, |
| 6803 | SCM_SUBSTRP, SCM_SUBSTR_STR, SCM_SUBSTR_OFFSET, SCM_COERCE_SUBSTR, |
| 6804 | SCM_ROSTRINGP, SCM_RWSTRINGP, SCM_VALIDATE_RWSTRING, SCM_ROCHARS, |
| 6805 | SCM_ROUCHARS, SCM_SETLENGTH, SCM_SETCHARS, SCM_LENGTH_MAX, SCM_GC8MARKP, |
| 6806 | SCM_SETGC8MARK, SCM_CLRGC8MARK, SCM_GCTYP16, SCM_GCCDR, SCM_SUBR_DOC, |
| 6807 | SCM_OPDIRP, SCM_VALIDATE_OPDIR, SCM_WTA, RETURN_SCM_WTA, SCM_CONST_LONG, |
| 6808 | SCM_WNA, SCM_FUNC_NAME, SCM_VALIDATE_NUMBER_COPY, |
| 6809 | SCM_VALIDATE_NUMBER_DEF_COPY, SCM_SLOPPY_CONSP, SCM_SLOPPY_NCONSP, |
| 6810 | SCM_SETAND_CDR, SCM_SETOR_CDR, SCM_SETAND_CAR, SCM_SETOR_CAR |
| 6811 | |
| 6812 | Use SCM_ASSERT_RANGE or SCM_VALIDATE_XXX_RANGE instead of SCM_OUTOFRANGE. |
| 6813 | Use scm_memory_error instead of SCM_NALLOC. |
| 6814 | Use SCM_STRINGP instead of SCM_SLOPPY_STRINGP. |
| 6815 | Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_STRINGORSUBSTR. |
| 6816 | Use SCM_FREE_CELL_P instead of SCM_FREEP/SCM_NFREEP |
| 6817 | Use a type specific accessor macro instead of SCM_CHARS/SCM_UCHARS. |
| 6818 | Use a type specific accessor instead of SCM(_|_RO|_HUGE_)LENGTH. |
| 6819 | Use SCM_VALIDATE_(SYMBOL|STRING) instead of SCM_VALIDATE_ROSTRING. |
| 6820 | Use SCM_STRING_COERCE_0TERMINATION_X instead of SCM_COERCE_SUBSTR. |
| 6821 | Use SCM_STRINGP or SCM_SYMBOLP instead of SCM_ROSTRINGP. |
| 6822 | Use SCM_STRINGP instead of SCM_RWSTRINGP. |
| 6823 | Use SCM_VALIDATE_STRING instead of SCM_VALIDATE_RWSTRING. |
| 6824 | Use SCM_STRING_CHARS instead of SCM_ROCHARS. |
| 6825 | Use SCM_STRING_UCHARS instead of SCM_ROUCHARS. |
| 6826 | Use a type specific setter macro instead of SCM_SETLENGTH. |
| 6827 | Use a type specific setter macro instead of SCM_SETCHARS. |
| 6828 | Use a type specific length macro instead of SCM_LENGTH_MAX. |
| 6829 | Use SCM_GCMARKP instead of SCM_GC8MARKP. |
| 6830 | Use SCM_SETGCMARK instead of SCM_SETGC8MARK. |
| 6831 | Use SCM_CLRGCMARK instead of SCM_CLRGC8MARK. |
| 6832 | Use SCM_TYP16 instead of SCM_GCTYP16. |
| 6833 | Use SCM_CDR instead of SCM_GCCDR. |
| 6834 | Use SCM_DIR_OPEN_P instead of SCM_OPDIRP. |
| 6835 | Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of SCM_WTA. |
| 6836 | Use SCM_MISC_ERROR or SCM_WRONG_TYPE_ARG instead of RETURN_SCM_WTA. |
| 6837 | Use SCM_VCELL_INIT instead of SCM_CONST_LONG. |
| 6838 | Use SCM_WRONG_NUM_ARGS instead of SCM_WNA. |
| 6839 | Use SCM_CONSP instead of SCM_SLOPPY_CONSP. |
| 6840 | Use !SCM_CONSP instead of SCM_SLOPPY_NCONSP. |
| 6841 | |
| 6842 | ** Removed function: scm_struct_init |
| 6843 | |
| 6844 | ** Removed variable: scm_symhash_dim |
| 6845 | |
| 6846 | ** Renamed function: scm_make_cont has been replaced by |
| 6847 | scm_make_continuation, which has a different interface. |
| 6848 | |
| 6849 | ** Deprecated function: scm_call_catching_errors |
| 6850 | |
| 6851 | Use scm_catch or scm_lazy_catch from throw.[ch] instead. |
| 6852 | |
| 6853 | ** Deprecated function: scm_strhash |
| 6854 | |
| 6855 | Use scm_string_hash instead. |
| 6856 | |
| 6857 | ** Deprecated function: scm_vector_set_length_x |
| 6858 | |
| 6859 | Instead, create a fresh vector of the desired size and copy the contents. |
| 6860 | |
| 6861 | ** scm_gensym has changed prototype |
| 6862 | |
| 6863 | scm_gensym now only takes one argument. |
| 6864 | |
| 6865 | ** Deprecated type tags: scm_tc7_ssymbol, scm_tc7_msymbol, scm_tcs_symbols, |
| 6866 | scm_tc7_lvector |
| 6867 | |
| 6868 | There is now only a single symbol type scm_tc7_symbol. |
| 6869 | The tag scm_tc7_lvector was not used anyway. |
| 6870 | |
| 6871 | ** Deprecated function: scm_make_smob_type_mfpe, scm_set_smob_mfpe. |
| 6872 | |
| 6873 | Use scm_make_smob_type and scm_set_smob_XXX instead. |
| 6874 | |
| 6875 | ** New function scm_set_smob_apply. |
| 6876 | |
| 6877 | This can be used to set an apply function to a smob type. |
| 6878 | |
| 6879 | ** Deprecated function: scm_strprint_obj |
| 6880 | |
| 6881 | Use scm_object_to_string instead. |
| 6882 | |
| 6883 | ** Deprecated function: scm_wta |
| 6884 | |
| 6885 | Use scm_wrong_type_arg, or another appropriate error signalling function |
| 6886 | instead. |
| 6887 | |
| 6888 | ** Explicit support for obarrays has been deprecated. |
| 6889 | |
| 6890 | Use `scm_str2symbol' and the generic hashtable functions instead. |
| 6891 | |
| 6892 | ** The concept of `vcells' has been deprecated. |
| 6893 | |
| 6894 | The data type `variable' is now used exclusively. `Vcells' have been |
| 6895 | a low-level concept so you are likely not affected by this change. |
| 6896 | |
| 6897 | *** Deprecated functions: scm_sym2vcell, scm_sysintern, |
| 6898 | scm_sysintern0, scm_symbol_value0, scm_intern, scm_intern0. |
| 6899 | |
| 6900 | Use scm_c_define or scm_c_lookup instead, as appropriate. |
| 6901 | |
| 6902 | *** New functions: scm_c_module_lookup, scm_c_lookup, |
| 6903 | scm_c_module_define, scm_c_define, scm_module_lookup, scm_lookup, |
| 6904 | scm_module_define, scm_define. |
| 6905 | |
| 6906 | These functions work with variables instead of with vcells. |
| 6907 | |
| 6908 | ** New functions for creating and defining `subr's and `gsubr's. |
| 6909 | |
| 6910 | The new functions more clearly distinguish between creating a subr (or |
| 6911 | gsubr) object and adding it to the current module. |
| 6912 | |
| 6913 | These new functions are available: scm_c_make_subr, scm_c_define_subr, |
| 6914 | scm_c_make_subr_with_generic, scm_c_define_subr_with_generic, |
| 6915 | scm_c_make_gsubr, scm_c_define_gsubr, scm_c_make_gsubr_with_generic, |
| 6916 | scm_c_define_gsubr_with_generic. |
| 6917 | |
| 6918 | ** Deprecated functions: scm_make_subr, scm_make_subr_opt, |
| 6919 | scm_make_subr_with_generic, scm_make_gsubr, |
| 6920 | scm_make_gsubr_with_generic. |
| 6921 | |
| 6922 | Use the new ones from above instead. |
| 6923 | |
| 6924 | ** C interface to the module system has changed. |
| 6925 | |
| 6926 | While we suggest that you avoid as many explicit module system |
| 6927 | operations from C as possible for the time being, the C interface has |
| 6928 | been made more similar to the high-level Scheme module system. |
| 6929 | |
| 6930 | *** New functions: scm_c_define_module, scm_c_use_module, |
| 6931 | scm_c_export, scm_c_resolve_module. |
| 6932 | |
| 6933 | They mostly work like their Scheme namesakes. scm_c_define_module |
| 6934 | takes a function that is called a context where the new module is |
| 6935 | current. |
| 6936 | |
| 6937 | *** Deprecated functions: scm_the_root_module, scm_make_module, |
| 6938 | scm_ensure_user_module, scm_load_scheme_module. |
| 6939 | |
| 6940 | Use the new functions instead. |
| 6941 | |
| 6942 | ** Renamed function: scm_internal_with_fluids becomes |
| 6943 | scm_c_with_fluids. |
| 6944 | |
| 6945 | scm_internal_with_fluids is available as a deprecated function. |
| 6946 | |
| 6947 | ** New function: scm_c_with_fluid. |
| 6948 | |
| 6949 | Just like scm_c_with_fluids, but takes one fluid and one value instead |
| 6950 | of lists of same. |
| 6951 | |
| 6952 | ** Deprecated typedefs: long_long, ulong_long. |
| 6953 | |
| 6954 | They are of questionable utility and they pollute the global |
| 6955 | namespace. |
| 6956 | |
| 6957 | ** Deprecated typedef: scm_sizet |
| 6958 | |
| 6959 | It is of questionable utility now that Guile requires ANSI C, and is |
| 6960 | oddly named. |
| 6961 | |
| 6962 | ** Deprecated typedefs: scm_port_rw_active, scm_port, |
| 6963 | scm_ptob_descriptor, scm_debug_info, scm_debug_frame, scm_fport, |
| 6964 | scm_option, scm_rstate, scm_rng, scm_array, scm_array_dim. |
| 6965 | |
| 6966 | Made more compliant with the naming policy by adding a _t at the end. |
| 6967 | |
| 6968 | ** Deprecated functions: scm_mkbig, scm_big2num, scm_adjbig, |
| 6969 | scm_normbig, scm_copybig, scm_2ulong2big, scm_dbl2big, scm_big2dbl |
| 6970 | |
| 6971 | With the exception of the mysterious scm_2ulong2big, they are still |
| 6972 | available under new names (scm_i_mkbig etc). These functions are not |
| 6973 | intended to be used in user code. You should avoid dealing with |
| 6974 | bignums directly, and should deal with numbers in general (which can |
| 6975 | be bignums). |
| 6976 | |
| 6977 | ** Change in behavior: scm_num2long, scm_num2ulong |
| 6978 | |
| 6979 | The scm_num2[u]long functions don't any longer accept an inexact |
| 6980 | argument. This change in behavior is motivated by concordance with |
| 6981 | R5RS: It is more common that a primitive doesn't want to accept an |
| 6982 | inexact for an exact. |
| 6983 | |
| 6984 | ** New functions: scm_short2num, scm_ushort2num, scm_int2num, |
| 6985 | scm_uint2num, scm_size2num, scm_ptrdiff2num, scm_num2short, |
| 6986 | scm_num2ushort, scm_num2int, scm_num2uint, scm_num2ptrdiff, |
| 6987 | scm_num2size. |
| 6988 | |
| 6989 | These are conversion functions between the various ANSI C integral |
| 6990 | types and Scheme numbers. NOTE: The scm_num2xxx functions don't |
| 6991 | accept an inexact argument. |
| 6992 | |
| 6993 | ** New functions: scm_float2num, scm_double2num, |
| 6994 | scm_num2float, scm_num2double. |
| 6995 | |
| 6996 | These are conversion functions between the two ANSI C float types and |
| 6997 | Scheme numbers. |
| 6998 | |
| 6999 | ** New number validation macros: |
| 7000 | SCM_NUM2{SIZE,PTRDIFF,SHORT,USHORT,INT,UINT}[_DEF] |
| 7001 | |
| 7002 | See above. |
| 7003 | |
| 7004 | ** New functions: scm_gc_protect_object, scm_gc_unprotect_object |
| 7005 | |
| 7006 | These are just nicer-named old scm_protect_object and |
| 7007 | scm_unprotect_object. |
| 7008 | |
| 7009 | ** Deprecated functions: scm_protect_object, scm_unprotect_object |
| 7010 | |
| 7011 | ** New functions: scm_gc_[un]register_root, scm_gc_[un]register_roots |
| 7012 | |
| 7013 | These functions can be used to register pointers to locations that |
| 7014 | hold SCM values. |
| 7015 | |
| 7016 | ** Deprecated function: scm_create_hook. |
| 7017 | |
| 7018 | Its sins are: misleading name, non-modularity and lack of general |
| 7019 | usefulness. |
| 7020 | |
| 7021 | \f |
| 7022 | Changes since Guile 1.3.4: |
| 7023 | |
| 7024 | * Changes to the distribution |
| 7025 | |
| 7026 | ** Trees from nightly snapshots and CVS now require you to run autogen.sh. |
| 7027 | |
| 7028 | We've changed the way we handle generated files in the Guile source |
| 7029 | repository. As a result, the procedure for building trees obtained |
| 7030 | from the nightly FTP snapshots or via CVS has changed: |
| 7031 | - You must have appropriate versions of autoconf, automake, and |
| 7032 | libtool installed on your system. See README for info on how to |
| 7033 | obtain these programs. |
| 7034 | - Before configuring the tree, you must first run the script |
| 7035 | `autogen.sh' at the top of the source tree. |
| 7036 | |
| 7037 | The Guile repository used to contain not only source files, written by |
| 7038 | humans, but also some generated files, like configure scripts and |
| 7039 | Makefile.in files. Even though the contents of these files could be |
| 7040 | derived mechanically from other files present, we thought it would |
| 7041 | make the tree easier to build if we checked them into CVS. |
| 7042 | |
| 7043 | However, this approach means that minor differences between |
| 7044 | developer's installed tools and habits affected the whole team. |
| 7045 | So we have removed the generated files from the repository, and |
| 7046 | added the autogen.sh script, which will reconstruct them |
| 7047 | appropriately. |
| 7048 | |
| 7049 | |
| 7050 | ** configure now has experimental options to remove support for certain |
| 7051 | features: |
| 7052 | |
| 7053 | --disable-arrays omit array and uniform array support |
| 7054 | --disable-posix omit posix interfaces |
| 7055 | --disable-networking omit networking interfaces |
| 7056 | --disable-regex omit regular expression interfaces |
| 7057 | |
| 7058 | These are likely to become separate modules some day. |
| 7059 | |
| 7060 | ** New configure option --enable-debug-freelist |
| 7061 | |
| 7062 | This enables a debugging version of SCM_NEWCELL(), and also registers |
| 7063 | an extra primitive, the setter `gc-set-debug-check-freelist!'. |
| 7064 | |
| 7065 | Configure with the --enable-debug-freelist option to enable |
| 7066 | the gc-set-debug-check-freelist! primitive, and then use: |
| 7067 | |
| 7068 | (gc-set-debug-check-freelist! #t) # turn on checking of the freelist |
| 7069 | (gc-set-debug-check-freelist! #f) # turn off checking |
| 7070 | |
| 7071 | Checking of the freelist forces a traversal of the freelist and |
| 7072 | a garbage collection before each allocation of a cell. This can |
| 7073 | slow down the interpreter dramatically, so the setter should be used to |
| 7074 | turn on this extra processing only when necessary. |
| 7075 | |
| 7076 | ** New configure option --enable-debug-malloc |
| 7077 | |
| 7078 | Include code for debugging of calls to scm_must_malloc/realloc/free. |
| 7079 | |
| 7080 | Checks that |
| 7081 | |
| 7082 | 1. objects freed by scm_must_free has been mallocated by scm_must_malloc |
| 7083 | 2. objects reallocated by scm_must_realloc has been allocated by |
| 7084 | scm_must_malloc |
| 7085 | 3. reallocated objects are reallocated with the same what string |
| 7086 | |
| 7087 | But, most importantly, it records the number of allocated objects of |
| 7088 | each kind. This is useful when searching for memory leaks. |
| 7089 | |
| 7090 | A Guile compiled with this option provides the primitive |
| 7091 | `malloc-stats' which returns an alist with pairs of kind and the |
| 7092 | number of objects of that kind. |
| 7093 | |
| 7094 | ** All includes are now referenced relative to the root directory |
| 7095 | |
| 7096 | Since some users have had problems with mixups between Guile and |
| 7097 | system headers, we have decided to always refer to Guile headers via |
| 7098 | their parent directories. This essentially creates a "private name |
| 7099 | space" for Guile headers. This means that the compiler only is given |
| 7100 | -I options for the root build and root source directory. |
| 7101 | |
| 7102 | ** Header files kw.h and genio.h have been removed. |
| 7103 | |
| 7104 | ** The module (ice-9 getopt-gnu-style) has been removed. |
| 7105 | |
| 7106 | ** New module (ice-9 documentation) |
| 7107 | |
| 7108 | Implements the interface to documentation strings associated with |
| 7109 | objects. |
| 7110 | |
| 7111 | ** New module (ice-9 time) |
| 7112 | |
| 7113 | Provides a macro `time', which displays execution time of a given form. |
| 7114 | |
| 7115 | ** New module (ice-9 history) |
| 7116 | |
| 7117 | Loading this module enables value history in the repl. |
| 7118 | |
| 7119 | * Changes to the stand-alone interpreter |
| 7120 | |
| 7121 | ** New command line option --debug |
| 7122 | |
| 7123 | Start Guile with debugging evaluator and backtraces enabled. |
| 7124 | |
| 7125 | This is useful when debugging your .guile init file or scripts. |
| 7126 | |
| 7127 | ** New help facility |
| 7128 | |
| 7129 | Usage: (help NAME) gives documentation about objects named NAME (a symbol) |
| 7130 | (help REGEXP) ditto for objects with names matching REGEXP (a string) |
| 7131 | (help 'NAME) gives documentation for NAME, even if it is not an object |
| 7132 | (help ,EXPR) gives documentation for object returned by EXPR |
| 7133 | (help (my module)) gives module commentary for `(my module)' |
| 7134 | (help) gives this text |
| 7135 | |
| 7136 | `help' searches among bindings exported from loaded modules, while |
| 7137 | `apropos' searches among bindings visible from the "current" module. |
| 7138 | |
| 7139 | Examples: (help help) |
| 7140 | (help cons) |
| 7141 | (help "output-string") |
| 7142 | |
| 7143 | ** `help' and `apropos' now prints full module names |
| 7144 | |
| 7145 | ** Dynamic linking now uses libltdl from the libtool package. |
| 7146 | |
| 7147 | The old system dependent code for doing dynamic linking has been |
| 7148 | replaced with calls to the libltdl functions which do all the hairy |
| 7149 | details for us. |
| 7150 | |
| 7151 | The major improvement is that you can now directly pass libtool |
| 7152 | library names like "libfoo.la" to `dynamic-link' and `dynamic-link' |
| 7153 | will be able to do the best shared library job you can get, via |
| 7154 | libltdl. |
| 7155 | |
| 7156 | The way dynamic libraries are found has changed and is not really |
| 7157 | portable across platforms, probably. It is therefore recommended to |
| 7158 | use absolute filenames when possible. |
| 7159 | |
| 7160 | If you pass a filename without an extension to `dynamic-link', it will |
| 7161 | try a few appropriate ones. Thus, the most platform ignorant way is |
| 7162 | to specify a name like "libfoo", without any directories and |
| 7163 | extensions. |
| 7164 | |
| 7165 | ** Guile COOP threads are now compatible with LinuxThreads |
| 7166 | |
| 7167 | Previously, COOP threading wasn't possible in applications linked with |
| 7168 | Linux POSIX threads due to their use of the stack pointer to find the |
| 7169 | thread context. This has now been fixed with a workaround which uses |
| 7170 | the pthreads to allocate the stack. |
| 7171 | |
| 7172 | ** New primitives: `pkgdata-dir', `site-dir', `library-dir' |
| 7173 | |
| 7174 | ** Positions of erring expression in scripts |
| 7175 | |
| 7176 | With version 1.3.4, the location of the erring expression in Guile |
| 7177 | scipts is no longer automatically reported. (This should have been |
| 7178 | documented before the 1.3.4 release.) |
| 7179 | |
| 7180 | You can get this information by enabling recording of positions of |
| 7181 | source expressions and running the debugging evaluator. Put this at |
| 7182 | the top of your script (or in your "site" file): |
| 7183 | |
| 7184 | (read-enable 'positions) |
| 7185 | (debug-enable 'debug) |
| 7186 | |
| 7187 | ** Backtraces in scripts |
| 7188 | |
| 7189 | It is now possible to get backtraces in scripts. |
| 7190 | |
| 7191 | Put |
| 7192 | |
| 7193 | (debug-enable 'debug 'backtrace) |
| 7194 | |
| 7195 | at the top of the script. |
| 7196 | |
| 7197 | (The first options enables the debugging evaluator. |
| 7198 | The second enables backtraces.) |
| 7199 | |
| 7200 | ** Part of module system symbol lookup now implemented in C |
| 7201 | |
| 7202 | The eval closure of most modules is now implemented in C. Since this |
| 7203 | was one of the bottlenecks for loading speed, Guile now loads code |
| 7204 | substantially faster than before. |
| 7205 | |
| 7206 | ** Attempting to get the value of an unbound variable now produces |
| 7207 | an exception with a key of 'unbound-variable instead of 'misc-error. |
| 7208 | |
| 7209 | ** The initial default output port is now unbuffered if it's using a |
| 7210 | tty device. Previously in this situation it was line-buffered. |
| 7211 | |
| 7212 | ** New hook: after-gc-hook |
| 7213 | |
| 7214 | after-gc-hook takes over the role of gc-thunk. This hook is run at |
| 7215 | the first SCM_TICK after a GC. (Thus, the code is run at the same |
| 7216 | point during evaluation as signal handlers.) |
| 7217 | |
| 7218 | Note that this hook should be used only for diagnostic and debugging |
| 7219 | purposes. It is not certain that it will continue to be well-defined |
| 7220 | when this hook is run in the future. |
| 7221 | |
| 7222 | C programmers: Note the new C level hooks scm_before_gc_c_hook, |
| 7223 | scm_before_sweep_c_hook, scm_after_gc_c_hook. |
| 7224 | |
| 7225 | ** Improvements to garbage collector |
| 7226 | |
| 7227 | Guile 1.4 has a new policy for triggering heap allocation and |
| 7228 | determining the sizes of heap segments. It fixes a number of problems |
| 7229 | in the old GC. |
| 7230 | |
| 7231 | 1. The new policy can handle two separate pools of cells |
| 7232 | (2-word/4-word) better. (The old policy would run wild, allocating |
| 7233 | more and more memory for certain programs.) |
| 7234 | |
| 7235 | 2. The old code would sometimes allocate far too much heap so that the |
| 7236 | Guile process became gigantic. The new code avoids this. |
| 7237 | |
| 7238 | 3. The old code would sometimes allocate too little so that few cells |
| 7239 | were freed at GC so that, in turn, too much time was spent in GC. |
| 7240 | |
| 7241 | 4. The old code would often trigger heap allocation several times in a |
| 7242 | row. (The new scheme predicts how large the segments needs to be |
| 7243 | in order not to need further allocation.) |
| 7244 | |
| 7245 | All in all, the new GC policy will make larger applications more |
| 7246 | efficient. |
| 7247 | |
| 7248 | The new GC scheme also is prepared for POSIX threading. Threads can |
| 7249 | allocate private pools of cells ("clusters") with just a single |
| 7250 | function call. Allocation of single cells from such a cluster can |
| 7251 | then proceed without any need of inter-thread synchronization. |
| 7252 | |
| 7253 | ** New environment variables controlling GC parameters |
| 7254 | |
| 7255 | GUILE_MAX_SEGMENT_SIZE Maximal segment size |
| 7256 | (default = 2097000) |
| 7257 | |
| 7258 | Allocation of 2-word cell heaps: |
| 7259 | |
| 7260 | GUILE_INIT_SEGMENT_SIZE_1 Size of initial heap segment in bytes |
| 7261 | (default = 360000) |
| 7262 | |
| 7263 | GUILE_MIN_YIELD_1 Minimum number of freed cells at each |
| 7264 | GC in percent of total heap size |
| 7265 | (default = 40) |
| 7266 | |
| 7267 | Allocation of 4-word cell heaps |
| 7268 | (used for real numbers and misc other objects): |
| 7269 | |
| 7270 | GUILE_INIT_SEGMENT_SIZE_2, GUILE_MIN_YIELD_2 |
| 7271 | |
| 7272 | (See entry "Way for application to customize GC parameters" under |
| 7273 | section "Changes to the scm_ interface" below.) |
| 7274 | |
| 7275 | ** Guile now implements reals using 4-word cells |
| 7276 | |
| 7277 | This speeds up computation with reals. (They were earlier allocated |
| 7278 | with `malloc'.) There is still some room for optimizations, however. |
| 7279 | |
| 7280 | ** Some further steps toward POSIX thread support have been taken |
| 7281 | |
| 7282 | *** Guile's critical sections (SCM_DEFER/ALLOW_INTS) |
| 7283 | don't have much effect any longer, and many of them will be removed in |
| 7284 | next release. |
| 7285 | |
| 7286 | *** Signals |
| 7287 | are only handled at the top of the evaluator loop, immediately after |
| 7288 | I/O, and in scm_equalp. |
| 7289 | |
| 7290 | *** The GC can allocate thread private pools of pairs. |
| 7291 | |
| 7292 | * Changes to Scheme functions and syntax |
| 7293 | |
| 7294 | ** close-input-port and close-output-port are now R5RS |
| 7295 | |
| 7296 | These procedures have been turned into primitives and have R5RS behaviour. |
| 7297 | |
| 7298 | ** New procedure: simple-format PORT MESSAGE ARG1 ... |
| 7299 | |
| 7300 | (ice-9 boot) makes `format' an alias for `simple-format' until possibly |
| 7301 | extended by the more sophisticated version in (ice-9 format) |
| 7302 | |
| 7303 | (simple-format port message . args) |
| 7304 | Write MESSAGE to DESTINATION, defaulting to `current-output-port'. |
| 7305 | MESSAGE can contain ~A (was %s) and ~S (was %S) escapes. When printed, |
| 7306 | the escapes are replaced with corresponding members of ARGS: |
| 7307 | ~A formats using `display' and ~S formats using `write'. |
| 7308 | If DESTINATION is #t, then use the `current-output-port', |
| 7309 | if DESTINATION is #f, then return a string containing the formatted text. |
| 7310 | Does not add a trailing newline." |
| 7311 | |
| 7312 | ** string-ref: the second argument is no longer optional. |
| 7313 | |
| 7314 | ** string, list->string: no longer accept strings in their arguments, |
| 7315 | only characters, for compatibility with R5RS. |
| 7316 | |
| 7317 | ** New procedure: port-closed? PORT |
| 7318 | Returns #t if PORT is closed or #f if it is open. |
| 7319 | |
| 7320 | ** Deprecated: list* |
| 7321 | |
| 7322 | The list* functionality is now provided by cons* (SRFI-1 compliant) |
| 7323 | |
| 7324 | ** New procedure: cons* ARG1 ARG2 ... ARGn |
| 7325 | |
| 7326 | Like `list', but the last arg provides the tail of the constructed list, |
| 7327 | returning (cons ARG1 (cons ARG2 (cons ... ARGn))). |
| 7328 | |
| 7329 | Requires at least one argument. If given one argument, that argument |
| 7330 | is returned as result. |
| 7331 | |
| 7332 | This function is called `list*' in some other Schemes and in Common LISP. |
| 7333 | |
| 7334 | ** Removed deprecated: serial-map, serial-array-copy!, serial-array-map! |
| 7335 | |
| 7336 | ** New procedure: object-documentation OBJECT |
| 7337 | |
| 7338 | Returns the documentation string associated with OBJECT. The |
| 7339 | procedure uses a caching mechanism so that subsequent lookups are |
| 7340 | faster. |
| 7341 | |
| 7342 | Exported by (ice-9 documentation). |
| 7343 | |
| 7344 | ** module-name now returns full names of modules |
| 7345 | |
| 7346 | Previously, only the last part of the name was returned (`session' for |
| 7347 | `(ice-9 session)'). Ex: `(ice-9 session)'. |
| 7348 | |
| 7349 | * Changes to the gh_ interface |
| 7350 | |
| 7351 | ** Deprecated: gh_int2scmb |
| 7352 | |
| 7353 | Use gh_bool2scm instead. |
| 7354 | |
| 7355 | * Changes to the scm_ interface |
| 7356 | |
| 7357 | ** Guile primitives now carry docstrings! |
| 7358 | |
| 7359 | Thanks to Greg Badros! |
| 7360 | |
| 7361 | ** Guile primitives are defined in a new way: SCM_DEFINE/SCM_DEFINE1/SCM_PROC |
| 7362 | |
| 7363 | Now Guile primitives are defined using the SCM_DEFINE/SCM_DEFINE1/SCM_PROC |
| 7364 | macros and must contain a docstring that is extracted into foo.doc using a new |
| 7365 | guile-doc-snarf script (that uses guile-doc-snarf.awk). |
| 7366 | |
| 7367 | However, a major overhaul of these macros is scheduled for the next release of |
| 7368 | guile. |
| 7369 | |
| 7370 | ** Guile primitives use a new technique for validation of arguments |
| 7371 | |
| 7372 | SCM_VALIDATE_* macros are defined to ease the redundancy and improve |
| 7373 | the readability of argument checking. |
| 7374 | |
| 7375 | ** All (nearly?) K&R prototypes for functions replaced with ANSI C equivalents. |
| 7376 | |
| 7377 | ** New macros: SCM_PACK, SCM_UNPACK |
| 7378 | |
| 7379 | Compose/decompose an SCM value. |
| 7380 | |
| 7381 | The SCM type is now treated as an abstract data type and may be defined as a |
| 7382 | long, a void* or as a struct, depending on the architecture and compile time |
| 7383 | options. This makes it easier to find several types of bugs, for example when |
| 7384 | SCM values are treated as integers without conversion. Values of the SCM type |
| 7385 | should be treated as "atomic" values. These macros are used when |
| 7386 | composing/decomposing an SCM value, either because you want to access |
| 7387 | individual bits, or because you want to treat it as an integer value. |
| 7388 | |
| 7389 | E.g., in order to set bit 7 in an SCM value x, use the expression |
| 7390 | |
| 7391 | SCM_PACK (SCM_UNPACK (x) | 0x80) |
| 7392 | |
| 7393 | ** The name property of hooks is deprecated. |
| 7394 | Thus, the use of SCM_HOOK_NAME and scm_make_hook_with_name is deprecated. |
| 7395 | |
| 7396 | You can emulate this feature by using object properties. |
| 7397 | |
| 7398 | ** Deprecated macros: SCM_INPORTP, SCM_OUTPORTP, SCM_CRDY, SCM_ICHRP, |
| 7399 | SCM_ICHR, SCM_MAKICHR, SCM_SETJMPBUF, SCM_NSTRINGP, SCM_NRWSTRINGP, |
| 7400 | SCM_NVECTORP |
| 7401 | |
| 7402 | These macros will be removed in a future release of Guile. |
| 7403 | |
| 7404 | ** The following types, functions and macros from numbers.h are deprecated: |
| 7405 | scm_dblproc, SCM_UNEGFIXABLE, SCM_FLOBUFLEN, SCM_INEXP, SCM_CPLXP, SCM_REAL, |
| 7406 | SCM_IMAG, SCM_REALPART, scm_makdbl, SCM_SINGP, SCM_NUM2DBL, SCM_NO_BIGDIG |
| 7407 | |
| 7408 | ** Port internals: the rw_random variable in the scm_port structure |
| 7409 | must be set to non-zero in any random access port. In recent Guile |
| 7410 | releases it was only set for bidirectional random-access ports. |
| 7411 | |
| 7412 | ** Port internals: the seek ptob procedure is now responsible for |
| 7413 | resetting the buffers if required. The change was made so that in the |
| 7414 | special case of reading the current position (i.e., seek p 0 SEEK_CUR) |
| 7415 | the fport and strport ptobs can avoid resetting the buffers, |
| 7416 | in particular to avoid discarding unread chars. An existing port |
| 7417 | type can be fixed by adding something like the following to the |
| 7418 | beginning of the ptob seek procedure: |
| 7419 | |
| 7420 | if (pt->rw_active == SCM_PORT_READ) |
| 7421 | scm_end_input (object); |
| 7422 | else if (pt->rw_active == SCM_PORT_WRITE) |
| 7423 | ptob->flush (object); |
| 7424 | |
| 7425 | although to actually avoid resetting the buffers and discard unread |
| 7426 | chars requires further hacking that depends on the characteristics |
| 7427 | of the ptob. |
| 7428 | |
| 7429 | ** Deprecated functions: scm_fseek, scm_tag |
| 7430 | |
| 7431 | These functions are no longer used and will be removed in a future version. |
| 7432 | |
| 7433 | ** The scm_sysmissing procedure is no longer used in libguile. |
| 7434 | Unless it turns out to be unexpectedly useful to somebody, it will be |
| 7435 | removed in a future version. |
| 7436 | |
| 7437 | ** The format of error message strings has changed |
| 7438 | |
| 7439 | The two C procedures: scm_display_error and scm_error, as well as the |
| 7440 | primitive `scm-error', now use scm_simple_format to do their work. |
| 7441 | This means that the message strings of all code must be updated to use |
| 7442 | ~A where %s was used before, and ~S where %S was used before. |
| 7443 | |
| 7444 | During the period when there still are a lot of old Guiles out there, |
| 7445 | you might want to support both old and new versions of Guile. |
| 7446 | |
| 7447 | There are basically two methods to achieve this. Both methods use |
| 7448 | autoconf. Put |
| 7449 | |
| 7450 | AC_CHECK_FUNCS(scm_simple_format) |
| 7451 | |
| 7452 | in your configure.in. |
| 7453 | |
| 7454 | Method 1: Use the string concatenation features of ANSI C's |
| 7455 | preprocessor. |
| 7456 | |
| 7457 | In C: |
| 7458 | |
| 7459 | #ifdef HAVE_SCM_SIMPLE_FORMAT |
| 7460 | #define FMT_S "~S" |
| 7461 | #else |
| 7462 | #define FMT_S "%S" |
| 7463 | #endif |
| 7464 | |
| 7465 | Then represent each of your error messages using a preprocessor macro: |
| 7466 | |
| 7467 | #define E_SPIDER_ERROR "There's a spider in your " ## FMT_S ## "!!!" |
| 7468 | |
| 7469 | In Scheme: |
| 7470 | |
| 7471 | (define fmt-s (if (defined? 'simple-format) "~S" "%S")) |
| 7472 | (define make-message string-append) |
| 7473 | |
| 7474 | (define e-spider-error (make-message "There's a spider in your " fmt-s "!!!")) |
| 7475 | |
| 7476 | Method 2: Use the oldfmt function found in doc/oldfmt.c. |
| 7477 | |
| 7478 | In C: |
| 7479 | |
| 7480 | scm_misc_error ("picnic", scm_c_oldfmt0 ("There's a spider in your ~S!!!"), |
| 7481 | ...); |
| 7482 | |
| 7483 | In Scheme: |
| 7484 | |
| 7485 | (scm-error 'misc-error "picnic" (oldfmt "There's a spider in your ~S!!!") |
| 7486 | ...) |
| 7487 | |
| 7488 | |
| 7489 | ** Deprecated: coop_mutex_init, coop_condition_variable_init |
| 7490 | |
| 7491 | Don't use the functions coop_mutex_init and |
| 7492 | coop_condition_variable_init. They will change. |
| 7493 | |
| 7494 | Use scm_mutex_init and scm_cond_init instead. |
| 7495 | |
| 7496 | ** New function: int scm_cond_timedwait (scm_cond_t *COND, scm_mutex_t *MUTEX, const struct timespec *ABSTIME) |
| 7497 | `scm_cond_timedwait' atomically unlocks MUTEX and waits on |
| 7498 | COND, as `scm_cond_wait' does, but it also bounds the duration |
| 7499 | of the wait. If COND has not been signaled before time ABSTIME, |
| 7500 | the mutex MUTEX is re-acquired and `scm_cond_timedwait' |
| 7501 | returns the error code `ETIMEDOUT'. |
| 7502 | |
| 7503 | The ABSTIME parameter specifies an absolute time, with the same |
| 7504 | origin as `time' and `gettimeofday': an ABSTIME of 0 corresponds |
| 7505 | to 00:00:00 GMT, January 1, 1970. |
| 7506 | |
| 7507 | ** New function: scm_cond_broadcast (scm_cond_t *COND) |
| 7508 | `scm_cond_broadcast' restarts all the threads that are waiting |
| 7509 | on the condition variable COND. Nothing happens if no threads are |
| 7510 | waiting on COND. |
| 7511 | |
| 7512 | ** New function: scm_key_create (scm_key_t *KEY, void (*destr_function) (void *)) |
| 7513 | `scm_key_create' allocates a new TSD key. The key is stored in |
| 7514 | the location pointed to by KEY. There is no limit on the number |
| 7515 | of keys allocated at a given time. The value initially associated |
| 7516 | with the returned key is `NULL' in all currently executing threads. |
| 7517 | |
| 7518 | The DESTR_FUNCTION argument, if not `NULL', specifies a destructor |
| 7519 | function associated with the key. When a thread terminates, |
| 7520 | DESTR_FUNCTION is called on the value associated with the key in |
| 7521 | that thread. The DESTR_FUNCTION is not called if a key is deleted |
| 7522 | with `scm_key_delete' or a value is changed with |
| 7523 | `scm_setspecific'. The order in which destructor functions are |
| 7524 | called at thread termination time is unspecified. |
| 7525 | |
| 7526 | Destructors are not yet implemented. |
| 7527 | |
| 7528 | ** New function: scm_setspecific (scm_key_t KEY, const void *POINTER) |
| 7529 | `scm_setspecific' changes the value associated with KEY in the |
| 7530 | calling thread, storing the given POINTER instead. |
| 7531 | |
| 7532 | ** New function: scm_getspecific (scm_key_t KEY) |
| 7533 | `scm_getspecific' returns the value currently associated with |
| 7534 | KEY in the calling thread. |
| 7535 | |
| 7536 | ** New function: scm_key_delete (scm_key_t KEY) |
| 7537 | `scm_key_delete' deallocates a TSD key. It does not check |
| 7538 | whether non-`NULL' values are associated with that key in the |
| 7539 | currently executing threads, nor call the destructor function |
| 7540 | associated with the key. |
| 7541 | |
| 7542 | ** New function: scm_c_hook_init (scm_c_hook_t *HOOK, void *HOOK_DATA, scm_c_hook_type_t TYPE) |
| 7543 | |
| 7544 | Initialize a C level hook HOOK with associated HOOK_DATA and type |
| 7545 | TYPE. (See scm_c_hook_run ().) |
| 7546 | |
| 7547 | ** New function: scm_c_hook_add (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA, int APPENDP) |
| 7548 | |
| 7549 | Add hook function FUNC with associated FUNC_DATA to HOOK. If APPENDP |
| 7550 | is true, add it last, otherwise first. The same FUNC can be added |
| 7551 | multiple times if FUNC_DATA differ and vice versa. |
| 7552 | |
| 7553 | ** New function: scm_c_hook_remove (scm_c_hook_t *HOOK, scm_c_hook_function_t FUNC, void *FUNC_DATA) |
| 7554 | |
| 7555 | Remove hook function FUNC with associated FUNC_DATA from HOOK. A |
| 7556 | function is only removed if both FUNC and FUNC_DATA matches. |
| 7557 | |
| 7558 | ** New function: void *scm_c_hook_run (scm_c_hook_t *HOOK, void *DATA) |
| 7559 | |
| 7560 | Run hook HOOK passing DATA to the hook functions. |
| 7561 | |
| 7562 | If TYPE is SCM_C_HOOK_NORMAL, all hook functions are run. The value |
| 7563 | returned is undefined. |
| 7564 | |
| 7565 | If TYPE is SCM_C_HOOK_OR, hook functions are run until a function |
| 7566 | returns a non-NULL value. This value is returned as the result of |
| 7567 | scm_c_hook_run. If all functions return NULL, NULL is returned. |
| 7568 | |
| 7569 | If TYPE is SCM_C_HOOK_AND, hook functions are run until a function |
| 7570 | returns a NULL value, and NULL is returned. If all functions returns |
| 7571 | a non-NULL value, the last value is returned. |
| 7572 | |
| 7573 | ** New C level GC hooks |
| 7574 | |
| 7575 | Five new C level hooks has been added to the garbage collector. |
| 7576 | |
| 7577 | scm_before_gc_c_hook |
| 7578 | scm_after_gc_c_hook |
| 7579 | |
| 7580 | are run before locking and after unlocking the heap. The system is |
| 7581 | thus in a mode where evaluation can take place. (Except that |
| 7582 | scm_before_gc_c_hook must not allocate new cells.) |
| 7583 | |
| 7584 | scm_before_mark_c_hook |
| 7585 | scm_before_sweep_c_hook |
| 7586 | scm_after_sweep_c_hook |
| 7587 | |
| 7588 | are run when the heap is locked. These are intended for extension of |
| 7589 | the GC in a modular fashion. Examples are the weaks and guardians |
| 7590 | modules. |
| 7591 | |
| 7592 | ** Way for application to customize GC parameters |
| 7593 | |
| 7594 | The application can set up other default values for the GC heap |
| 7595 | allocation parameters |
| 7596 | |
| 7597 | GUILE_INIT_HEAP_SIZE_1, GUILE_MIN_YIELD_1, |
| 7598 | GUILE_INIT_HEAP_SIZE_2, GUILE_MIN_YIELD_2, |
| 7599 | GUILE_MAX_SEGMENT_SIZE, |
| 7600 | |
| 7601 | by setting |
| 7602 | |
| 7603 | scm_default_init_heap_size_1, scm_default_min_yield_1, |
| 7604 | scm_default_init_heap_size_2, scm_default_min_yield_2, |
| 7605 | scm_default_max_segment_size |
| 7606 | |
| 7607 | respectively before callong scm_boot_guile. |
| 7608 | |
| 7609 | (See entry "New environment variables ..." in section |
| 7610 | "Changes to the stand-alone interpreter" above.) |
| 7611 | |
| 7612 | ** scm_protect_object/scm_unprotect_object now nest |
| 7613 | |
| 7614 | This means that you can call scm_protect_object multiple times on an |
| 7615 | object and count on the object being protected until |
| 7616 | scm_unprotect_object has been call the same number of times. |
| 7617 | |
| 7618 | The functions also have better time complexity. |
| 7619 | |
| 7620 | Still, it is usually possible to structure the application in a way |
| 7621 | that you don't need to use these functions. For example, if you use a |
| 7622 | protected standard Guile list to keep track of live objects rather |
| 7623 | than some custom data type, objects will die a natural death when they |
| 7624 | are no longer needed. |
| 7625 | |
| 7626 | ** Deprecated type tags: scm_tc16_flo, scm_tc_flo, scm_tc_dblr, scm_tc_dblc |
| 7627 | |
| 7628 | Guile does not provide the float representation for inexact real numbers any |
| 7629 | more. Now, only doubles are used to represent inexact real numbers. Further, |
| 7630 | the tag names scm_tc_dblr and scm_tc_dblc have been changed to scm_tc16_real |
| 7631 | and scm_tc16_complex, respectively. |
| 7632 | |
| 7633 | ** Removed deprecated type scm_smobfuns |
| 7634 | |
| 7635 | ** Removed deprecated function scm_newsmob |
| 7636 | |
| 7637 | ** Warning: scm_make_smob_type_mfpe might become deprecated in a future release |
| 7638 | |
| 7639 | There is an ongoing discussion among the developers whether to |
| 7640 | deprecate `scm_make_smob_type_mfpe' or not. Please use the current |
| 7641 | standard interface (scm_make_smob_type, scm_set_smob_XXX) in new code |
| 7642 | until this issue has been settled. |
| 7643 | |
| 7644 | ** Removed deprecated type tag scm_tc16_kw |
| 7645 | |
| 7646 | ** Added type tag scm_tc16_keyword |
| 7647 | |
| 7648 | (This was introduced already in release 1.3.4 but was not documented |
| 7649 | until now.) |
| 7650 | |
| 7651 | ** gdb_print now prints "*** Guile not initialized ***" until Guile initialized |
| 7652 | |
| 7653 | * Changes to system call interfaces: |
| 7654 | |
| 7655 | ** The "select" procedure now tests port buffers for the ability to |
| 7656 | provide input or accept output. Previously only the underlying file |
| 7657 | descriptors were checked. |
| 7658 | |
| 7659 | ** New variable PIPE_BUF: the maximum number of bytes that can be |
| 7660 | atomically written to a pipe. |
| 7661 | |
| 7662 | ** If a facility is not available on the system when Guile is |
| 7663 | compiled, the corresponding primitive procedure will not be defined. |
| 7664 | Previously it would have been defined but would throw a system-error |
| 7665 | exception if called. Exception handlers which catch this case may |
| 7666 | need minor modification: an error will be thrown with key |
| 7667 | 'unbound-variable instead of 'system-error. Alternatively it's |
| 7668 | now possible to use `defined?' to check whether the facility is |
| 7669 | available. |
| 7670 | |
| 7671 | ** Procedures which depend on the timezone should now give the correct |
| 7672 | result on systems which cache the TZ environment variable, even if TZ |
| 7673 | is changed without calling tzset. |
| 7674 | |
| 7675 | * Changes to the networking interfaces: |
| 7676 | |
| 7677 | ** New functions: htons, ntohs, htonl, ntohl: for converting short and |
| 7678 | long integers between network and host format. For now, it's not |
| 7679 | particularly convenient to do this kind of thing, but consider: |
| 7680 | |
| 7681 | (define write-network-long |
| 7682 | (lambda (value port) |
| 7683 | (let ((v (make-uniform-vector 1 1 0))) |
| 7684 | (uniform-vector-set! v 0 (htonl value)) |
| 7685 | (uniform-vector-write v port)))) |
| 7686 | |
| 7687 | (define read-network-long |
| 7688 | (lambda (port) |
| 7689 | (let ((v (make-uniform-vector 1 1 0))) |
| 7690 | (uniform-vector-read! v port) |
| 7691 | (ntohl (uniform-vector-ref v 0))))) |
| 7692 | |
| 7693 | ** If inet-aton fails, it now throws an error with key 'misc-error |
| 7694 | instead of 'system-error, since errno is not relevant. |
| 7695 | |
| 7696 | ** Certain gethostbyname/gethostbyaddr failures now throw errors with |
| 7697 | specific keys instead of 'system-error. The latter is inappropriate |
| 7698 | since errno will not have been set. The keys are: |
| 7699 | 'host-not-found, 'try-again, 'no-recovery and 'no-data. |
| 7700 | |
| 7701 | ** sethostent, setnetent, setprotoent, setservent: now take an |
| 7702 | optional argument STAYOPEN, which specifies whether the database |
| 7703 | remains open after a database entry is accessed randomly (e.g., using |
| 7704 | gethostbyname for the hosts database.) The default is #f. Previously |
| 7705 | #t was always used. |
| 7706 | |
| 7707 | \f |
| 7708 | Changes since Guile 1.3.2: |
| 7709 | |
| 7710 | * Changes to the stand-alone interpreter |
| 7711 | |
| 7712 | ** Debugger |
| 7713 | |
| 7714 | An initial version of the Guile debugger written by Chris Hanson has |
| 7715 | been added. The debugger is still under development but is included |
| 7716 | in the distribution anyway since it is already quite useful. |
| 7717 | |
| 7718 | Type |
| 7719 | |
| 7720 | (debug) |
| 7721 | |
| 7722 | after an error to enter the debugger. Type `help' inside the debugger |
| 7723 | for a description of available commands. |
| 7724 | |
| 7725 | If you prefer to have stack frames numbered and printed in |
| 7726 | anti-chronological order and prefer up in the stack to be down on the |
| 7727 | screen as is the case in gdb, you can put |
| 7728 | |
| 7729 | (debug-enable 'backwards) |
| 7730 | |
| 7731 | in your .guile startup file. (However, this means that Guile can't |
| 7732 | use indentation to indicate stack level.) |
| 7733 | |
| 7734 | The debugger is autoloaded into Guile at the first use. |
| 7735 | |
| 7736 | ** Further enhancements to backtraces |
| 7737 | |
| 7738 | There is a new debug option `width' which controls the maximum width |
| 7739 | on the screen of printed stack frames. Fancy printing parameters |
| 7740 | ("level" and "length" as in Common LISP) are adaptively adjusted for |
| 7741 | each stack frame to give maximum information while still fitting |
| 7742 | within the bounds. If the stack frame can't be made to fit by |
| 7743 | adjusting parameters, it is simply cut off at the end. This is marked |
| 7744 | with a `$'. |
| 7745 | |
| 7746 | ** Some modules are now only loaded when the repl is started |
| 7747 | |
| 7748 | The modules (ice-9 debug), (ice-9 session), (ice-9 threads) and (ice-9 |
| 7749 | regex) are now loaded into (guile-user) only if the repl has been |
| 7750 | started. The effect is that the startup time for scripts has been |
| 7751 | reduced to 30% of what it was previously. |
| 7752 | |
| 7753 | Correctly written scripts load the modules they require at the top of |
| 7754 | the file and should not be affected by this change. |
| 7755 | |
| 7756 | ** Hooks are now represented as smobs |
| 7757 | |
| 7758 | * Changes to Scheme functions and syntax |
| 7759 | |
| 7760 | ** Readline support has changed again. |
| 7761 | |
| 7762 | The old (readline-activator) module is gone. Use (ice-9 readline) |
| 7763 | instead, which now contains all readline functionality. So the code |
| 7764 | to activate readline is now |
| 7765 | |
| 7766 | (use-modules (ice-9 readline)) |
| 7767 | (activate-readline) |
| 7768 | |
| 7769 | This should work at any time, including from the guile prompt. |
| 7770 | |
| 7771 | To avoid confusion about the terms of Guile's license, please only |
| 7772 | enable readline for your personal use; please don't make it the |
| 7773 | default for others. Here is why we make this rather odd-sounding |
| 7774 | request: |
| 7775 | |
| 7776 | Guile is normally licensed under a weakened form of the GNU General |
| 7777 | Public License, which allows you to link code with Guile without |
| 7778 | placing that code under the GPL. This exception is important to some |
| 7779 | people. |
| 7780 | |
| 7781 | However, since readline is distributed under the GNU General Public |
| 7782 | License, when you link Guile with readline, either statically or |
| 7783 | dynamically, you effectively change Guile's license to the strict GPL. |
| 7784 | Whenever you link any strictly GPL'd code into Guile, uses of Guile |
| 7785 | which are normally permitted become forbidden. This is a rather |
| 7786 | non-obvious consequence of the licensing terms. |
| 7787 | |
| 7788 | So, to make sure things remain clear, please let people choose for |
| 7789 | themselves whether to link GPL'd libraries like readline with Guile. |
| 7790 | |
| 7791 | ** regexp-substitute/global has changed slightly, but incompatibly. |
| 7792 | |
| 7793 | If you include a function in the item list, the string of the match |
| 7794 | object it receives is the same string passed to |
| 7795 | regexp-substitute/global, not some suffix of that string. |
| 7796 | Correspondingly, the match's positions are relative to the entire |
| 7797 | string, not the suffix. |
| 7798 | |
| 7799 | If the regexp can match the empty string, the way matches are chosen |
| 7800 | from the string has changed. regexp-substitute/global recognizes the |
| 7801 | same set of matches that list-matches does; see below. |
| 7802 | |
| 7803 | ** New function: list-matches REGEXP STRING [FLAGS] |
| 7804 | |
| 7805 | Return a list of match objects, one for every non-overlapping, maximal |
| 7806 | match of REGEXP in STRING. The matches appear in left-to-right order. |
| 7807 | list-matches only reports matches of the empty string if there are no |
| 7808 | other matches which begin on, end at, or include the empty match's |
| 7809 | position. |
| 7810 | |
| 7811 | If present, FLAGS is passed as the FLAGS argument to regexp-exec. |
| 7812 | |
| 7813 | ** New function: fold-matches REGEXP STRING INIT PROC [FLAGS] |
| 7814 | |
| 7815 | For each match of REGEXP in STRING, apply PROC to the match object, |
| 7816 | and the last value PROC returned, or INIT for the first call. Return |
| 7817 | the last value returned by PROC. We apply PROC to the matches as they |
| 7818 | appear from left to right. |
| 7819 | |
| 7820 | This function recognizes matches according to the same criteria as |
| 7821 | list-matches. |
| 7822 | |
| 7823 | Thus, you could define list-matches like this: |
| 7824 | |
| 7825 | (define (list-matches regexp string . flags) |
| 7826 | (reverse! (apply fold-matches regexp string '() cons flags))) |
| 7827 | |
| 7828 | If present, FLAGS is passed as the FLAGS argument to regexp-exec. |
| 7829 | |
| 7830 | ** Hooks |
| 7831 | |
| 7832 | *** New function: hook? OBJ |
| 7833 | |
| 7834 | Return #t if OBJ is a hook, otherwise #f. |
| 7835 | |
| 7836 | *** New function: make-hook-with-name NAME [ARITY] |
| 7837 | |
| 7838 | Return a hook with name NAME and arity ARITY. The default value for |
| 7839 | ARITY is 0. The only effect of NAME is that it will appear when the |
| 7840 | hook object is printed to ease debugging. |
| 7841 | |
| 7842 | *** New function: hook-empty? HOOK |
| 7843 | |
| 7844 | Return #t if HOOK doesn't contain any procedures, otherwise #f. |
| 7845 | |
| 7846 | *** New function: hook->list HOOK |
| 7847 | |
| 7848 | Return a list of the procedures that are called when run-hook is |
| 7849 | applied to HOOK. |
| 7850 | |
| 7851 | ** `map' signals an error if its argument lists are not all the same length. |
| 7852 | |
| 7853 | This is the behavior required by R5RS, so this change is really a bug |
| 7854 | fix. But it seems to affect a lot of people's code, so we're |
| 7855 | mentioning it here anyway. |
| 7856 | |
| 7857 | ** Print-state handling has been made more transparent |
| 7858 | |
| 7859 | Under certain circumstances, ports are represented as a port with an |
| 7860 | associated print state. Earlier, this pair was represented as a pair |
| 7861 | (see "Some magic has been added to the printer" below). It is now |
| 7862 | indistinguishable (almost; see `get-print-state') from a port on the |
| 7863 | user level. |
| 7864 | |
| 7865 | *** New function: port-with-print-state OUTPUT-PORT PRINT-STATE |
| 7866 | |
| 7867 | Return a new port with the associated print state PRINT-STATE. |
| 7868 | |
| 7869 | *** New function: get-print-state OUTPUT-PORT |
| 7870 | |
| 7871 | Return the print state associated with this port if it exists, |
| 7872 | otherwise return #f. |
| 7873 | |
| 7874 | *** New function: directory-stream? OBJECT |
| 7875 | |
| 7876 | Returns true iff OBJECT is a directory stream --- the sort of object |
| 7877 | returned by `opendir'. |
| 7878 | |
| 7879 | ** New function: using-readline? |
| 7880 | |
| 7881 | Return #t if readline is in use in the current repl. |
| 7882 | |
| 7883 | ** structs will be removed in 1.4 |
| 7884 | |
| 7885 | Structs will be replaced in Guile 1.4. We will merge GOOPS into Guile |
| 7886 | and use GOOPS objects as the fundamental record type. |
| 7887 | |
| 7888 | * Changes to the scm_ interface |
| 7889 | |
| 7890 | ** structs will be removed in 1.4 |
| 7891 | |
| 7892 | The entire current struct interface (struct.c, struct.h) will be |
| 7893 | replaced in Guile 1.4. We will merge GOOPS into libguile and use |
| 7894 | GOOPS objects as the fundamental record type. |
| 7895 | |
| 7896 | ** The internal representation of subr's has changed |
| 7897 | |
| 7898 | Instead of giving a hint to the subr name, the CAR field of the subr |
| 7899 | now contains an index to a subr entry in scm_subr_table. |
| 7900 | |
| 7901 | *** New variable: scm_subr_table |
| 7902 | |
| 7903 | An array of subr entries. A subr entry contains the name, properties |
| 7904 | and documentation associated with the subr. The properties and |
| 7905 | documentation slots are not yet used. |
| 7906 | |
| 7907 | ** A new scheme for "forwarding" calls to a builtin to a generic function |
| 7908 | |
| 7909 | It is now possible to extend the functionality of some Guile |
| 7910 | primitives by letting them defer a call to a GOOPS generic function on |
| 7911 | argument mismatch. This means that there is no loss of efficiency in |
| 7912 | normal evaluation. |
| 7913 | |
| 7914 | Example: |
| 7915 | |
| 7916 | (use-modules (oop goops)) ; Must be GOOPS version 0.2. |
| 7917 | (define-method + ((x <string>) (y <string>)) |
| 7918 | (string-append x y)) |
| 7919 | |
| 7920 | + will still be as efficient as usual in numerical calculations, but |
| 7921 | can also be used for concatenating strings. |
| 7922 | |
| 7923 | Who will be the first one to extend Guile's numerical tower to |
| 7924 | rationals? :) [OK, there a few other things to fix before this can |
| 7925 | be made in a clean way.] |
| 7926 | |
| 7927 | *** New snarf macros for defining primitives: SCM_GPROC, SCM_GPROC1 |
| 7928 | |
| 7929 | New macro: SCM_GPROC (CNAME, SNAME, REQ, OPT, VAR, CFUNC, GENERIC) |
| 7930 | |
| 7931 | New macro: SCM_GPROC1 (CNAME, SNAME, TYPE, CFUNC, GENERIC) |
| 7932 | |
| 7933 | These do the same job as SCM_PROC and SCM_PROC1, but they also define |
| 7934 | a variable GENERIC which can be used by the dispatch macros below. |
| 7935 | |
| 7936 | [This is experimental code which may change soon.] |
| 7937 | |
| 7938 | *** New macros for forwarding control to a generic on arg type error |
| 7939 | |
| 7940 | New macro: SCM_WTA_DISPATCH_1 (GENERIC, ARG1, POS, SUBR) |
| 7941 | |
| 7942 | New macro: SCM_WTA_DISPATCH_2 (GENERIC, ARG1, ARG2, POS, SUBR) |
| 7943 | |
| 7944 | These correspond to the scm_wta function call, and have the same |
| 7945 | behaviour until the user has called the GOOPS primitive |
| 7946 | `enable-primitive-generic!'. After that, these macros will apply the |
| 7947 | generic function GENERIC to the argument(s) instead of calling |
| 7948 | scm_wta. |
| 7949 | |
| 7950 | [This is experimental code which may change soon.] |
| 7951 | |
| 7952 | *** New macros for argument testing with generic dispatch |
| 7953 | |
| 7954 | New macro: SCM_GASSERT1 (COND, GENERIC, ARG1, POS, SUBR) |
| 7955 | |
| 7956 | New macro: SCM_GASSERT2 (COND, GENERIC, ARG1, ARG2, POS, SUBR) |
| 7957 | |
| 7958 | These correspond to the SCM_ASSERT macro, but will defer control to |
| 7959 | GENERIC on error after `enable-primitive-generic!' has been called. |
| 7960 | |
| 7961 | [This is experimental code which may change soon.] |
| 7962 | |
| 7963 | ** New function: SCM scm_eval_body (SCM body, SCM env) |
| 7964 | |
| 7965 | Evaluates the body of a special form. |
| 7966 | |
| 7967 | ** The internal representation of struct's has changed |
| 7968 | |
| 7969 | Previously, four slots were allocated for the procedure(s) of entities |
| 7970 | and operators. The motivation for this representation had to do with |
| 7971 | the structure of the evaluator, the wish to support tail-recursive |
| 7972 | generic functions, and efficiency. Since the generic function |
| 7973 | dispatch mechanism has changed, there is no longer a need for such an |
| 7974 | expensive representation, and the representation has been simplified. |
| 7975 | |
| 7976 | This should not make any difference for most users. |
| 7977 | |
| 7978 | ** GOOPS support has been cleaned up. |
| 7979 | |
| 7980 | Some code has been moved from eval.c to objects.c and code in both of |
| 7981 | these compilation units has been cleaned up and better structured. |
| 7982 | |
| 7983 | *** New functions for applying generic functions |
| 7984 | |
| 7985 | New function: SCM scm_apply_generic (GENERIC, ARGS) |
| 7986 | New function: SCM scm_call_generic_0 (GENERIC) |
| 7987 | New function: SCM scm_call_generic_1 (GENERIC, ARG1) |
| 7988 | New function: SCM scm_call_generic_2 (GENERIC, ARG1, ARG2) |
| 7989 | New function: SCM scm_call_generic_3 (GENERIC, ARG1, ARG2, ARG3) |
| 7990 | |
| 7991 | ** Deprecated function: scm_make_named_hook |
| 7992 | |
| 7993 | It is now replaced by: |
| 7994 | |
| 7995 | ** New function: SCM scm_create_hook (const char *name, int arity) |
| 7996 | |
| 7997 | Creates a hook in the same way as make-hook above but also |
| 7998 | binds a variable named NAME to it. |
| 7999 | |
| 8000 | This is the typical way of creating a hook from C code. |
| 8001 | |
| 8002 | Currently, the variable is created in the "current" module. |
| 8003 | This might change when we get the new module system. |
| 8004 | |
| 8005 | [The behaviour is identical to scm_make_named_hook.] |
| 8006 | |
| 8007 | |
| 8008 | \f |
| 8009 | Changes since Guile 1.3: |
| 8010 | |
| 8011 | * Changes to mailing lists |
| 8012 | |
| 8013 | ** Some of the Guile mailing lists have moved to sourceware.cygnus.com. |
| 8014 | |
| 8015 | See the README file to find current addresses for all the Guile |
| 8016 | mailing lists. |
| 8017 | |
| 8018 | * Changes to the distribution |
| 8019 | |
| 8020 | ** Readline support is no longer included with Guile by default. |
| 8021 | |
| 8022 | Based on the different license terms of Guile and Readline, we |
| 8023 | concluded that Guile should not *by default* cause the linking of |
| 8024 | Readline into an application program. Readline support is now offered |
| 8025 | as a separate module, which is linked into an application only when |
| 8026 | you explicitly specify it. |
| 8027 | |
| 8028 | Although Guile is GNU software, its distribution terms add a special |
| 8029 | exception to the usual GNU General Public License (GPL). Guile's |
| 8030 | license includes a clause that allows you to link Guile with non-free |
| 8031 | programs. We add this exception so as not to put Guile at a |
| 8032 | disadvantage vis-a-vis other extensibility packages that support other |
| 8033 | languages. |
| 8034 | |
| 8035 | In contrast, the GNU Readline library is distributed under the GNU |
| 8036 | General Public License pure and simple. This means that you may not |
| 8037 | link Readline, even dynamically, into an application unless it is |
| 8038 | distributed under a free software license that is compatible the GPL. |
| 8039 | |
| 8040 | Because of this difference in distribution terms, an application that |
| 8041 | can use Guile may not be able to use Readline. Now users will be |
| 8042 | explicitly offered two independent decisions about the use of these |
| 8043 | two packages. |
| 8044 | |
| 8045 | You can activate the readline support by issuing |
| 8046 | |
| 8047 | (use-modules (readline-activator)) |
| 8048 | (activate-readline) |
| 8049 | |
| 8050 | from your ".guile" file, for example. |
| 8051 | |
| 8052 | * Changes to the stand-alone interpreter |
| 8053 | |
| 8054 | ** All builtins now print as primitives. |
| 8055 | Previously builtin procedures not belonging to the fundamental subr |
| 8056 | types printed as #<compiled closure #<primitive-procedure gsubr-apply>>. |
| 8057 | Now, they print as #<primitive-procedure NAME>. |
| 8058 | |
| 8059 | ** Backtraces slightly more intelligible. |
| 8060 | gsubr-apply and macro transformer application frames no longer appear |
| 8061 | in backtraces. |
| 8062 | |
| 8063 | * Changes to Scheme functions and syntax |
| 8064 | |
| 8065 | ** Guile now correctly handles internal defines by rewriting them into |
| 8066 | their equivalent letrec. Previously, internal defines would |
| 8067 | incrementally add to the innermost environment, without checking |
| 8068 | whether the restrictions specified in RnRS were met. This lead to the |
| 8069 | correct behaviour when these restriction actually were met, but didn't |
| 8070 | catch all illegal uses. Such an illegal use could lead to crashes of |
| 8071 | the Guile interpreter or other unwanted results. An example of |
| 8072 | incorrect internal defines that made Guile behave erratically: |
| 8073 | |
| 8074 | (let () |
| 8075 | (define a 1) |
| 8076 | (define (b) a) |
| 8077 | (define c (1+ (b))) |
| 8078 | (define d 3) |
| 8079 | |
| 8080 | (b)) |
| 8081 | |
| 8082 | => 2 |
| 8083 | |
| 8084 | The problem with this example is that the definition of `c' uses the |
| 8085 | value of `b' directly. This confuses the meoization machine of Guile |
| 8086 | so that the second call of `b' (this time in a larger environment that |
| 8087 | also contains bindings for `c' and `d') refers to the binding of `c' |
| 8088 | instead of `a'. You could also make Guile crash with a variation on |
| 8089 | this theme: |
| 8090 | |
| 8091 | (define (foo flag) |
| 8092 | (define a 1) |
| 8093 | (define (b flag) (if flag a 1)) |
| 8094 | (define c (1+ (b flag))) |
| 8095 | (define d 3) |
| 8096 | |
| 8097 | (b #t)) |
| 8098 | |
| 8099 | (foo #f) |
| 8100 | (foo #t) |
| 8101 | |
| 8102 | From now on, Guile will issue an `Unbound variable: b' error message |
| 8103 | for both examples. |
| 8104 | |
| 8105 | ** Hooks |
| 8106 | |
| 8107 | A hook contains a list of functions which should be called on |
| 8108 | particular occasions in an existing program. Hooks are used for |
| 8109 | customization. |
| 8110 | |
| 8111 | A window manager might have a hook before-window-map-hook. The window |
| 8112 | manager uses the function run-hooks to call all functions stored in |
| 8113 | before-window-map-hook each time a window is mapped. The user can |
| 8114 | store functions in the hook using add-hook!. |
| 8115 | |
| 8116 | In Guile, hooks are first class objects. |
| 8117 | |
| 8118 | *** New function: make-hook [N_ARGS] |
| 8119 | |
| 8120 | Return a hook for hook functions which can take N_ARGS arguments. |
| 8121 | The default value for N_ARGS is 0. |
| 8122 | |
| 8123 | (See also scm_make_named_hook below.) |
| 8124 | |
| 8125 | *** New function: add-hook! HOOK PROC [APPEND_P] |
| 8126 | |
| 8127 | Put PROC at the beginning of the list of functions stored in HOOK. |
| 8128 | If APPEND_P is supplied, and non-false, put PROC at the end instead. |
| 8129 | |
| 8130 | PROC must be able to take the number of arguments specified when the |
| 8131 | hook was created. |
| 8132 | |
| 8133 | If PROC already exists in HOOK, then remove it first. |
| 8134 | |
| 8135 | *** New function: remove-hook! HOOK PROC |
| 8136 | |
| 8137 | Remove PROC from the list of functions in HOOK. |
| 8138 | |
| 8139 | *** New function: reset-hook! HOOK |
| 8140 | |
| 8141 | Clear the list of hook functions stored in HOOK. |
| 8142 | |
| 8143 | *** New function: run-hook HOOK ARG1 ... |
| 8144 | |
| 8145 | Run all hook functions stored in HOOK with arguments ARG1 ... . |
| 8146 | The number of arguments supplied must correspond to the number given |
| 8147 | when the hook was created. |
| 8148 | |
| 8149 | ** The function `dynamic-link' now takes optional keyword arguments. |
| 8150 | The only keyword argument that is currently defined is `:global |
| 8151 | BOOL'. With it, you can control whether the shared library will be |
| 8152 | linked in global mode or not. In global mode, the symbols from the |
| 8153 | linked library can be used to resolve references from other |
| 8154 | dynamically linked libraries. In non-global mode, the linked |
| 8155 | library is essentially invisible and can only be accessed via |
| 8156 | `dynamic-func', etc. The default is now to link in global mode. |
| 8157 | Previously, the default has been non-global mode. |
| 8158 | |
| 8159 | The `#:global' keyword is only effective on platforms that support |
| 8160 | the dlopen family of functions. |
| 8161 | |
| 8162 | ** New function `provided?' |
| 8163 | |
| 8164 | - Function: provided? FEATURE |
| 8165 | Return true iff FEATURE is supported by this installation of |
| 8166 | Guile. FEATURE must be a symbol naming a feature; the global |
| 8167 | variable `*features*' is a list of available features. |
| 8168 | |
| 8169 | ** Changes to the module (ice-9 expect): |
| 8170 | |
| 8171 | *** The expect-strings macro now matches `$' in a regular expression |
| 8172 | only at a line-break or end-of-file by default. Previously it would |
| 8173 | match the end of the string accumulated so far. The old behaviour |
| 8174 | can be obtained by setting the variable `expect-strings-exec-flags' |
| 8175 | to 0. |
| 8176 | |
| 8177 | *** The expect-strings macro now uses a variable `expect-strings-exec-flags' |
| 8178 | for the regexp-exec flags. If `regexp/noteol' is included, then `$' |
| 8179 | in a regular expression will still match before a line-break or |
| 8180 | end-of-file. The default is `regexp/noteol'. |
| 8181 | |
| 8182 | *** The expect-strings macro now uses a variable |
| 8183 | `expect-strings-compile-flags' for the flags to be supplied to |
| 8184 | `make-regexp'. The default is `regexp/newline', which was previously |
| 8185 | hard-coded. |
| 8186 | |
| 8187 | *** The expect macro now supplies two arguments to a match procedure: |
| 8188 | the current accumulated string and a flag to indicate whether |
| 8189 | end-of-file has been reached. Previously only the string was supplied. |
| 8190 | If end-of-file is reached, the match procedure will be called an |
| 8191 | additional time with the same accumulated string as the previous call |
| 8192 | but with the flag set. |
| 8193 | |
| 8194 | ** New module (ice-9 format), implementing the Common Lisp `format' function. |
| 8195 | |
| 8196 | This code, and the documentation for it that appears here, was |
| 8197 | borrowed from SLIB, with minor adaptations for Guile. |
| 8198 | |
| 8199 | - Function: format DESTINATION FORMAT-STRING . ARGUMENTS |
| 8200 | An almost complete implementation of Common LISP format description |
| 8201 | according to the CL reference book `Common LISP' from Guy L. |
| 8202 | Steele, Digital Press. Backward compatible to most of the |
| 8203 | available Scheme format implementations. |
| 8204 | |
| 8205 | Returns `#t', `#f' or a string; has side effect of printing |
| 8206 | according to FORMAT-STRING. If DESTINATION is `#t', the output is |
| 8207 | to the current output port and `#t' is returned. If DESTINATION |
| 8208 | is `#f', a formatted string is returned as the result of the call. |
| 8209 | NEW: If DESTINATION is a string, DESTINATION is regarded as the |
| 8210 | format string; FORMAT-STRING is then the first argument and the |
| 8211 | output is returned as a string. If DESTINATION is a number, the |
| 8212 | output is to the current error port if available by the |
| 8213 | implementation. Otherwise DESTINATION must be an output port and |
| 8214 | `#t' is returned. |
| 8215 | |
| 8216 | FORMAT-STRING must be a string. In case of a formatting error |
| 8217 | format returns `#f' and prints a message on the current output or |
| 8218 | error port. Characters are output as if the string were output by |
| 8219 | the `display' function with the exception of those prefixed by a |
| 8220 | tilde (~). For a detailed description of the FORMAT-STRING syntax |
| 8221 | please consult a Common LISP format reference manual. For a test |
| 8222 | suite to verify this format implementation load `formatst.scm'. |
| 8223 | Please send bug reports to `lutzeb@cs.tu-berlin.de'. |
| 8224 | |
| 8225 | Note: `format' is not reentrant, i.e. only one `format'-call may |
| 8226 | be executed at a time. |
| 8227 | |
| 8228 | |
| 8229 | *** Format Specification (Format version 3.0) |
| 8230 | |
| 8231 | Please consult a Common LISP format reference manual for a detailed |
| 8232 | description of the format string syntax. For a demonstration of the |
| 8233 | implemented directives see `formatst.scm'. |
| 8234 | |
| 8235 | This implementation supports directive parameters and modifiers (`:' |
| 8236 | and `@' characters). Multiple parameters must be separated by a comma |
| 8237 | (`,'). Parameters can be numerical parameters (positive or negative), |
| 8238 | character parameters (prefixed by a quote character (`''), variable |
| 8239 | parameters (`v'), number of rest arguments parameter (`#'), empty and |
| 8240 | default parameters. Directive characters are case independent. The |
| 8241 | general form of a directive is: |
| 8242 | |
| 8243 | DIRECTIVE ::= ~{DIRECTIVE-PARAMETER,}[:][@]DIRECTIVE-CHARACTER |
| 8244 | |
| 8245 | DIRECTIVE-PARAMETER ::= [ [-|+]{0-9}+ | 'CHARACTER | v | # ] |
| 8246 | |
| 8247 | *** Implemented CL Format Control Directives |
| 8248 | |
| 8249 | Documentation syntax: Uppercase characters represent the |
| 8250 | corresponding control directive characters. Lowercase characters |
| 8251 | represent control directive parameter descriptions. |
| 8252 | |
| 8253 | `~A' |
| 8254 | Any (print as `display' does). |
| 8255 | `~@A' |
| 8256 | left pad. |
| 8257 | |
| 8258 | `~MINCOL,COLINC,MINPAD,PADCHARA' |
| 8259 | full padding. |
| 8260 | |
| 8261 | `~S' |
| 8262 | S-expression (print as `write' does). |
| 8263 | `~@S' |
| 8264 | left pad. |
| 8265 | |
| 8266 | `~MINCOL,COLINC,MINPAD,PADCHARS' |
| 8267 | full padding. |
| 8268 | |
| 8269 | `~D' |
| 8270 | Decimal. |
| 8271 | `~@D' |
| 8272 | print number sign always. |
| 8273 | |
| 8274 | `~:D' |
| 8275 | print comma separated. |
| 8276 | |
| 8277 | `~MINCOL,PADCHAR,COMMACHARD' |
| 8278 | padding. |
| 8279 | |
| 8280 | `~X' |
| 8281 | Hexadecimal. |
| 8282 | `~@X' |
| 8283 | print number sign always. |
| 8284 | |
| 8285 | `~:X' |
| 8286 | print comma separated. |
| 8287 | |
| 8288 | `~MINCOL,PADCHAR,COMMACHARX' |
| 8289 | padding. |
| 8290 | |
| 8291 | `~O' |
| 8292 | Octal. |
| 8293 | `~@O' |
| 8294 | print number sign always. |
| 8295 | |
| 8296 | `~:O' |
| 8297 | print comma separated. |
| 8298 | |
| 8299 | `~MINCOL,PADCHAR,COMMACHARO' |
| 8300 | padding. |
| 8301 | |
| 8302 | `~B' |
| 8303 | Binary. |
| 8304 | `~@B' |
| 8305 | print number sign always. |
| 8306 | |
| 8307 | `~:B' |
| 8308 | print comma separated. |
| 8309 | |
| 8310 | `~MINCOL,PADCHAR,COMMACHARB' |
| 8311 | padding. |
| 8312 | |
| 8313 | `~NR' |
| 8314 | Radix N. |
| 8315 | `~N,MINCOL,PADCHAR,COMMACHARR' |
| 8316 | padding. |
| 8317 | |
| 8318 | `~@R' |
| 8319 | print a number as a Roman numeral. |
| 8320 | |
| 8321 | `~:@R' |
| 8322 | print a number as an "old fashioned" Roman numeral. |
| 8323 | |
| 8324 | `~:R' |
| 8325 | print a number as an ordinal English number. |
| 8326 | |
| 8327 | `~:@R' |
| 8328 | print a number as a cardinal English number. |
| 8329 | |
| 8330 | `~P' |
| 8331 | Plural. |
| 8332 | `~@P' |
| 8333 | prints `y' and `ies'. |
| 8334 | |
| 8335 | `~:P' |
| 8336 | as `~P but jumps 1 argument backward.' |
| 8337 | |
| 8338 | `~:@P' |
| 8339 | as `~@P but jumps 1 argument backward.' |
| 8340 | |
| 8341 | `~C' |
| 8342 | Character. |
| 8343 | `~@C' |
| 8344 | prints a character as the reader can understand it (i.e. `#\' |
| 8345 | prefixing). |
| 8346 | |
| 8347 | `~:C' |
| 8348 | prints a character as emacs does (eg. `^C' for ASCII 03). |
| 8349 | |
| 8350 | `~F' |
| 8351 | Fixed-format floating-point (prints a flonum like MMM.NNN). |
| 8352 | `~WIDTH,DIGITS,SCALE,OVERFLOWCHAR,PADCHARF' |
| 8353 | `~@F' |
| 8354 | If the number is positive a plus sign is printed. |
| 8355 | |
| 8356 | `~E' |
| 8357 | Exponential floating-point (prints a flonum like MMM.NNN`E'EE). |
| 8358 | `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARE' |
| 8359 | `~@E' |
| 8360 | If the number is positive a plus sign is printed. |
| 8361 | |
| 8362 | `~G' |
| 8363 | General floating-point (prints a flonum either fixed or |
| 8364 | exponential). |
| 8365 | `~WIDTH,DIGITS,EXPONENTDIGITS,SCALE,OVERFLOWCHAR,PADCHAR,EXPONENTCHARG' |
| 8366 | `~@G' |
| 8367 | If the number is positive a plus sign is printed. |
| 8368 | |
| 8369 | `~$' |
| 8370 | Dollars floating-point (prints a flonum in fixed with signs |
| 8371 | separated). |
| 8372 | `~DIGITS,SCALE,WIDTH,PADCHAR$' |
| 8373 | `~@$' |
| 8374 | If the number is positive a plus sign is printed. |
| 8375 | |
| 8376 | `~:@$' |
| 8377 | A sign is always printed and appears before the padding. |
| 8378 | |
| 8379 | `~:$' |
| 8380 | The sign appears before the padding. |
| 8381 | |
| 8382 | `~%' |
| 8383 | Newline. |
| 8384 | `~N%' |
| 8385 | print N newlines. |
| 8386 | |
| 8387 | `~&' |
| 8388 | print newline if not at the beginning of the output line. |
| 8389 | `~N&' |
| 8390 | prints `~&' and then N-1 newlines. |
| 8391 | |
| 8392 | `~|' |
| 8393 | Page Separator. |
| 8394 | `~N|' |
| 8395 | print N page separators. |
| 8396 | |
| 8397 | `~~' |
| 8398 | Tilde. |
| 8399 | `~N~' |
| 8400 | print N tildes. |
| 8401 | |
| 8402 | `~'<newline> |
| 8403 | Continuation Line. |
| 8404 | `~:'<newline> |
| 8405 | newline is ignored, white space left. |
| 8406 | |
| 8407 | `~@'<newline> |
| 8408 | newline is left, white space ignored. |
| 8409 | |
| 8410 | `~T' |
| 8411 | Tabulation. |
| 8412 | `~@T' |
| 8413 | relative tabulation. |
| 8414 | |
| 8415 | `~COLNUM,COLINCT' |
| 8416 | full tabulation. |
| 8417 | |
| 8418 | `~?' |
| 8419 | Indirection (expects indirect arguments as a list). |
| 8420 | `~@?' |
| 8421 | extracts indirect arguments from format arguments. |
| 8422 | |
| 8423 | `~(STR~)' |
| 8424 | Case conversion (converts by `string-downcase'). |
| 8425 | `~:(STR~)' |
| 8426 | converts by `string-capitalize'. |
| 8427 | |
| 8428 | `~@(STR~)' |
| 8429 | converts by `string-capitalize-first'. |
| 8430 | |
| 8431 | `~:@(STR~)' |
| 8432 | converts by `string-upcase'. |
| 8433 | |
| 8434 | `~*' |
| 8435 | Argument Jumping (jumps 1 argument forward). |
| 8436 | `~N*' |
| 8437 | jumps N arguments forward. |
| 8438 | |
| 8439 | `~:*' |
| 8440 | jumps 1 argument backward. |
| 8441 | |
| 8442 | `~N:*' |
| 8443 | jumps N arguments backward. |
| 8444 | |
| 8445 | `~@*' |
| 8446 | jumps to the 0th argument. |
| 8447 | |
| 8448 | `~N@*' |
| 8449 | jumps to the Nth argument (beginning from 0) |
| 8450 | |
| 8451 | `~[STR0~;STR1~;...~;STRN~]' |
| 8452 | Conditional Expression (numerical clause conditional). |
| 8453 | `~N[' |
| 8454 | take argument from N. |
| 8455 | |
| 8456 | `~@[' |
| 8457 | true test conditional. |
| 8458 | |
| 8459 | `~:[' |
| 8460 | if-else-then conditional. |
| 8461 | |
| 8462 | `~;' |
| 8463 | clause separator. |
| 8464 | |
| 8465 | `~:;' |
| 8466 | default clause follows. |
| 8467 | |
| 8468 | `~{STR~}' |
| 8469 | Iteration (args come from the next argument (a list)). |
| 8470 | `~N{' |
| 8471 | at most N iterations. |
| 8472 | |
| 8473 | `~:{' |
| 8474 | args from next arg (a list of lists). |
| 8475 | |
| 8476 | `~@{' |
| 8477 | args from the rest of arguments. |
| 8478 | |
| 8479 | `~:@{' |
| 8480 | args from the rest args (lists). |
| 8481 | |
| 8482 | `~^' |
| 8483 | Up and out. |
| 8484 | `~N^' |
| 8485 | aborts if N = 0 |
| 8486 | |
| 8487 | `~N,M^' |
| 8488 | aborts if N = M |
| 8489 | |
| 8490 | `~N,M,K^' |
| 8491 | aborts if N <= M <= K |
| 8492 | |
| 8493 | *** Not Implemented CL Format Control Directives |
| 8494 | |
| 8495 | `~:A' |
| 8496 | print `#f' as an empty list (see below). |
| 8497 | |
| 8498 | `~:S' |
| 8499 | print `#f' as an empty list (see below). |
| 8500 | |
| 8501 | `~<~>' |
| 8502 | Justification. |
| 8503 | |
| 8504 | `~:^' |
| 8505 | (sorry I don't understand its semantics completely) |
| 8506 | |
| 8507 | *** Extended, Replaced and Additional Control Directives |
| 8508 | |
| 8509 | `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHD' |
| 8510 | `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHX' |
| 8511 | `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHO' |
| 8512 | `~MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHB' |
| 8513 | `~N,MINCOL,PADCHAR,COMMACHAR,COMMAWIDTHR' |
| 8514 | COMMAWIDTH is the number of characters between two comma |
| 8515 | characters. |
| 8516 | |
| 8517 | `~I' |
| 8518 | print a R4RS complex number as `~F~@Fi' with passed parameters for |
| 8519 | `~F'. |
| 8520 | |
| 8521 | `~Y' |
| 8522 | Pretty print formatting of an argument for scheme code lists. |
| 8523 | |
| 8524 | `~K' |
| 8525 | Same as `~?.' |
| 8526 | |
| 8527 | `~!' |
| 8528 | Flushes the output if format DESTINATION is a port. |
| 8529 | |
| 8530 | `~_' |
| 8531 | Print a `#\space' character |
| 8532 | `~N_' |
| 8533 | print N `#\space' characters. |
| 8534 | |
| 8535 | `~/' |
| 8536 | Print a `#\tab' character |
| 8537 | `~N/' |
| 8538 | print N `#\tab' characters. |
| 8539 | |
| 8540 | `~NC' |
| 8541 | Takes N as an integer representation for a character. No arguments |
| 8542 | are consumed. N is converted to a character by `integer->char'. N |
| 8543 | must be a positive decimal number. |
| 8544 | |
| 8545 | `~:S' |
| 8546 | Print out readproof. Prints out internal objects represented as |
| 8547 | `#<...>' as strings `"#<...>"' so that the format output can always |
| 8548 | be processed by `read'. |
| 8549 | |
| 8550 | `~:A' |
| 8551 | Print out readproof. Prints out internal objects represented as |
| 8552 | `#<...>' as strings `"#<...>"' so that the format output can always |
| 8553 | be processed by `read'. |
| 8554 | |
| 8555 | `~Q' |
| 8556 | Prints information and a copyright notice on the format |
| 8557 | implementation. |
| 8558 | `~:Q' |
| 8559 | prints format version. |
| 8560 | |
| 8561 | `~F, ~E, ~G, ~$' |
| 8562 | may also print number strings, i.e. passing a number as a string |
| 8563 | and format it accordingly. |
| 8564 | |
| 8565 | *** Configuration Variables |
| 8566 | |
| 8567 | The format module exports some configuration variables to suit the |
| 8568 | systems and users needs. There should be no modification necessary for |
| 8569 | the configuration that comes with Guile. Format detects automatically |
| 8570 | if the running scheme system implements floating point numbers and |
| 8571 | complex numbers. |
| 8572 | |
| 8573 | format:symbol-case-conv |
| 8574 | Symbols are converted by `symbol->string' so the case type of the |
| 8575 | printed symbols is implementation dependent. |
| 8576 | `format:symbol-case-conv' is a one arg closure which is either |
| 8577 | `#f' (no conversion), `string-upcase', `string-downcase' or |
| 8578 | `string-capitalize'. (default `#f') |
| 8579 | |
| 8580 | format:iobj-case-conv |
| 8581 | As FORMAT:SYMBOL-CASE-CONV but applies for the representation of |
| 8582 | implementation internal objects. (default `#f') |
| 8583 | |
| 8584 | format:expch |
| 8585 | The character prefixing the exponent value in `~E' printing. |
| 8586 | (default `#\E') |
| 8587 | |
| 8588 | *** Compatibility With Other Format Implementations |
| 8589 | |
| 8590 | SLIB format 2.x: |
| 8591 | See `format.doc'. |
| 8592 | |
| 8593 | SLIB format 1.4: |
| 8594 | Downward compatible except for padding support and `~A', `~S', |
| 8595 | `~P', `~X' uppercase printing. SLIB format 1.4 uses C-style |
| 8596 | `printf' padding support which is completely replaced by the CL |
| 8597 | `format' padding style. |
| 8598 | |
| 8599 | MIT C-Scheme 7.1: |
| 8600 | Downward compatible except for `~', which is not documented |
| 8601 | (ignores all characters inside the format string up to a newline |
| 8602 | character). (7.1 implements `~a', `~s', ~NEWLINE, `~~', `~%', |
| 8603 | numerical and variable parameters and `:/@' modifiers in the CL |
| 8604 | sense). |
| 8605 | |
| 8606 | Elk 1.5/2.0: |
| 8607 | Downward compatible except for `~A' and `~S' which print in |
| 8608 | uppercase. (Elk implements `~a', `~s', `~~', and `~%' (no |
| 8609 | directive parameters or modifiers)). |
| 8610 | |
| 8611 | Scheme->C 01nov91: |
| 8612 | Downward compatible except for an optional destination parameter: |
| 8613 | S2C accepts a format call without a destination which returns a |
| 8614 | formatted string. This is equivalent to a #f destination in S2C. |
| 8615 | (S2C implements `~a', `~s', `~c', `~%', and `~~' (no directive |
| 8616 | parameters or modifiers)). |
| 8617 | |
| 8618 | |
| 8619 | ** Changes to string-handling functions. |
| 8620 | |
| 8621 | These functions were added to support the (ice-9 format) module, above. |
| 8622 | |
| 8623 | *** New function: string-upcase STRING |
| 8624 | *** New function: string-downcase STRING |
| 8625 | |
| 8626 | These are non-destructive versions of the existing string-upcase! and |
| 8627 | string-downcase! functions. |
| 8628 | |
| 8629 | *** New function: string-capitalize! STRING |
| 8630 | *** New function: string-capitalize STRING |
| 8631 | |
| 8632 | These functions convert the first letter of each word in the string to |
| 8633 | upper case. Thus: |
| 8634 | |
| 8635 | (string-capitalize "howdy there") |
| 8636 | => "Howdy There" |
| 8637 | |
| 8638 | As with the other functions, string-capitalize! modifies the string in |
| 8639 | place, while string-capitalize returns a modified copy of its argument. |
| 8640 | |
| 8641 | *** New function: string-ci->symbol STRING |
| 8642 | |
| 8643 | Return a symbol whose name is STRING, but having the same case as if |
| 8644 | the symbol had be read by `read'. |
| 8645 | |
| 8646 | Guile can be configured to be sensitive or insensitive to case |
| 8647 | differences in Scheme identifiers. If Guile is case-insensitive, all |
| 8648 | symbols are converted to lower case on input. The `string-ci->symbol' |
| 8649 | function returns a symbol whose name in STRING, transformed as Guile |
| 8650 | would if STRING were input. |
| 8651 | |
| 8652 | *** New function: substring-move! STRING1 START END STRING2 START |
| 8653 | |
| 8654 | Copy the substring of STRING1 from START (inclusive) to END |
| 8655 | (exclusive) to STRING2 at START. STRING1 and STRING2 may be the same |
| 8656 | string, and the source and destination areas may overlap; in all |
| 8657 | cases, the function behaves as if all the characters were copied |
| 8658 | simultanously. |
| 8659 | |
| 8660 | *** Extended functions: substring-move-left! substring-move-right! |
| 8661 | |
| 8662 | These functions now correctly copy arbitrarily overlapping substrings; |
| 8663 | they are both synonyms for substring-move!. |
| 8664 | |
| 8665 | |
| 8666 | ** New module (ice-9 getopt-long), with the function `getopt-long'. |
| 8667 | |
| 8668 | getopt-long is a function for parsing command-line arguments in a |
| 8669 | manner consistent with other GNU programs. |
| 8670 | |
| 8671 | (getopt-long ARGS GRAMMAR) |
| 8672 | Parse the arguments ARGS according to the argument list grammar GRAMMAR. |
| 8673 | |
| 8674 | ARGS should be a list of strings. Its first element should be the |
| 8675 | name of the program; subsequent elements should be the arguments |
| 8676 | that were passed to the program on the command line. The |
| 8677 | `program-arguments' procedure returns a list of this form. |
| 8678 | |
| 8679 | GRAMMAR is a list of the form: |
| 8680 | ((OPTION (PROPERTY VALUE) ...) ...) |
| 8681 | |
| 8682 | Each OPTION should be a symbol. `getopt-long' will accept a |
| 8683 | command-line option named `--OPTION'. |
| 8684 | Each option can have the following (PROPERTY VALUE) pairs: |
| 8685 | |
| 8686 | (single-char CHAR) --- Accept `-CHAR' as a single-character |
| 8687 | equivalent to `--OPTION'. This is how to specify traditional |
| 8688 | Unix-style flags. |
| 8689 | (required? BOOL) --- If BOOL is true, the option is required. |
| 8690 | getopt-long will raise an error if it is not found in ARGS. |
| 8691 | (value BOOL) --- If BOOL is #t, the option accepts a value; if |
| 8692 | it is #f, it does not; and if it is the symbol |
| 8693 | `optional', the option may appear in ARGS with or |
| 8694 | without a value. |
| 8695 | (predicate FUNC) --- If the option accepts a value (i.e. you |
| 8696 | specified `(value #t)' for this option), then getopt |
| 8697 | will apply FUNC to the value, and throw an exception |
| 8698 | if it returns #f. FUNC should be a procedure which |
| 8699 | accepts a string and returns a boolean value; you may |
| 8700 | need to use quasiquotes to get it into GRAMMAR. |
| 8701 | |
| 8702 | The (PROPERTY VALUE) pairs may occur in any order, but each |
| 8703 | property may occur only once. By default, options do not have |
| 8704 | single-character equivalents, are not required, and do not take |
| 8705 | values. |
| 8706 | |
| 8707 | In ARGS, single-character options may be combined, in the usual |
| 8708 | Unix fashion: ("-x" "-y") is equivalent to ("-xy"). If an option |
| 8709 | accepts values, then it must be the last option in the |
| 8710 | combination; the value is the next argument. So, for example, using |
| 8711 | the following grammar: |
| 8712 | ((apples (single-char #\a)) |
| 8713 | (blimps (single-char #\b) (value #t)) |
| 8714 | (catalexis (single-char #\c) (value #t))) |
| 8715 | the following argument lists would be acceptable: |
| 8716 | ("-a" "-b" "bang" "-c" "couth") ("bang" and "couth" are the values |
| 8717 | for "blimps" and "catalexis") |
| 8718 | ("-ab" "bang" "-c" "couth") (same) |
| 8719 | ("-ac" "couth" "-b" "bang") (same) |
| 8720 | ("-abc" "couth" "bang") (an error, since `-b' is not the |
| 8721 | last option in its combination) |
| 8722 | |
| 8723 | If an option's value is optional, then `getopt-long' decides |
| 8724 | whether it has a value by looking at what follows it in ARGS. If |
| 8725 | the next element is a string, and it does not appear to be an |
| 8726 | option itself, then that string is the option's value. |
| 8727 | |
| 8728 | The value of a long option can appear as the next element in ARGS, |
| 8729 | or it can follow the option name, separated by an `=' character. |
| 8730 | Thus, using the same grammar as above, the following argument lists |
| 8731 | are equivalent: |
| 8732 | ("--apples" "Braeburn" "--blimps" "Goodyear") |
| 8733 | ("--apples=Braeburn" "--blimps" "Goodyear") |
| 8734 | ("--blimps" "Goodyear" "--apples=Braeburn") |
| 8735 | |
| 8736 | If the option "--" appears in ARGS, argument parsing stops there; |
| 8737 | subsequent arguments are returned as ordinary arguments, even if |
| 8738 | they resemble options. So, in the argument list: |
| 8739 | ("--apples" "Granny Smith" "--" "--blimp" "Goodyear") |
| 8740 | `getopt-long' will recognize the `apples' option as having the |
| 8741 | value "Granny Smith", but it will not recognize the `blimp' |
| 8742 | option; it will return the strings "--blimp" and "Goodyear" as |
| 8743 | ordinary argument strings. |
| 8744 | |
| 8745 | The `getopt-long' function returns the parsed argument list as an |
| 8746 | assocation list, mapping option names --- the symbols from GRAMMAR |
| 8747 | --- onto their values, or #t if the option does not accept a value. |
| 8748 | Unused options do not appear in the alist. |
| 8749 | |
| 8750 | All arguments that are not the value of any option are returned |
| 8751 | as a list, associated with the empty list. |
| 8752 | |
| 8753 | `getopt-long' throws an exception if: |
| 8754 | - it finds an unrecognized option in ARGS |
| 8755 | - a required option is omitted |
| 8756 | - an option that requires an argument doesn't get one |
| 8757 | - an option that doesn't accept an argument does get one (this can |
| 8758 | only happen using the long option `--opt=value' syntax) |
| 8759 | - an option predicate fails |
| 8760 | |
| 8761 | So, for example: |
| 8762 | |
| 8763 | (define grammar |
| 8764 | `((lockfile-dir (required? #t) |
| 8765 | (value #t) |
| 8766 | (single-char #\k) |
| 8767 | (predicate ,file-is-directory?)) |
| 8768 | (verbose (required? #f) |
| 8769 | (single-char #\v) |
| 8770 | (value #f)) |
| 8771 | (x-includes (single-char #\x)) |
| 8772 | (rnet-server (single-char #\y) |
| 8773 | (predicate ,string?)))) |
| 8774 | |
| 8775 | (getopt-long '("my-prog" "-vk" "/tmp" "foo1" "--x-includes=/usr/include" |
| 8776 | "--rnet-server=lamprod" "--" "-fred" "foo2" "foo3") |
| 8777 | grammar) |
| 8778 | => ((() "foo1" "-fred" "foo2" "foo3") |
| 8779 | (rnet-server . "lamprod") |
| 8780 | (x-includes . "/usr/include") |
| 8781 | (lockfile-dir . "/tmp") |
| 8782 | (verbose . #t)) |
| 8783 | |
| 8784 | ** The (ice-9 getopt-gnu-style) module is obsolete; use (ice-9 getopt-long). |
| 8785 | |
| 8786 | It will be removed in a few releases. |
| 8787 | |
| 8788 | ** New syntax: lambda* |
| 8789 | ** New syntax: define* |
| 8790 | ** New syntax: define*-public |
| 8791 | ** New syntax: defmacro* |
| 8792 | ** New syntax: defmacro*-public |
| 8793 | Guile now supports optional arguments. |
| 8794 | |
| 8795 | `lambda*', `define*', `define*-public', `defmacro*' and |
| 8796 | `defmacro*-public' are identical to the non-* versions except that |
| 8797 | they use an extended type of parameter list that has the following BNF |
| 8798 | syntax (parentheses are literal, square brackets indicate grouping, |
| 8799 | and `*', `+' and `?' have the usual meaning): |
| 8800 | |
| 8801 | ext-param-list ::= ( [identifier]* [#&optional [ext-var-decl]+]? |
| 8802 | [#&key [ext-var-decl]+ [#&allow-other-keys]?]? |
| 8803 | [[#&rest identifier]|[. identifier]]? ) | [identifier] |
| 8804 | |
| 8805 | ext-var-decl ::= identifier | ( identifier expression ) |
| 8806 | |
| 8807 | The semantics are best illustrated with the following documentation |
| 8808 | and examples for `lambda*': |
| 8809 | |
| 8810 | lambda* args . body |
| 8811 | lambda extended for optional and keyword arguments |
| 8812 | |
| 8813 | lambda* creates a procedure that takes optional arguments. These |
| 8814 | are specified by putting them inside brackets at the end of the |
| 8815 | paramater list, but before any dotted rest argument. For example, |
| 8816 | (lambda* (a b #&optional c d . e) '()) |
| 8817 | creates a procedure with fixed arguments a and b, optional arguments c |
| 8818 | and d, and rest argument e. If the optional arguments are omitted |
| 8819 | in a call, the variables for them are unbound in the procedure. This |
| 8820 | can be checked with the bound? macro. |
| 8821 | |
| 8822 | lambda* can also take keyword arguments. For example, a procedure |
| 8823 | defined like this: |
| 8824 | (lambda* (#&key xyzzy larch) '()) |
| 8825 | can be called with any of the argument lists (#:xyzzy 11) |
| 8826 | (#:larch 13) (#:larch 42 #:xyzzy 19) (). Whichever arguments |
| 8827 | are given as keywords are bound to values. |
| 8828 | |
| 8829 | Optional and keyword arguments can also be given default values |
| 8830 | which they take on when they are not present in a call, by giving a |
| 8831 | two-item list in place of an optional argument, for example in: |
| 8832 | (lambda* (foo #&optional (bar 42) #&key (baz 73)) (list foo bar baz)) |
| 8833 | foo is a fixed argument, bar is an optional argument with default |
| 8834 | value 42, and baz is a keyword argument with default value 73. |
| 8835 | Default value expressions are not evaluated unless they are needed |
| 8836 | and until the procedure is called. |
| 8837 | |
| 8838 | lambda* now supports two more special parameter list keywords. |
| 8839 | |
| 8840 | lambda*-defined procedures now throw an error by default if a |
| 8841 | keyword other than one of those specified is found in the actual |
| 8842 | passed arguments. However, specifying #&allow-other-keys |
| 8843 | immediately after the kyword argument declarations restores the |
| 8844 | previous behavior of ignoring unknown keywords. lambda* also now |
| 8845 | guarantees that if the same keyword is passed more than once, the |
| 8846 | last one passed is the one that takes effect. For example, |
| 8847 | ((lambda* (#&key (heads 0) (tails 0)) (display (list heads tails))) |
| 8848 | #:heads 37 #:tails 42 #:heads 99) |
| 8849 | would result in (99 47) being displayed. |
| 8850 | |
| 8851 | #&rest is also now provided as a synonym for the dotted syntax rest |
| 8852 | argument. The argument lists (a . b) and (a #&rest b) are equivalent in |
| 8853 | all respects to lambda*. This is provided for more similarity to DSSSL, |
| 8854 | MIT-Scheme and Kawa among others, as well as for refugees from other |
| 8855 | Lisp dialects. |
| 8856 | |
| 8857 | Further documentation may be found in the optargs.scm file itself. |
| 8858 | |
| 8859 | The optional argument module also exports the macros `let-optional', |
| 8860 | `let-optional*', `let-keywords', `let-keywords*' and `bound?'. These |
| 8861 | are not documented here because they may be removed in the future, but |
| 8862 | full documentation is still available in optargs.scm. |
| 8863 | |
| 8864 | ** New syntax: and-let* |
| 8865 | Guile now supports the `and-let*' form, described in the draft SRFI-2. |
| 8866 | |
| 8867 | Syntax: (land* (<clause> ...) <body> ...) |
| 8868 | Each <clause> should have one of the following forms: |
| 8869 | (<variable> <expression>) |
| 8870 | (<expression>) |
| 8871 | <bound-variable> |
| 8872 | Each <variable> or <bound-variable> should be an identifier. Each |
| 8873 | <expression> should be a valid expression. The <body> should be a |
| 8874 | possibly empty sequence of expressions, like the <body> of a |
| 8875 | lambda form. |
| 8876 | |
| 8877 | Semantics: A LAND* expression is evaluated by evaluating the |
| 8878 | <expression> or <bound-variable> of each of the <clause>s from |
| 8879 | left to right. The value of the first <expression> or |
| 8880 | <bound-variable> that evaluates to a false value is returned; the |
| 8881 | remaining <expression>s and <bound-variable>s are not evaluated. |
| 8882 | The <body> forms are evaluated iff all the <expression>s and |
| 8883 | <bound-variable>s evaluate to true values. |
| 8884 | |
| 8885 | The <expression>s and the <body> are evaluated in an environment |
| 8886 | binding each <variable> of the preceding (<variable> <expression>) |
| 8887 | clauses to the value of the <expression>. Later bindings |
| 8888 | shadow earlier bindings. |
| 8889 | |
| 8890 | Guile's and-let* macro was contributed by Michael Livshin. |
| 8891 | |
| 8892 | ** New sorting functions |
| 8893 | |
| 8894 | *** New function: sorted? SEQUENCE LESS? |
| 8895 | Returns `#t' when the sequence argument is in non-decreasing order |
| 8896 | according to LESS? (that is, there is no adjacent pair `... x y |
| 8897 | ...' for which `(less? y x)'). |
| 8898 | |
| 8899 | Returns `#f' when the sequence contains at least one out-of-order |
| 8900 | pair. It is an error if the sequence is neither a list nor a |
| 8901 | vector. |
| 8902 | |
| 8903 | *** New function: merge LIST1 LIST2 LESS? |
| 8904 | LIST1 and LIST2 are sorted lists. |
| 8905 | Returns the sorted list of all elements in LIST1 and LIST2. |
| 8906 | |
| 8907 | Assume that the elements a and b1 in LIST1 and b2 in LIST2 are "equal" |
| 8908 | in the sense that (LESS? x y) --> #f for x, y in {a, b1, b2}, |
| 8909 | and that a < b1 in LIST1. Then a < b1 < b2 in the result. |
| 8910 | (Here "<" should read "comes before".) |
| 8911 | |
| 8912 | *** New procedure: merge! LIST1 LIST2 LESS? |
| 8913 | Merges two lists, re-using the pairs of LIST1 and LIST2 to build |
| 8914 | the result. If the code is compiled, and LESS? constructs no new |
| 8915 | pairs, no pairs at all will be allocated. The first pair of the |
| 8916 | result will be either the first pair of LIST1 or the first pair of |
| 8917 | LIST2. |
| 8918 | |
| 8919 | *** New function: sort SEQUENCE LESS? |
| 8920 | Accepts either a list or a vector, and returns a new sequence |
| 8921 | which is sorted. The new sequence is the same type as the input. |
| 8922 | Always `(sorted? (sort sequence less?) less?)'. The original |
| 8923 | sequence is not altered in any way. The new sequence shares its |
| 8924 | elements with the old one; no elements are copied. |
| 8925 | |
| 8926 | *** New procedure: sort! SEQUENCE LESS |
| 8927 | Returns its sorted result in the original boxes. No new storage is |
| 8928 | allocated at all. Proper usage: (set! slist (sort! slist <)) |
| 8929 | |
| 8930 | *** New function: stable-sort SEQUENCE LESS? |
| 8931 | Similar to `sort' but stable. That is, if "equal" elements are |
| 8932 | ordered a < b in the original sequence, they will have the same order |
| 8933 | in the result. |
| 8934 | |
| 8935 | *** New function: stable-sort! SEQUENCE LESS? |
| 8936 | Similar to `sort!' but stable. |
| 8937 | Uses temporary storage when sorting vectors. |
| 8938 | |
| 8939 | *** New functions: sort-list, sort-list! |
| 8940 | Added for compatibility with scsh. |
| 8941 | |
| 8942 | ** New built-in random number support |
| 8943 | |
| 8944 | *** New function: random N [STATE] |
| 8945 | Accepts a positive integer or real N and returns a number of the |
| 8946 | same type between zero (inclusive) and N (exclusive). The values |
| 8947 | returned have a uniform distribution. |
| 8948 | |
| 8949 | The optional argument STATE must be of the type produced by |
| 8950 | `copy-random-state' or `seed->random-state'. It defaults to the value |
| 8951 | of the variable `*random-state*'. This object is used to maintain the |
| 8952 | state of the pseudo-random-number generator and is altered as a side |
| 8953 | effect of the `random' operation. |
| 8954 | |
| 8955 | *** New variable: *random-state* |
| 8956 | Holds a data structure that encodes the internal state of the |
| 8957 | random-number generator that `random' uses by default. The nature |
| 8958 | of this data structure is implementation-dependent. It may be |
| 8959 | printed out and successfully read back in, but may or may not |
| 8960 | function correctly as a random-number state object in another |
| 8961 | implementation. |
| 8962 | |
| 8963 | *** New function: copy-random-state [STATE] |
| 8964 | Returns a new object of type suitable for use as the value of the |
| 8965 | variable `*random-state*' and as a second argument to `random'. |
| 8966 | If argument STATE is given, a copy of it is returned. Otherwise a |
| 8967 | copy of `*random-state*' is returned. |
| 8968 | |
| 8969 | *** New function: seed->random-state SEED |
| 8970 | Returns a new object of type suitable for use as the value of the |
| 8971 | variable `*random-state*' and as a second argument to `random'. |
| 8972 | SEED is a string or a number. A new state is generated and |
| 8973 | initialized using SEED. |
| 8974 | |
| 8975 | *** New function: random:uniform [STATE] |
| 8976 | Returns an uniformly distributed inexact real random number in the |
| 8977 | range between 0 and 1. |
| 8978 | |
| 8979 | *** New procedure: random:solid-sphere! VECT [STATE] |
| 8980 | Fills VECT with inexact real random numbers the sum of whose |
| 8981 | squares is less than 1.0. Thinking of VECT as coordinates in |
| 8982 | space of dimension N = `(vector-length VECT)', the coordinates are |
| 8983 | uniformly distributed within the unit N-shere. The sum of the |
| 8984 | squares of the numbers is returned. VECT can be either a vector |
| 8985 | or a uniform vector of doubles. |
| 8986 | |
| 8987 | *** New procedure: random:hollow-sphere! VECT [STATE] |
| 8988 | Fills VECT with inexact real random numbers the sum of whose squares |
| 8989 | is equal to 1.0. Thinking of VECT as coordinates in space of |
| 8990 | dimension n = `(vector-length VECT)', the coordinates are uniformly |
| 8991 | distributed over the surface of the unit n-shere. VECT can be either |
| 8992 | a vector or a uniform vector of doubles. |
| 8993 | |
| 8994 | *** New function: random:normal [STATE] |
| 8995 | Returns an inexact real in a normal distribution with mean 0 and |
| 8996 | standard deviation 1. For a normal distribution with mean M and |
| 8997 | standard deviation D use `(+ M (* D (random:normal)))'. |
| 8998 | |
| 8999 | *** New procedure: random:normal-vector! VECT [STATE] |
| 9000 | Fills VECT with inexact real random numbers which are independent and |
| 9001 | standard normally distributed (i.e., with mean 0 and variance 1). |
| 9002 | VECT can be either a vector or a uniform vector of doubles. |
| 9003 | |
| 9004 | *** New function: random:exp STATE |
| 9005 | Returns an inexact real in an exponential distribution with mean 1. |
| 9006 | For an exponential distribution with mean U use (* U (random:exp)). |
| 9007 | |
| 9008 | ** The range of logand, logior, logxor, logtest, and logbit? have changed. |
| 9009 | |
| 9010 | These functions now operate on numbers in the range of a C unsigned |
| 9011 | long. |
| 9012 | |
| 9013 | These functions used to operate on numbers in the range of a C signed |
| 9014 | long; however, this seems inappropriate, because Guile integers don't |
| 9015 | overflow. |
| 9016 | |
| 9017 | ** New function: make-guardian |
| 9018 | This is an implementation of guardians as described in |
| 9019 | R. Kent Dybvig, Carl Bruggeman, and David Eby (1993) "Guardians in a |
| 9020 | Generation-Based Garbage Collector" ACM SIGPLAN Conference on |
| 9021 | Programming Language Design and Implementation, June 1993 |
| 9022 | ftp://ftp.cs.indiana.edu/pub/scheme-repository/doc/pubs/guardians.ps.gz |
| 9023 | |
| 9024 | ** New functions: delq1!, delv1!, delete1! |
| 9025 | These procedures behave similar to delq! and friends but delete only |
| 9026 | one object if at all. |
| 9027 | |
| 9028 | ** New function: unread-string STRING PORT |
| 9029 | Unread STRING to PORT, that is, push it back onto the port so that |
| 9030 | next read operation will work on the pushed back characters. |
| 9031 | |
| 9032 | ** unread-char can now be called multiple times |
| 9033 | If unread-char is called multiple times, the unread characters will be |
| 9034 | read again in last-in first-out order. |
| 9035 | |
| 9036 | ** the procedures uniform-array-read! and uniform-array-write! now |
| 9037 | work on any kind of port, not just ports which are open on a file. |
| 9038 | |
| 9039 | ** Now 'l' in a port mode requests line buffering. |
| 9040 | |
| 9041 | ** The procedure truncate-file now works on string ports as well |
| 9042 | as file ports. If the size argument is omitted, the current |
| 9043 | file position is used. |
| 9044 | |
| 9045 | ** new procedure: seek PORT/FDES OFFSET WHENCE |
| 9046 | The arguments are the same as for the old fseek procedure, but it |
| 9047 | works on string ports as well as random-access file ports. |
| 9048 | |
| 9049 | ** the fseek procedure now works on string ports, since it has been |
| 9050 | redefined using seek. |
| 9051 | |
| 9052 | ** the setvbuf procedure now uses a default size if mode is _IOFBF and |
| 9053 | size is not supplied. |
| 9054 | |
| 9055 | ** the newline procedure no longer flushes the port if it's not |
| 9056 | line-buffered: previously it did if it was the current output port. |
| 9057 | |
| 9058 | ** open-pipe and close-pipe are no longer primitive procedures, but |
| 9059 | an emulation can be obtained using `(use-modules (ice-9 popen))'. |
| 9060 | |
| 9061 | ** the freopen procedure has been removed. |
| 9062 | |
| 9063 | ** new procedure: drain-input PORT |
| 9064 | Drains PORT's read buffers (including any pushed-back characters) |
| 9065 | and returns the contents as a single string. |
| 9066 | |
| 9067 | ** New function: map-in-order PROC LIST1 LIST2 ... |
| 9068 | Version of `map' which guarantees that the procedure is applied to the |
| 9069 | lists in serial order. |
| 9070 | |
| 9071 | ** Renamed `serial-array-copy!' and `serial-array-map!' to |
| 9072 | `array-copy-in-order!' and `array-map-in-order!'. The old names are |
| 9073 | now obsolete and will go away in release 1.5. |
| 9074 | |
| 9075 | ** New syntax: collect BODY1 ... |
| 9076 | Version of `begin' which returns a list of the results of the body |
| 9077 | forms instead of the result of the last body form. In contrast to |
| 9078 | `begin', `collect' allows an empty body. |
| 9079 | |
| 9080 | ** New functions: read-history FILENAME, write-history FILENAME |
| 9081 | Read/write command line history from/to file. Returns #t on success |
| 9082 | and #f if an error occured. |
| 9083 | |
| 9084 | ** `ls' and `lls' in module (ice-9 ls) now handle no arguments. |
| 9085 | |
| 9086 | These procedures return a list of definitions available in the specified |
| 9087 | argument, a relative module reference. In the case of no argument, |
| 9088 | `(current-module)' is now consulted for definitions to return, instead |
| 9089 | of simply returning #f, the former behavior. |
| 9090 | |
| 9091 | ** The #/ syntax for lists is no longer supported. |
| 9092 | |
| 9093 | Earlier versions of Scheme accepted this syntax, but printed a |
| 9094 | warning. |
| 9095 | |
| 9096 | ** Guile no longer consults the SCHEME_LOAD_PATH environment variable. |
| 9097 | |
| 9098 | Instead, you should set GUILE_LOAD_PATH to tell Guile where to find |
| 9099 | modules. |
| 9100 | |
| 9101 | * Changes to the gh_ interface |
| 9102 | |
| 9103 | ** gh_scm2doubles |
| 9104 | |
| 9105 | Now takes a second argument which is the result array. If this |
| 9106 | pointer is NULL, a new array is malloced (the old behaviour). |
| 9107 | |
| 9108 | ** gh_chars2byvect, gh_shorts2svect, gh_floats2fvect, gh_scm2chars, |
| 9109 | gh_scm2shorts, gh_scm2longs, gh_scm2floats |
| 9110 | |
| 9111 | New functions. |
| 9112 | |
| 9113 | * Changes to the scm_ interface |
| 9114 | |
| 9115 | ** Function: scm_make_named_hook (char* name, int n_args) |
| 9116 | |
| 9117 | Creates a hook in the same way as make-hook above but also |
| 9118 | binds a variable named NAME to it. |
| 9119 | |
| 9120 | This is the typical way of creating a hook from C code. |
| 9121 | |
| 9122 | Currently, the variable is created in the "current" module. This |
| 9123 | might change when we get the new module system. |
| 9124 | |
| 9125 | ** The smob interface |
| 9126 | |
| 9127 | The interface for creating smobs has changed. For documentation, see |
| 9128 | data-rep.info (made from guile-core/doc/data-rep.texi). |
| 9129 | |
| 9130 | *** Deprecated function: SCM scm_newsmob (scm_smobfuns *) |
| 9131 | |
| 9132 | >>> This function will be removed in 1.3.4. <<< |
| 9133 | |
| 9134 | It is replaced by: |
| 9135 | |
| 9136 | *** Function: SCM scm_make_smob_type (const char *name, scm_sizet size) |
| 9137 | This function adds a new smob type, named NAME, with instance size |
| 9138 | SIZE to the system. The return value is a tag that is used in |
| 9139 | creating instances of the type. If SIZE is 0, then no memory will |
| 9140 | be allocated when instances of the smob are created, and nothing |
| 9141 | will be freed by the default free function. |
| 9142 | |
| 9143 | *** Function: void scm_set_smob_mark (long tc, SCM (*mark) (SCM)) |
| 9144 | This function sets the smob marking procedure for the smob type |
| 9145 | specified by the tag TC. TC is the tag returned by |
| 9146 | `scm_make_smob_type'. |
| 9147 | |
| 9148 | *** Function: void scm_set_smob_free (long tc, SCM (*mark) (SCM)) |
| 9149 | This function sets the smob freeing procedure for the smob type |
| 9150 | specified by the tag TC. TC is the tag returned by |
| 9151 | `scm_make_smob_type'. |
| 9152 | |
| 9153 | *** Function: void scm_set_smob_print (tc, print) |
| 9154 | |
| 9155 | - Function: void scm_set_smob_print (long tc, |
| 9156 | scm_sizet (*print) (SCM, |
| 9157 | SCM, |
| 9158 | scm_print_state *)) |
| 9159 | |
| 9160 | This function sets the smob printing procedure for the smob type |
| 9161 | specified by the tag TC. TC is the tag returned by |
| 9162 | `scm_make_smob_type'. |
| 9163 | |
| 9164 | *** Function: void scm_set_smob_equalp (long tc, SCM (*equalp) (SCM, SCM)) |
| 9165 | This function sets the smob equality-testing predicate for the |
| 9166 | smob type specified by the tag TC. TC is the tag returned by |
| 9167 | `scm_make_smob_type'. |
| 9168 | |
| 9169 | *** Macro: void SCM_NEWSMOB (SCM var, long tc, void *data) |
| 9170 | Make VALUE contain a smob instance of the type with type code TC and |
| 9171 | smob data DATA. VALUE must be previously declared as C type `SCM'. |
| 9172 | |
| 9173 | *** Macro: fn_returns SCM_RETURN_NEWSMOB (long tc, void *data) |
| 9174 | This macro expands to a block of code that creates a smob instance |
| 9175 | of the type with type code TC and smob data DATA, and returns that |
| 9176 | `SCM' value. It should be the last piece of code in a block. |
| 9177 | |
| 9178 | ** The interfaces for using I/O ports and implementing port types |
| 9179 | (ptobs) have changed significantly. The new interface is based on |
| 9180 | shared access to buffers and a new set of ptob procedures. |
| 9181 | |
| 9182 | *** scm_newptob has been removed |
| 9183 | |
| 9184 | It is replaced by: |
| 9185 | |
| 9186 | *** Function: SCM scm_make_port_type (type_name, fill_buffer, write_flush) |
| 9187 | |
| 9188 | - Function: SCM scm_make_port_type (char *type_name, |
| 9189 | int (*fill_buffer) (SCM port), |
| 9190 | void (*write_flush) (SCM port)); |
| 9191 | |
| 9192 | Similarly to the new smob interface, there is a set of function |
| 9193 | setters by which the user can customize the behaviour of his port |
| 9194 | type. See ports.h (scm_set_port_XXX). |
| 9195 | |
| 9196 | ** scm_strport_to_string: New function: creates a new string from |
| 9197 | a string port's buffer. |
| 9198 | |
| 9199 | ** Plug in interface for random number generators |
| 9200 | The variable `scm_the_rng' in random.c contains a value and three |
| 9201 | function pointers which together define the current random number |
| 9202 | generator being used by the Scheme level interface and the random |
| 9203 | number library functions. |
| 9204 | |
| 9205 | The user is free to replace the default generator with the generator |
| 9206 | of his own choice. |
| 9207 | |
| 9208 | *** Variable: size_t scm_the_rng.rstate_size |
| 9209 | The size of the random state type used by the current RNG |
| 9210 | measured in chars. |
| 9211 | |
| 9212 | *** Function: unsigned long scm_the_rng.random_bits (scm_rstate *STATE) |
| 9213 | Given the random STATE, return 32 random bits. |
| 9214 | |
| 9215 | *** Function: void scm_the_rng.init_rstate (scm_rstate *STATE, chars *S, int N) |
| 9216 | Seed random state STATE using string S of length N. |
| 9217 | |
| 9218 | *** Function: scm_rstate *scm_the_rng.copy_rstate (scm_rstate *STATE) |
| 9219 | Given random state STATE, return a malloced copy. |
| 9220 | |
| 9221 | ** Default RNG |
| 9222 | The default RNG is the MWC (Multiply With Carry) random number |
| 9223 | generator described by George Marsaglia at the Department of |
| 9224 | Statistics and Supercomputer Computations Research Institute, The |
| 9225 | Florida State University (http://stat.fsu.edu/~geo). |
| 9226 | |
| 9227 | It uses 64 bits, has a period of 4578426017172946943 (4.6e18), and |
| 9228 | passes all tests in the DIEHARD test suite |
| 9229 | (http://stat.fsu.edu/~geo/diehard.html). The generation of 32 bits |
| 9230 | costs one multiply and one add on platforms which either supports long |
| 9231 | longs (gcc does this on most systems) or have 64 bit longs. The cost |
| 9232 | is four multiply on other systems but this can be optimized by writing |
| 9233 | scm_i_uniform32 in assembler. |
| 9234 | |
| 9235 | These functions are provided through the scm_the_rng interface for use |
| 9236 | by libguile and the application. |
| 9237 | |
| 9238 | *** Function: unsigned long scm_i_uniform32 (scm_i_rstate *STATE) |
| 9239 | Given the random STATE, return 32 random bits. |
| 9240 | Don't use this function directly. Instead go through the plugin |
| 9241 | interface (see "Plug in interface" above). |
| 9242 | |
| 9243 | *** Function: void scm_i_init_rstate (scm_i_rstate *STATE, char *SEED, int N) |
| 9244 | Initialize STATE using SEED of length N. |
| 9245 | |
| 9246 | *** Function: scm_i_rstate *scm_i_copy_rstate (scm_i_rstate *STATE) |
| 9247 | Return a malloc:ed copy of STATE. This function can easily be re-used |
| 9248 | in the interfaces to other RNGs. |
| 9249 | |
| 9250 | ** Random number library functions |
| 9251 | These functions use the current RNG through the scm_the_rng interface. |
| 9252 | It might be a good idea to use these functions from your C code so |
| 9253 | that only one random generator is used by all code in your program. |
| 9254 | |
| 9255 | The default random state is stored in: |
| 9256 | |
| 9257 | *** Variable: SCM scm_var_random_state |
| 9258 | Contains the vcell of the Scheme variable "*random-state*" which is |
| 9259 | used as default state by all random number functions in the Scheme |
| 9260 | level interface. |
| 9261 | |
| 9262 | Example: |
| 9263 | |
| 9264 | double x = scm_c_uniform01 (SCM_RSTATE (SCM_CDR (scm_var_random_state))); |
| 9265 | |
| 9266 | *** Function: scm_rstate *scm_c_default_rstate (void) |
| 9267 | This is a convenience function which returns the value of |
| 9268 | scm_var_random_state. An error message is generated if this value |
| 9269 | isn't a random state. |
| 9270 | |
| 9271 | *** Function: scm_rstate *scm_c_make_rstate (char *SEED, int LENGTH) |
| 9272 | Make a new random state from the string SEED of length LENGTH. |
| 9273 | |
| 9274 | It is generally not a good idea to use multiple random states in a |
| 9275 | program. While subsequent random numbers generated from one random |
| 9276 | state are guaranteed to be reasonably independent, there is no such |
| 9277 | guarantee for numbers generated from different random states. |
| 9278 | |
| 9279 | *** Macro: unsigned long scm_c_uniform32 (scm_rstate *STATE) |
| 9280 | Return 32 random bits. |
| 9281 | |
| 9282 | *** Function: double scm_c_uniform01 (scm_rstate *STATE) |
| 9283 | Return a sample from the uniform(0,1) distribution. |
| 9284 | |
| 9285 | *** Function: double scm_c_normal01 (scm_rstate *STATE) |
| 9286 | Return a sample from the normal(0,1) distribution. |
| 9287 | |
| 9288 | *** Function: double scm_c_exp1 (scm_rstate *STATE) |
| 9289 | Return a sample from the exp(1) distribution. |
| 9290 | |
| 9291 | *** Function: unsigned long scm_c_random (scm_rstate *STATE, unsigned long M) |
| 9292 | Return a sample from the discrete uniform(0,M) distribution. |
| 9293 | |
| 9294 | *** Function: SCM scm_c_random_bignum (scm_rstate *STATE, SCM M) |
| 9295 | Return a sample from the discrete uniform(0,M) distribution. |
| 9296 | M must be a bignum object. The returned value may be an INUM. |
| 9297 | |
| 9298 | |
| 9299 | \f |
| 9300 | Changes in Guile 1.3 (released Monday, October 19, 1998): |
| 9301 | |
| 9302 | * Changes to the distribution |
| 9303 | |
| 9304 | ** We renamed the SCHEME_LOAD_PATH environment variable to GUILE_LOAD_PATH. |
| 9305 | To avoid conflicts, programs should name environment variables after |
| 9306 | themselves, except when there's a common practice establishing some |
| 9307 | other convention. |
| 9308 | |
| 9309 | For now, Guile supports both GUILE_LOAD_PATH and SCHEME_LOAD_PATH, |
| 9310 | giving the former precedence, and printing a warning message if the |
| 9311 | latter is set. Guile 1.4 will not recognize SCHEME_LOAD_PATH at all. |
| 9312 | |
| 9313 | ** The header files related to multi-byte characters have been removed. |
| 9314 | They were: libguile/extchrs.h and libguile/mbstrings.h. Any C code |
| 9315 | which referred to these explicitly will probably need to be rewritten, |
| 9316 | since the support for the variant string types has been removed; see |
| 9317 | below. |
| 9318 | |
| 9319 | ** The header files append.h and sequences.h have been removed. These |
| 9320 | files implemented non-R4RS operations which would encourage |
| 9321 | non-portable programming style and less easy-to-read code. |
| 9322 | |
| 9323 | * Changes to the stand-alone interpreter |
| 9324 | |
| 9325 | ** New procedures have been added to implement a "batch mode": |
| 9326 | |
| 9327 | *** Function: batch-mode? |
| 9328 | |
| 9329 | Returns a boolean indicating whether the interpreter is in batch |
| 9330 | mode. |
| 9331 | |
| 9332 | *** Function: set-batch-mode?! ARG |
| 9333 | |
| 9334 | If ARG is true, switches the interpreter to batch mode. The `#f' |
| 9335 | case has not been implemented. |
| 9336 | |
| 9337 | ** Guile now provides full command-line editing, when run interactively. |
| 9338 | To use this feature, you must have the readline library installed. |
| 9339 | The Guile build process will notice it, and automatically include |
| 9340 | support for it. |
| 9341 | |
| 9342 | The readline library is available via anonymous FTP from any GNU |
| 9343 | mirror site; the canonical location is "ftp://prep.ai.mit.edu/pub/gnu". |
| 9344 | |
| 9345 | ** the-last-stack is now a fluid. |
| 9346 | |
| 9347 | * Changes to the procedure for linking libguile with your programs |
| 9348 | |
| 9349 | ** You can now use the `guile-config' utility to build programs that use Guile. |
| 9350 | |
| 9351 | Guile now includes a command-line utility called `guile-config', which |
| 9352 | can provide information about how to compile and link programs that |
| 9353 | use Guile. |
| 9354 | |
| 9355 | *** `guile-config compile' prints any C compiler flags needed to use Guile. |
| 9356 | You should include this command's output on the command line you use |
| 9357 | to compile C or C++ code that #includes the Guile header files. It's |
| 9358 | usually just a `-I' flag to help the compiler find the Guile headers. |
| 9359 | |
| 9360 | |
| 9361 | *** `guile-config link' prints any linker flags necessary to link with Guile. |
| 9362 | |
| 9363 | This command writes to its standard output a list of flags which you |
| 9364 | must pass to the linker to link your code against the Guile library. |
| 9365 | The flags include '-lguile' itself, any other libraries the Guile |
| 9366 | library depends upon, and any `-L' flags needed to help the linker |
| 9367 | find those libraries. |
| 9368 | |
| 9369 | For example, here is a Makefile rule that builds a program named 'foo' |
| 9370 | from the object files ${FOO_OBJECTS}, and links them against Guile: |
| 9371 | |
| 9372 | foo: ${FOO_OBJECTS} |
| 9373 | ${CC} ${CFLAGS} ${FOO_OBJECTS} `guile-config link` -o foo |
| 9374 | |
| 9375 | Previous Guile releases recommended that you use autoconf to detect |
| 9376 | which of a predefined set of libraries were present on your system. |
| 9377 | It is more robust to use `guile-config', since it records exactly which |
| 9378 | libraries the installed Guile library requires. |
| 9379 | |
| 9380 | This was originally called `build-guile', but was renamed to |
| 9381 | `guile-config' before Guile 1.3 was released, to be consistent with |
| 9382 | the analogous script for the GTK+ GUI toolkit, which is called |
| 9383 | `gtk-config'. |
| 9384 | |
| 9385 | |
| 9386 | ** Use the GUILE_FLAGS macro in your configure.in file to find Guile. |
| 9387 | |
| 9388 | If you are using the GNU autoconf package to configure your program, |
| 9389 | you can use the GUILE_FLAGS autoconf macro to call `guile-config' |
| 9390 | (described above) and gather the necessary values for use in your |
| 9391 | Makefiles. |
| 9392 | |
| 9393 | The GUILE_FLAGS macro expands to configure script code which runs the |
| 9394 | `guile-config' script, to find out where Guile's header files and |
| 9395 | libraries are installed. It sets two variables, marked for |
| 9396 | substitution, as by AC_SUBST. |
| 9397 | |
| 9398 | GUILE_CFLAGS --- flags to pass to a C or C++ compiler to build |
| 9399 | code that uses Guile header files. This is almost always just a |
| 9400 | -I flag. |
| 9401 | |
| 9402 | GUILE_LDFLAGS --- flags to pass to the linker to link a |
| 9403 | program against Guile. This includes `-lguile' for the Guile |
| 9404 | library itself, any libraries that Guile itself requires (like |
| 9405 | -lqthreads), and so on. It may also include a -L flag to tell the |
| 9406 | compiler where to find the libraries. |
| 9407 | |
| 9408 | GUILE_FLAGS is defined in the file guile.m4, in the top-level |
| 9409 | directory of the Guile distribution. You can copy it into your |
| 9410 | package's aclocal.m4 file, and then use it in your configure.in file. |
| 9411 | |
| 9412 | If you are using the `aclocal' program, distributed with GNU automake, |
| 9413 | to maintain your aclocal.m4 file, the Guile installation process |
| 9414 | installs guile.m4 where aclocal will find it. All you need to do is |
| 9415 | use GUILE_FLAGS in your configure.in file, and then run `aclocal'; |
| 9416 | this will copy the definition of GUILE_FLAGS into your aclocal.m4 |
| 9417 | file. |
| 9418 | |
| 9419 | |
| 9420 | * Changes to Scheme functions and syntax |
| 9421 | |
| 9422 | ** Multi-byte strings have been removed, as have multi-byte and wide |
| 9423 | ports. We felt that these were the wrong approach to |
| 9424 | internationalization support. |
| 9425 | |
| 9426 | ** New function: readline [PROMPT] |
| 9427 | Read a line from the terminal, and allow the user to edit it, |
| 9428 | prompting with PROMPT. READLINE provides a large set of Emacs-like |
| 9429 | editing commands, lets the user recall previously typed lines, and |
| 9430 | works on almost every kind of terminal, including dumb terminals. |
| 9431 | |
| 9432 | READLINE assumes that the cursor is at the beginning of the line when |
| 9433 | it is invoked. Thus, you can't print a prompt yourself, and then call |
| 9434 | READLINE; you need to package up your prompt as a string, pass it to |
| 9435 | the function, and let READLINE print the prompt itself. This is |
| 9436 | because READLINE needs to know the prompt's screen width. |
| 9437 | |
| 9438 | For Guile to provide this function, you must have the readline |
| 9439 | library, version 2.1 or later, installed on your system. Readline is |
| 9440 | available via anonymous FTP from prep.ai.mit.edu in pub/gnu, or from |
| 9441 | any GNU mirror site. |
| 9442 | |
| 9443 | See also ADD-HISTORY function. |
| 9444 | |
| 9445 | ** New function: add-history STRING |
| 9446 | Add STRING as the most recent line in the history used by the READLINE |
| 9447 | command. READLINE does not add lines to the history itself; you must |
| 9448 | call ADD-HISTORY to make previous input available to the user. |
| 9449 | |
| 9450 | ** The behavior of the read-line function has changed. |
| 9451 | |
| 9452 | This function now uses standard C library functions to read the line, |
| 9453 | for speed. This means that it doesn not respect the value of |
| 9454 | scm-line-incrementors; it assumes that lines are delimited with |
| 9455 | #\newline. |
| 9456 | |
| 9457 | (Note that this is read-line, the function that reads a line of text |
| 9458 | from a port, not readline, the function that reads a line from a |
| 9459 | terminal, providing full editing capabilities.) |
| 9460 | |
| 9461 | ** New module (ice-9 getopt-gnu-style): Parse command-line arguments. |
| 9462 | |
| 9463 | This module provides some simple argument parsing. It exports one |
| 9464 | function: |
| 9465 | |
| 9466 | Function: getopt-gnu-style ARG-LS |
| 9467 | Parse a list of program arguments into an alist of option |
| 9468 | descriptions. |
| 9469 | |
| 9470 | Each item in the list of program arguments is examined to see if |
| 9471 | it meets the syntax of a GNU long-named option. An argument like |
| 9472 | `--MUMBLE' produces an element of the form (MUMBLE . #t) in the |
| 9473 | returned alist, where MUMBLE is a keyword object with the same |
| 9474 | name as the argument. An argument like `--MUMBLE=FROB' produces |
| 9475 | an element of the form (MUMBLE . FROB), where FROB is a string. |
| 9476 | |
| 9477 | As a special case, the returned alist also contains a pair whose |
| 9478 | car is the symbol `rest'. The cdr of this pair is a list |
| 9479 | containing all the items in the argument list that are not options |
| 9480 | of the form mentioned above. |
| 9481 | |
| 9482 | The argument `--' is treated specially: all items in the argument |
| 9483 | list appearing after such an argument are not examined, and are |
| 9484 | returned in the special `rest' list. |
| 9485 | |
| 9486 | This function does not parse normal single-character switches. |
| 9487 | You will need to parse them out of the `rest' list yourself. |
| 9488 | |
| 9489 | ** The read syntax for byte vectors and short vectors has changed. |
| 9490 | |
| 9491 | Instead of #bytes(...), write #y(...). |
| 9492 | |
| 9493 | Instead of #short(...), write #h(...). |
| 9494 | |
| 9495 | This may seem nutty, but, like the other uniform vectors, byte vectors |
| 9496 | and short vectors want to have the same print and read syntax (and, |
| 9497 | more basic, want to have read syntax!). Changing the read syntax to |
| 9498 | use multiple characters after the hash sign breaks with the |
| 9499 | conventions used in R5RS and the conventions used for the other |
| 9500 | uniform vectors. It also introduces complexity in the current reader, |
| 9501 | both on the C and Scheme levels. (The Right solution is probably to |
| 9502 | change the syntax and prototypes for uniform vectors entirely.) |
| 9503 | |
| 9504 | |
| 9505 | ** The new module (ice-9 session) provides useful interactive functions. |
| 9506 | |
| 9507 | *** New procedure: (apropos REGEXP OPTION ...) |
| 9508 | |
| 9509 | Display a list of top-level variables whose names match REGEXP, and |
| 9510 | the modules they are imported from. Each OPTION should be one of the |
| 9511 | following symbols: |
| 9512 | |
| 9513 | value --- Show the value of each matching variable. |
| 9514 | shadow --- Show bindings shadowed by subsequently imported modules. |
| 9515 | full --- Same as both `shadow' and `value'. |
| 9516 | |
| 9517 | For example: |
| 9518 | |
| 9519 | guile> (apropos "trace" 'full) |
| 9520 | debug: trace #<procedure trace args> |
| 9521 | debug: untrace #<procedure untrace args> |
| 9522 | the-scm-module: display-backtrace #<compiled-closure #<primitive-procedure gsubr-apply>> |
| 9523 | the-scm-module: before-backtrace-hook () |
| 9524 | the-scm-module: backtrace #<primitive-procedure backtrace> |
| 9525 | the-scm-module: after-backtrace-hook () |
| 9526 | the-scm-module: has-shown-backtrace-hint? #f |
| 9527 | guile> |
| 9528 | |
| 9529 | ** There are new functions and syntax for working with macros. |
| 9530 | |
| 9531 | Guile implements macros as a special object type. Any variable whose |
| 9532 | top-level binding is a macro object acts as a macro. The macro object |
| 9533 | specifies how the expression should be transformed before evaluation. |
| 9534 | |
| 9535 | *** Macro objects now print in a reasonable way, resembling procedures. |
| 9536 | |
| 9537 | *** New function: (macro? OBJ) |
| 9538 | True iff OBJ is a macro object. |
| 9539 | |
| 9540 | *** New function: (primitive-macro? OBJ) |
| 9541 | Like (macro? OBJ), but true only if OBJ is one of the Guile primitive |
| 9542 | macro transformers, implemented in eval.c rather than Scheme code. |
| 9543 | |
| 9544 | Why do we have this function? |
| 9545 | - For symmetry with procedure? and primitive-procedure?, |
| 9546 | - to allow custom print procedures to tell whether a macro is |
| 9547 | primitive, and display it differently, and |
| 9548 | - to allow compilers and user-written evaluators to distinguish |
| 9549 | builtin special forms from user-defined ones, which could be |
| 9550 | compiled. |
| 9551 | |
| 9552 | *** New function: (macro-type OBJ) |
| 9553 | Return a value indicating what kind of macro OBJ is. Possible return |
| 9554 | values are: |
| 9555 | |
| 9556 | The symbol `syntax' --- a macro created by procedure->syntax. |
| 9557 | The symbol `macro' --- a macro created by procedure->macro. |
| 9558 | The symbol `macro!' --- a macro created by procedure->memoizing-macro. |
| 9559 | The boolean #f --- if OBJ is not a macro object. |
| 9560 | |
| 9561 | *** New function: (macro-name MACRO) |
| 9562 | Return the name of the macro object MACRO's procedure, as returned by |
| 9563 | procedure-name. |
| 9564 | |
| 9565 | *** New function: (macro-transformer MACRO) |
| 9566 | Return the transformer procedure for MACRO. |
| 9567 | |
| 9568 | *** New syntax: (use-syntax MODULE ... TRANSFORMER) |
| 9569 | |
| 9570 | Specify a new macro expander to use in the current module. Each |
| 9571 | MODULE is a module name, with the same meaning as in the `use-modules' |
| 9572 | form; each named module's exported bindings are added to the current |
| 9573 | top-level environment. TRANSFORMER is an expression evaluated in the |
| 9574 | resulting environment which must yield a procedure to use as the |
| 9575 | module's eval transformer: every expression evaluated in this module |
| 9576 | is passed to this function, and the result passed to the Guile |
| 9577 | interpreter. |
| 9578 | |
| 9579 | *** macro-eval! is removed. Use local-eval instead. |
| 9580 | |
| 9581 | ** Some magic has been added to the printer to better handle user |
| 9582 | written printing routines (like record printers, closure printers). |
| 9583 | |
| 9584 | The problem is that these user written routines must have access to |
| 9585 | the current `print-state' to be able to handle fancy things like |
| 9586 | detection of circular references. These print-states have to be |
| 9587 | passed to the builtin printing routines (display, write, etc) to |
| 9588 | properly continue the print chain. |
| 9589 | |
| 9590 | We didn't want to change all existing print code so that it |
| 9591 | explicitly passes thru a print state in addition to a port. Instead, |
| 9592 | we extented the possible values that the builtin printing routines |
| 9593 | accept as a `port'. In addition to a normal port, they now also take |
| 9594 | a pair of a normal port and a print-state. Printing will go to the |
| 9595 | port and the print-state will be used to control the detection of |
| 9596 | circular references, etc. If the builtin function does not care for a |
| 9597 | print-state, it is simply ignored. |
| 9598 | |
| 9599 | User written callbacks are now called with such a pair as their |
| 9600 | `port', but because every function now accepts this pair as a PORT |
| 9601 | argument, you don't have to worry about that. In fact, it is probably |
| 9602 | safest to not check for these pairs. |
| 9603 | |
| 9604 | However, it is sometimes necessary to continue a print chain on a |
| 9605 | different port, for example to get a intermediate string |
| 9606 | representation of the printed value, mangle that string somehow, and |
| 9607 | then to finally print the mangled string. Use the new function |
| 9608 | |
| 9609 | inherit-print-state OLD-PORT NEW-PORT |
| 9610 | |
| 9611 | for this. It constructs a new `port' that prints to NEW-PORT but |
| 9612 | inherits the print-state of OLD-PORT. |
| 9613 | |
| 9614 | ** struct-vtable-offset renamed to vtable-offset-user |
| 9615 | |
| 9616 | ** New constants: vtable-index-layout, vtable-index-vtable, vtable-index-printer |
| 9617 | |
| 9618 | ** There is now a third optional argument to make-vtable-vtable |
| 9619 | (and fourth to make-struct) when constructing new types (vtables). |
| 9620 | This argument initializes field vtable-index-printer of the vtable. |
| 9621 | |
| 9622 | ** The detection of circular references has been extended to structs. |
| 9623 | That is, a structure that -- in the process of being printed -- prints |
| 9624 | itself does not lead to infinite recursion. |
| 9625 | |
| 9626 | ** There is now some basic support for fluids. Please read |
| 9627 | "libguile/fluid.h" to find out more. It is accessible from Scheme with |
| 9628 | the following functions and macros: |
| 9629 | |
| 9630 | Function: make-fluid |
| 9631 | |
| 9632 | Create a new fluid object. Fluids are not special variables or |
| 9633 | some other extension to the semantics of Scheme, but rather |
| 9634 | ordinary Scheme objects. You can store them into variables (that |
| 9635 | are still lexically scoped, of course) or into any other place you |
| 9636 | like. Every fluid has a initial value of `#f'. |
| 9637 | |
| 9638 | Function: fluid? OBJ |
| 9639 | |
| 9640 | Test whether OBJ is a fluid. |
| 9641 | |
| 9642 | Function: fluid-ref FLUID |
| 9643 | Function: fluid-set! FLUID VAL |
| 9644 | |
| 9645 | Access/modify the fluid FLUID. Modifications are only visible |
| 9646 | within the current dynamic root (that includes threads). |
| 9647 | |
| 9648 | Function: with-fluids* FLUIDS VALUES THUNK |
| 9649 | |
| 9650 | FLUIDS is a list of fluids and VALUES a corresponding list of |
| 9651 | values for these fluids. Before THUNK gets called the values are |
| 9652 | installed in the fluids and the old values of the fluids are |
| 9653 | saved in the VALUES list. When the flow of control leaves THUNK |
| 9654 | or reenters it, the values get swapped again. You might think of |
| 9655 | this as a `safe-fluid-excursion'. Note that the VALUES list is |
| 9656 | modified by `with-fluids*'. |
| 9657 | |
| 9658 | Macro: with-fluids ((FLUID VALUE) ...) FORM ... |
| 9659 | |
| 9660 | The same as `with-fluids*' but with a different syntax. It looks |
| 9661 | just like `let', but both FLUID and VALUE are evaluated. Remember, |
| 9662 | fluids are not special variables but ordinary objects. FLUID |
| 9663 | should evaluate to a fluid. |
| 9664 | |
| 9665 | ** Changes to system call interfaces: |
| 9666 | |
| 9667 | *** close-port, close-input-port and close-output-port now return a |
| 9668 | boolean instead of an `unspecified' object. #t means that the port |
| 9669 | was successfully closed, while #f means it was already closed. It is |
| 9670 | also now possible for these procedures to raise an exception if an |
| 9671 | error occurs (some errors from write can be delayed until close.) |
| 9672 | |
| 9673 | *** the first argument to chmod, fcntl, ftell and fseek can now be a |
| 9674 | file descriptor. |
| 9675 | |
| 9676 | *** the third argument to fcntl is now optional. |
| 9677 | |
| 9678 | *** the first argument to chown can now be a file descriptor or a port. |
| 9679 | |
| 9680 | *** the argument to stat can now be a port. |
| 9681 | |
| 9682 | *** The following new procedures have been added (most use scsh |
| 9683 | interfaces): |
| 9684 | |
| 9685 | *** procedure: close PORT/FD |
| 9686 | Similar to close-port (*note close-port: Closing Ports.), but also |
| 9687 | works on file descriptors. A side effect of closing a file |
| 9688 | descriptor is that any ports using that file descriptor are moved |
| 9689 | to a different file descriptor and have their revealed counts set |
| 9690 | to zero. |
| 9691 | |
| 9692 | *** procedure: port->fdes PORT |
| 9693 | Returns the integer file descriptor underlying PORT. As a side |
| 9694 | effect the revealed count of PORT is incremented. |
| 9695 | |
| 9696 | *** procedure: fdes->ports FDES |
| 9697 | Returns a list of existing ports which have FDES as an underlying |
| 9698 | file descriptor, without changing their revealed counts. |
| 9699 | |
| 9700 | *** procedure: fdes->inport FDES |
| 9701 | Returns an existing input port which has FDES as its underlying |
| 9702 | file descriptor, if one exists, and increments its revealed count. |
| 9703 | Otherwise, returns a new input port with a revealed count of 1. |
| 9704 | |
| 9705 | *** procedure: fdes->outport FDES |
| 9706 | Returns an existing output port which has FDES as its underlying |
| 9707 | file descriptor, if one exists, and increments its revealed count. |
| 9708 | Otherwise, returns a new output port with a revealed count of 1. |
| 9709 | |
| 9710 | The next group of procedures perform a `dup2' system call, if NEWFD |
| 9711 | (an integer) is supplied, otherwise a `dup'. The file descriptor to be |
| 9712 | duplicated can be supplied as an integer or contained in a port. The |
| 9713 | type of value returned varies depending on which procedure is used. |
| 9714 | |
| 9715 | All procedures also have the side effect when performing `dup2' that |
| 9716 | any ports using NEWFD are moved to a different file descriptor and have |
| 9717 | their revealed counts set to zero. |
| 9718 | |
| 9719 | *** procedure: dup->fdes PORT/FD [NEWFD] |
| 9720 | Returns an integer file descriptor. |
| 9721 | |
| 9722 | *** procedure: dup->inport PORT/FD [NEWFD] |
| 9723 | Returns a new input port using the new file descriptor. |
| 9724 | |
| 9725 | *** procedure: dup->outport PORT/FD [NEWFD] |
| 9726 | Returns a new output port using the new file descriptor. |
| 9727 | |
| 9728 | *** procedure: dup PORT/FD [NEWFD] |
| 9729 | Returns a new port if PORT/FD is a port, with the same mode as the |
| 9730 | supplied port, otherwise returns an integer file descriptor. |
| 9731 | |
| 9732 | *** procedure: dup->port PORT/FD MODE [NEWFD] |
| 9733 | Returns a new port using the new file descriptor. MODE supplies a |
| 9734 | mode string for the port (*note open-file: File Ports.). |
| 9735 | |
| 9736 | *** procedure: setenv NAME VALUE |
| 9737 | Modifies the environment of the current process, which is also the |
| 9738 | default environment inherited by child processes. |
| 9739 | |
| 9740 | If VALUE is `#f', then NAME is removed from the environment. |
| 9741 | Otherwise, the string NAME=VALUE is added to the environment, |
| 9742 | replacing any existing string with name matching NAME. |
| 9743 | |
| 9744 | The return value is unspecified. |
| 9745 | |
| 9746 | *** procedure: truncate-file OBJ SIZE |
| 9747 | Truncates the file referred to by OBJ to at most SIZE bytes. OBJ |
| 9748 | can be a string containing a file name or an integer file |
| 9749 | descriptor or port open for output on the file. The underlying |
| 9750 | system calls are `truncate' and `ftruncate'. |
| 9751 | |
| 9752 | The return value is unspecified. |
| 9753 | |
| 9754 | *** procedure: setvbuf PORT MODE [SIZE] |
| 9755 | Set the buffering mode for PORT. MODE can be: |
| 9756 | `_IONBF' |
| 9757 | non-buffered |
| 9758 | |
| 9759 | `_IOLBF' |
| 9760 | line buffered |
| 9761 | |
| 9762 | `_IOFBF' |
| 9763 | block buffered, using a newly allocated buffer of SIZE bytes. |
| 9764 | However if SIZE is zero or unspecified, the port will be made |
| 9765 | non-buffered. |
| 9766 | |
| 9767 | This procedure should not be used after I/O has been performed with |
| 9768 | the port. |
| 9769 | |
| 9770 | Ports are usually block buffered by default, with a default buffer |
| 9771 | size. Procedures e.g., *Note open-file: File Ports, which accept a |
| 9772 | mode string allow `0' to be added to request an unbuffered port. |
| 9773 | |
| 9774 | *** procedure: fsync PORT/FD |
| 9775 | Copies any unwritten data for the specified output file descriptor |
| 9776 | to disk. If PORT/FD is a port, its buffer is flushed before the |
| 9777 | underlying file descriptor is fsync'd. The return value is |
| 9778 | unspecified. |
| 9779 | |
| 9780 | *** procedure: open-fdes PATH FLAGS [MODES] |
| 9781 | Similar to `open' but returns a file descriptor instead of a port. |
| 9782 | |
| 9783 | *** procedure: execle PATH ENV [ARG] ... |
| 9784 | Similar to `execl', but the environment of the new process is |
| 9785 | specified by ENV, which must be a list of strings as returned by |
| 9786 | the `environ' procedure. |
| 9787 | |
| 9788 | This procedure is currently implemented using the `execve' system |
| 9789 | call, but we call it `execle' because of its Scheme calling |
| 9790 | interface. |
| 9791 | |
| 9792 | *** procedure: strerror ERRNO |
| 9793 | Returns the Unix error message corresponding to ERRNO, an integer. |
| 9794 | |
| 9795 | *** procedure: primitive-exit [STATUS] |
| 9796 | Terminate the current process without unwinding the Scheme stack. |
| 9797 | This is would typically be useful after a fork. The exit status |
| 9798 | is STATUS if supplied, otherwise zero. |
| 9799 | |
| 9800 | *** procedure: times |
| 9801 | Returns an object with information about real and processor time. |
| 9802 | The following procedures accept such an object as an argument and |
| 9803 | return a selected component: |
| 9804 | |
| 9805 | `tms:clock' |
| 9806 | The current real time, expressed as time units relative to an |
| 9807 | arbitrary base. |
| 9808 | |
| 9809 | `tms:utime' |
| 9810 | The CPU time units used by the calling process. |
| 9811 | |
| 9812 | `tms:stime' |
| 9813 | The CPU time units used by the system on behalf of the |
| 9814 | calling process. |
| 9815 | |
| 9816 | `tms:cutime' |
| 9817 | The CPU time units used by terminated child processes of the |
| 9818 | calling process, whose status has been collected (e.g., using |
| 9819 | `waitpid'). |
| 9820 | |
| 9821 | `tms:cstime' |
| 9822 | Similarly, the CPU times units used by the system on behalf of |
| 9823 | terminated child processes. |
| 9824 | |
| 9825 | ** Removed: list-length |
| 9826 | ** Removed: list-append, list-append! |
| 9827 | ** Removed: list-reverse, list-reverse! |
| 9828 | |
| 9829 | ** array-map renamed to array-map! |
| 9830 | |
| 9831 | ** serial-array-map renamed to serial-array-map! |
| 9832 | |
| 9833 | ** catch doesn't take #f as first argument any longer |
| 9834 | |
| 9835 | Previously, it was possible to pass #f instead of a key to `catch'. |
| 9836 | That would cause `catch' to pass a jump buffer object to the procedure |
| 9837 | passed as second argument. The procedure could then use this jump |
| 9838 | buffer objekt as an argument to throw. |
| 9839 | |
| 9840 | This mechanism has been removed since its utility doesn't motivate the |
| 9841 | extra complexity it introduces. |
| 9842 | |
| 9843 | ** The `#/' notation for lists now provokes a warning message from Guile. |
| 9844 | This syntax will be removed from Guile in the near future. |
| 9845 | |
| 9846 | To disable the warning message, set the GUILE_HUSH environment |
| 9847 | variable to any non-empty value. |
| 9848 | |
| 9849 | ** The newline character now prints as `#\newline', following the |
| 9850 | normal Scheme notation, not `#\nl'. |
| 9851 | |
| 9852 | * Changes to the gh_ interface |
| 9853 | |
| 9854 | ** The gh_enter function now takes care of loading the Guile startup files. |
| 9855 | gh_enter works by calling scm_boot_guile; see the remarks below. |
| 9856 | |
| 9857 | ** Function: void gh_write (SCM x) |
| 9858 | |
| 9859 | Write the printed representation of the scheme object x to the current |
| 9860 | output port. Corresponds to the scheme level `write'. |
| 9861 | |
| 9862 | ** gh_list_length renamed to gh_length. |
| 9863 | |
| 9864 | ** vector handling routines |
| 9865 | |
| 9866 | Several major changes. In particular, gh_vector() now resembles |
| 9867 | (vector ...) (with a caveat -- see manual), and gh_make_vector() now |
| 9868 | exists and behaves like (make-vector ...). gh_vset() and gh_vref() |
| 9869 | have been renamed gh_vector_set_x() and gh_vector_ref(). Some missing |
| 9870 | vector-related gh_ functions have been implemented. |
| 9871 | |
| 9872 | ** pair and list routines |
| 9873 | |
| 9874 | Implemented several of the R4RS pair and list functions that were |
| 9875 | missing. |
| 9876 | |
| 9877 | ** gh_scm2doubles, gh_doubles2scm, gh_doubles2dvect |
| 9878 | |
| 9879 | New function. Converts double arrays back and forth between Scheme |
| 9880 | and C. |
| 9881 | |
| 9882 | * Changes to the scm_ interface |
| 9883 | |
| 9884 | ** The function scm_boot_guile now takes care of loading the startup files. |
| 9885 | |
| 9886 | Guile's primary initialization function, scm_boot_guile, now takes |
| 9887 | care of loading `boot-9.scm', in the `ice-9' module, to initialize |
| 9888 | Guile, define the module system, and put together some standard |
| 9889 | bindings. It also loads `init.scm', which is intended to hold |
| 9890 | site-specific initialization code. |
| 9891 | |
| 9892 | Since Guile cannot operate properly until boot-9.scm is loaded, there |
| 9893 | is no reason to separate loading boot-9.scm from Guile's other |
| 9894 | initialization processes. |
| 9895 | |
| 9896 | This job used to be done by scm_compile_shell_switches, which didn't |
| 9897 | make much sense; in particular, it meant that people using Guile for |
| 9898 | non-shell-like applications had to jump through hoops to get Guile |
| 9899 | initialized properly. |
| 9900 | |
| 9901 | ** The function scm_compile_shell_switches no longer loads the startup files. |
| 9902 | Now, Guile always loads the startup files, whenever it is initialized; |
| 9903 | see the notes above for scm_boot_guile and scm_load_startup_files. |
| 9904 | |
| 9905 | ** Function: scm_load_startup_files |
| 9906 | This new function takes care of loading Guile's initialization file |
| 9907 | (`boot-9.scm'), and the site initialization file, `init.scm'. Since |
| 9908 | this is always called by the Guile initialization process, it's |
| 9909 | probably not too useful to call this yourself, but it's there anyway. |
| 9910 | |
| 9911 | ** The semantics of smob marking have changed slightly. |
| 9912 | |
| 9913 | The smob marking function (the `mark' member of the scm_smobfuns |
| 9914 | structure) is no longer responsible for setting the mark bit on the |
| 9915 | smob. The generic smob handling code in the garbage collector will |
| 9916 | set this bit. The mark function need only ensure that any other |
| 9917 | objects the smob refers to get marked. |
| 9918 | |
| 9919 | Note that this change means that the smob's GC8MARK bit is typically |
| 9920 | already set upon entry to the mark function. Thus, marking functions |
| 9921 | which look like this: |
| 9922 | |
| 9923 | { |
| 9924 | if (SCM_GC8MARKP (ptr)) |
| 9925 | return SCM_BOOL_F; |
| 9926 | SCM_SETGC8MARK (ptr); |
| 9927 | ... mark objects to which the smob refers ... |
| 9928 | } |
| 9929 | |
| 9930 | are now incorrect, since they will return early, and fail to mark any |
| 9931 | other objects the smob refers to. Some code in the Guile library used |
| 9932 | to work this way. |
| 9933 | |
| 9934 | ** The semantics of the I/O port functions in scm_ptobfuns have changed. |
| 9935 | |
| 9936 | If you have implemented your own I/O port type, by writing the |
| 9937 | functions required by the scm_ptobfuns and then calling scm_newptob, |
| 9938 | you will need to change your functions slightly. |
| 9939 | |
| 9940 | The functions in a scm_ptobfuns structure now expect the port itself |
| 9941 | as their argument; they used to expect the `stream' member of the |
| 9942 | port's scm_port_table structure. This allows functions in an |
| 9943 | scm_ptobfuns structure to easily access the port's cell (and any flags |
| 9944 | it its CAR), and the port's scm_port_table structure. |
| 9945 | |
| 9946 | Guile now passes the I/O port itself as the `port' argument in the |
| 9947 | following scm_ptobfuns functions: |
| 9948 | |
| 9949 | int (*free) (SCM port); |
| 9950 | int (*fputc) (int, SCM port); |
| 9951 | int (*fputs) (char *, SCM port); |
| 9952 | scm_sizet (*fwrite) SCM_P ((char *ptr, |
| 9953 | scm_sizet size, |
| 9954 | scm_sizet nitems, |
| 9955 | SCM port)); |
| 9956 | int (*fflush) (SCM port); |
| 9957 | int (*fgetc) (SCM port); |
| 9958 | int (*fclose) (SCM port); |
| 9959 | |
| 9960 | The interfaces to the `mark', `print', `equalp', and `fgets' methods |
| 9961 | are unchanged. |
| 9962 | |
| 9963 | If you have existing code which defines its own port types, it is easy |
| 9964 | to convert your code to the new interface; simply apply SCM_STREAM to |
| 9965 | the port argument to yield the value you code used to expect. |
| 9966 | |
| 9967 | Note that since both the port and the stream have the same type in the |
| 9968 | C code --- they are both SCM values --- the C compiler will not remind |
| 9969 | you if you forget to update your scm_ptobfuns functions. |
| 9970 | |
| 9971 | |
| 9972 | ** Function: int scm_internal_select (int fds, |
| 9973 | SELECT_TYPE *rfds, |
| 9974 | SELECT_TYPE *wfds, |
| 9975 | SELECT_TYPE *efds, |
| 9976 | struct timeval *timeout); |
| 9977 | |
| 9978 | This is a replacement for the `select' function provided by the OS. |
| 9979 | It enables I/O blocking and sleeping to happen for one cooperative |
| 9980 | thread without blocking other threads. It also avoids busy-loops in |
| 9981 | these situations. It is intended that all I/O blocking and sleeping |
| 9982 | will finally go through this function. Currently, this function is |
| 9983 | only available on systems providing `gettimeofday' and `select'. |
| 9984 | |
| 9985 | ** Function: SCM scm_internal_stack_catch (SCM tag, |
| 9986 | scm_catch_body_t body, |
| 9987 | void *body_data, |
| 9988 | scm_catch_handler_t handler, |
| 9989 | void *handler_data) |
| 9990 | |
| 9991 | A new sibling to the other two C level `catch' functions |
| 9992 | scm_internal_catch and scm_internal_lazy_catch. Use it if you want |
| 9993 | the stack to be saved automatically into the variable `the-last-stack' |
| 9994 | (scm_the_last_stack_var) on error. This is necessary if you want to |
| 9995 | use advanced error reporting, such as calling scm_display_error and |
| 9996 | scm_display_backtrace. (They both take a stack object as argument.) |
| 9997 | |
| 9998 | ** Function: SCM scm_spawn_thread (scm_catch_body_t body, |
| 9999 | void *body_data, |
| 10000 | scm_catch_handler_t handler, |
| 10001 | void *handler_data) |
| 10002 | |
| 10003 | Spawns a new thread. It does a job similar to |
| 10004 | scm_call_with_new_thread but takes arguments more suitable when |
| 10005 | spawning threads from application C code. |
| 10006 | |
| 10007 | ** The hook scm_error_callback has been removed. It was originally |
| 10008 | intended as a way for the user to install his own error handler. But |
| 10009 | that method works badly since it intervenes between throw and catch, |
| 10010 | thereby changing the semantics of expressions like (catch #t ...). |
| 10011 | The correct way to do it is to use one of the C level catch functions |
| 10012 | in throw.c: scm_internal_catch/lazy_catch/stack_catch. |
| 10013 | |
| 10014 | ** Removed functions: |
| 10015 | |
| 10016 | scm_obj_length, scm_list_length, scm_list_append, scm_list_append_x, |
| 10017 | scm_list_reverse, scm_list_reverse_x |
| 10018 | |
| 10019 | ** New macros: SCM_LISTn where n is one of the integers 0-9. |
| 10020 | |
| 10021 | These can be used for pretty list creation from C. The idea is taken |
| 10022 | from Erick Gallesio's STk. |
| 10023 | |
| 10024 | ** scm_array_map renamed to scm_array_map_x |
| 10025 | |
| 10026 | ** mbstrings are now removed |
| 10027 | |
| 10028 | This means that the type codes scm_tc7_mb_string and |
| 10029 | scm_tc7_mb_substring has been removed. |
| 10030 | |
| 10031 | ** scm_gen_putc, scm_gen_puts, scm_gen_write, and scm_gen_getc have changed. |
| 10032 | |
| 10033 | Since we no longer support multi-byte strings, these I/O functions |
| 10034 | have been simplified, and renamed. Here are their old names, and |
| 10035 | their new names and arguments: |
| 10036 | |
| 10037 | scm_gen_putc -> void scm_putc (int c, SCM port); |
| 10038 | scm_gen_puts -> void scm_puts (char *s, SCM port); |
| 10039 | scm_gen_write -> void scm_lfwrite (char *ptr, scm_sizet size, SCM port); |
| 10040 | scm_gen_getc -> void scm_getc (SCM port); |
| 10041 | |
| 10042 | |
| 10043 | ** The macros SCM_TYP7D and SCM_TYP7SD has been removed. |
| 10044 | |
| 10045 | ** The macro SCM_TYP7S has taken the role of the old SCM_TYP7D |
| 10046 | |
| 10047 | SCM_TYP7S now masks away the bit which distinguishes substrings from |
| 10048 | strings. |
| 10049 | |
| 10050 | ** scm_catch_body_t: Backward incompatible change! |
| 10051 | |
| 10052 | Body functions to scm_internal_catch and friends do not any longer |
| 10053 | take a second argument. This is because it is no longer possible to |
| 10054 | pass a #f arg to catch. |
| 10055 | |
| 10056 | ** Calls to scm_protect_object and scm_unprotect now nest properly. |
| 10057 | |
| 10058 | The function scm_protect_object protects its argument from being freed |
| 10059 | by the garbage collector. scm_unprotect_object removes that |
| 10060 | protection. |
| 10061 | |
| 10062 | These functions now nest properly. That is, for every object O, there |
| 10063 | is a counter which scm_protect_object(O) increments and |
| 10064 | scm_unprotect_object(O) decrements, if the counter is greater than |
| 10065 | zero. Every object's counter is zero when it is first created. If an |
| 10066 | object's counter is greater than zero, the garbage collector will not |
| 10067 | reclaim its storage. |
| 10068 | |
| 10069 | This allows you to use scm_protect_object in your code without |
| 10070 | worrying that some other function you call will call |
| 10071 | scm_unprotect_object, and allow it to be freed. Assuming that the |
| 10072 | functions you call are well-behaved, and unprotect only those objects |
| 10073 | they protect, you can follow the same rule and have confidence that |
| 10074 | objects will be freed only at appropriate times. |
| 10075 | |
| 10076 | \f |
| 10077 | Changes in Guile 1.2 (released Tuesday, June 24 1997): |
| 10078 | |
| 10079 | * Changes to the distribution |
| 10080 | |
| 10081 | ** Nightly snapshots are now available from ftp.red-bean.com. |
| 10082 | The old server, ftp.cyclic.com, has been relinquished to its rightful |
| 10083 | owner. |
| 10084 | |
| 10085 | Nightly snapshots of the Guile development sources are now available via |
| 10086 | anonymous FTP from ftp.red-bean.com, as /pub/guile/guile-snap.tar.gz. |
| 10087 | |
| 10088 | Via the web, that's: ftp://ftp.red-bean.com/pub/guile/guile-snap.tar.gz |
| 10089 | For getit, that's: ftp.red-bean.com:/pub/guile/guile-snap.tar.gz |
| 10090 | |
| 10091 | ** To run Guile without installing it, the procedure has changed a bit. |
| 10092 | |
| 10093 | If you used a separate build directory to compile Guile, you'll need |
| 10094 | to include the build directory in SCHEME_LOAD_PATH, as well as the |
| 10095 | source directory. See the `INSTALL' file for examples. |
| 10096 | |
| 10097 | * Changes to the procedure for linking libguile with your programs |
| 10098 | |
| 10099 | ** The standard Guile load path for Scheme code now includes |
| 10100 | $(datadir)/guile (usually /usr/local/share/guile). This means that |
| 10101 | you can install your own Scheme files there, and Guile will find them. |
| 10102 | (Previous versions of Guile only checked a directory whose name |
| 10103 | contained the Guile version number, so you had to re-install or move |
| 10104 | your Scheme sources each time you installed a fresh version of Guile.) |
| 10105 | |
| 10106 | The load path also includes $(datadir)/guile/site; we recommend |
| 10107 | putting individual Scheme files there. If you want to install a |
| 10108 | package with multiple source files, create a directory for them under |
| 10109 | $(datadir)/guile. |
| 10110 | |
| 10111 | ** Guile 1.2 will now use the Rx regular expression library, if it is |
| 10112 | installed on your system. When you are linking libguile into your own |
| 10113 | programs, this means you will have to link against -lguile, -lqt (if |
| 10114 | you configured Guile with thread support), and -lrx. |
| 10115 | |
| 10116 | If you are using autoconf to generate configuration scripts for your |
| 10117 | application, the following lines should suffice to add the appropriate |
| 10118 | libraries to your link command: |
| 10119 | |
| 10120 | ### Find Rx, quickthreads and libguile. |
| 10121 | AC_CHECK_LIB(rx, main) |
| 10122 | AC_CHECK_LIB(qt, main) |
| 10123 | AC_CHECK_LIB(guile, scm_shell) |
| 10124 | |
| 10125 | The Guile 1.2 distribution does not contain sources for the Rx |
| 10126 | library, as Guile 1.0 did. If you want to use Rx, you'll need to |
| 10127 | retrieve it from a GNU FTP site and install it separately. |
| 10128 | |
| 10129 | * Changes to Scheme functions and syntax |
| 10130 | |
| 10131 | ** The dynamic linking features of Guile are now enabled by default. |
| 10132 | You can disable them by giving the `--disable-dynamic-linking' option |
| 10133 | to configure. |
| 10134 | |
| 10135 | (dynamic-link FILENAME) |
| 10136 | |
| 10137 | Find the object file denoted by FILENAME (a string) and link it |
| 10138 | into the running Guile application. When everything works out, |
| 10139 | return a Scheme object suitable for representing the linked object |
| 10140 | file. Otherwise an error is thrown. How object files are |
| 10141 | searched is system dependent. |
| 10142 | |
| 10143 | (dynamic-object? VAL) |
| 10144 | |
| 10145 | Determine whether VAL represents a dynamically linked object file. |
| 10146 | |
| 10147 | (dynamic-unlink DYNOBJ) |
| 10148 | |
| 10149 | Unlink the indicated object file from the application. DYNOBJ |
| 10150 | should be one of the values returned by `dynamic-link'. |
| 10151 | |
| 10152 | (dynamic-func FUNCTION DYNOBJ) |
| 10153 | |
| 10154 | Search the C function indicated by FUNCTION (a string or symbol) |
| 10155 | in DYNOBJ and return some Scheme object that can later be used |
| 10156 | with `dynamic-call' to actually call this function. Right now, |
| 10157 | these Scheme objects are formed by casting the address of the |
| 10158 | function to `long' and converting this number to its Scheme |
| 10159 | representation. |
| 10160 | |
| 10161 | (dynamic-call FUNCTION DYNOBJ) |
| 10162 | |
| 10163 | Call the C function indicated by FUNCTION and DYNOBJ. The |
| 10164 | function is passed no arguments and its return value is ignored. |
| 10165 | When FUNCTION is something returned by `dynamic-func', call that |
| 10166 | function and ignore DYNOBJ. When FUNCTION is a string (or symbol, |
| 10167 | etc.), look it up in DYNOBJ; this is equivalent to |
| 10168 | |
| 10169 | (dynamic-call (dynamic-func FUNCTION DYNOBJ) #f) |
| 10170 | |
| 10171 | Interrupts are deferred while the C function is executing (with |
| 10172 | SCM_DEFER_INTS/SCM_ALLOW_INTS). |
| 10173 | |
| 10174 | (dynamic-args-call FUNCTION DYNOBJ ARGS) |
| 10175 | |
| 10176 | Call the C function indicated by FUNCTION and DYNOBJ, but pass it |
| 10177 | some arguments and return its return value. The C function is |
| 10178 | expected to take two arguments and return an `int', just like |
| 10179 | `main': |
| 10180 | |
| 10181 | int c_func (int argc, char **argv); |
| 10182 | |
| 10183 | ARGS must be a list of strings and is converted into an array of |
| 10184 | `char *'. The array is passed in ARGV and its size in ARGC. The |
| 10185 | return value is converted to a Scheme number and returned from the |
| 10186 | call to `dynamic-args-call'. |
| 10187 | |
| 10188 | When dynamic linking is disabled or not supported on your system, |
| 10189 | the above functions throw errors, but they are still available. |
| 10190 | |
| 10191 | Here is a small example that works on GNU/Linux: |
| 10192 | |
| 10193 | (define libc-obj (dynamic-link "libc.so")) |
| 10194 | (dynamic-args-call 'rand libc-obj '()) |
| 10195 | |
| 10196 | See the file `libguile/DYNAMIC-LINKING' for additional comments. |
| 10197 | |
| 10198 | ** The #/ syntax for module names is depreciated, and will be removed |
| 10199 | in a future version of Guile. Instead of |
| 10200 | |
| 10201 | #/foo/bar/baz |
| 10202 | |
| 10203 | instead write |
| 10204 | |
| 10205 | (foo bar baz) |
| 10206 | |
| 10207 | The latter syntax is more consistent with existing Lisp practice. |
| 10208 | |
| 10209 | ** Guile now does fancier printing of structures. Structures are the |
| 10210 | underlying implementation for records, which in turn are used to |
| 10211 | implement modules, so all of these object now print differently and in |
| 10212 | a more informative way. |
| 10213 | |
| 10214 | The Scheme printer will examine the builtin variable *struct-printer* |
| 10215 | whenever it needs to print a structure object. When this variable is |
| 10216 | not `#f' it is deemed to be a procedure and will be applied to the |
| 10217 | structure object and the output port. When *struct-printer* is `#f' |
| 10218 | or the procedure return `#f' the structure object will be printed in |
| 10219 | the boring #<struct 80458270> form. |
| 10220 | |
| 10221 | This hook is used by some routines in ice-9/boot-9.scm to implement |
| 10222 | type specific printing routines. Please read the comments there about |
| 10223 | "printing structs". |
| 10224 | |
| 10225 | One of the more specific uses of structs are records. The printing |
| 10226 | procedure that could be passed to MAKE-RECORD-TYPE is now actually |
| 10227 | called. It should behave like a *struct-printer* procedure (described |
| 10228 | above). |
| 10229 | |
| 10230 | ** Guile now supports a new R4RS-compliant syntax for keywords. A |
| 10231 | token of the form #:NAME, where NAME has the same syntax as a Scheme |
| 10232 | symbol, is the external representation of the keyword named NAME. |
| 10233 | Keyword objects print using this syntax as well, so values containing |
| 10234 | keyword objects can be read back into Guile. When used in an |
| 10235 | expression, keywords are self-quoting objects. |
| 10236 | |
| 10237 | Guile suports this read syntax, and uses this print syntax, regardless |
| 10238 | of the current setting of the `keyword' read option. The `keyword' |
| 10239 | read option only controls whether Guile recognizes the `:NAME' syntax, |
| 10240 | which is incompatible with R4RS. (R4RS says such token represent |
| 10241 | symbols.) |
| 10242 | |
| 10243 | ** Guile has regular expression support again. Guile 1.0 included |
| 10244 | functions for matching regular expressions, based on the Rx library. |
| 10245 | In Guile 1.1, the Guile/Rx interface was removed to simplify the |
| 10246 | distribution, and thus Guile had no regular expression support. Guile |
| 10247 | 1.2 again supports the most commonly used functions, and supports all |
| 10248 | of SCSH's regular expression functions. |
| 10249 | |
| 10250 | If your system does not include a POSIX regular expression library, |
| 10251 | and you have not linked Guile with a third-party regexp library such as |
| 10252 | Rx, these functions will not be available. You can tell whether your |
| 10253 | Guile installation includes regular expression support by checking |
| 10254 | whether the `*features*' list includes the `regex' symbol. |
| 10255 | |
| 10256 | *** regexp functions |
| 10257 | |
| 10258 | By default, Guile supports POSIX extended regular expressions. That |
| 10259 | means that the characters `(', `)', `+' and `?' are special, and must |
| 10260 | be escaped if you wish to match the literal characters. |
| 10261 | |
| 10262 | This regular expression interface was modeled after that implemented |
| 10263 | by SCSH, the Scheme Shell. It is intended to be upwardly compatible |
| 10264 | with SCSH regular expressions. |
| 10265 | |
| 10266 | **** Function: string-match PATTERN STR [START] |
| 10267 | Compile the string PATTERN into a regular expression and compare |
| 10268 | it with STR. The optional numeric argument START specifies the |
| 10269 | position of STR at which to begin matching. |
| 10270 | |
| 10271 | `string-match' returns a "match structure" which describes what, |
| 10272 | if anything, was matched by the regular expression. *Note Match |
| 10273 | Structures::. If STR does not match PATTERN at all, |
| 10274 | `string-match' returns `#f'. |
| 10275 | |
| 10276 | Each time `string-match' is called, it must compile its PATTERN |
| 10277 | argument into a regular expression structure. This operation is |
| 10278 | expensive, which makes `string-match' inefficient if the same regular |
| 10279 | expression is used several times (for example, in a loop). For better |
| 10280 | performance, you can compile a regular expression in advance and then |
| 10281 | match strings against the compiled regexp. |
| 10282 | |
| 10283 | **** Function: make-regexp STR [FLAGS] |
| 10284 | Compile the regular expression described by STR, and return the |
| 10285 | compiled regexp structure. If STR does not describe a legal |
| 10286 | regular expression, `make-regexp' throws a |
| 10287 | `regular-expression-syntax' error. |
| 10288 | |
| 10289 | FLAGS may be the bitwise-or of one or more of the following: |
| 10290 | |
| 10291 | **** Constant: regexp/extended |
| 10292 | Use POSIX Extended Regular Expression syntax when interpreting |
| 10293 | STR. If not set, POSIX Basic Regular Expression syntax is used. |
| 10294 | If the FLAGS argument is omitted, we assume regexp/extended. |
| 10295 | |
| 10296 | **** Constant: regexp/icase |
| 10297 | Do not differentiate case. Subsequent searches using the |
| 10298 | returned regular expression will be case insensitive. |
| 10299 | |
| 10300 | **** Constant: regexp/newline |
| 10301 | Match-any-character operators don't match a newline. |
| 10302 | |
| 10303 | A non-matching list ([^...]) not containing a newline matches a |
| 10304 | newline. |
| 10305 | |
| 10306 | Match-beginning-of-line operator (^) matches the empty string |
| 10307 | immediately after a newline, regardless of whether the FLAGS |
| 10308 | passed to regexp-exec contain regexp/notbol. |
| 10309 | |
| 10310 | Match-end-of-line operator ($) matches the empty string |
| 10311 | immediately before a newline, regardless of whether the FLAGS |
| 10312 | passed to regexp-exec contain regexp/noteol. |
| 10313 | |
| 10314 | **** Function: regexp-exec REGEXP STR [START [FLAGS]] |
| 10315 | Match the compiled regular expression REGEXP against `str'. If |
| 10316 | the optional integer START argument is provided, begin matching |
| 10317 | from that position in the string. Return a match structure |
| 10318 | describing the results of the match, or `#f' if no match could be |
| 10319 | found. |
| 10320 | |
| 10321 | FLAGS may be the bitwise-or of one or more of the following: |
| 10322 | |
| 10323 | **** Constant: regexp/notbol |
| 10324 | The match-beginning-of-line operator always fails to match (but |
| 10325 | see the compilation flag regexp/newline above) This flag may be |
| 10326 | used when different portions of a string are passed to |
| 10327 | regexp-exec and the beginning of the string should not be |
| 10328 | interpreted as the beginning of the line. |
| 10329 | |
| 10330 | **** Constant: regexp/noteol |
| 10331 | The match-end-of-line operator always fails to match (but see the |
| 10332 | compilation flag regexp/newline above) |
| 10333 | |
| 10334 | **** Function: regexp? OBJ |
| 10335 | Return `#t' if OBJ is a compiled regular expression, or `#f' |
| 10336 | otherwise. |
| 10337 | |
| 10338 | Regular expressions are commonly used to find patterns in one string |
| 10339 | and replace them with the contents of another string. |
| 10340 | |
| 10341 | **** Function: regexp-substitute PORT MATCH [ITEM...] |
| 10342 | Write to the output port PORT selected contents of the match |
| 10343 | structure MATCH. Each ITEM specifies what should be written, and |
| 10344 | may be one of the following arguments: |
| 10345 | |
| 10346 | * A string. String arguments are written out verbatim. |
| 10347 | |
| 10348 | * An integer. The submatch with that number is written. |
| 10349 | |
| 10350 | * The symbol `pre'. The portion of the matched string preceding |
| 10351 | the regexp match is written. |
| 10352 | |
| 10353 | * The symbol `post'. The portion of the matched string |
| 10354 | following the regexp match is written. |
| 10355 | |
| 10356 | PORT may be `#f', in which case nothing is written; instead, |
| 10357 | `regexp-substitute' constructs a string from the specified ITEMs |
| 10358 | and returns that. |
| 10359 | |
| 10360 | **** Function: regexp-substitute/global PORT REGEXP TARGET [ITEM...] |
| 10361 | Similar to `regexp-substitute', but can be used to perform global |
| 10362 | substitutions on STR. Instead of taking a match structure as an |
| 10363 | argument, `regexp-substitute/global' takes two string arguments: a |
| 10364 | REGEXP string describing a regular expression, and a TARGET string |
| 10365 | which should be matched against this regular expression. |
| 10366 | |
| 10367 | Each ITEM behaves as in REGEXP-SUBSTITUTE, with the following |
| 10368 | exceptions: |
| 10369 | |
| 10370 | * A function may be supplied. When this function is called, it |
| 10371 | will be passed one argument: a match structure for a given |
| 10372 | regular expression match. It should return a string to be |
| 10373 | written out to PORT. |
| 10374 | |
| 10375 | * The `post' symbol causes `regexp-substitute/global' to recurse |
| 10376 | on the unmatched portion of STR. This *must* be supplied in |
| 10377 | order to perform global search-and-replace on STR; if it is |
| 10378 | not present among the ITEMs, then `regexp-substitute/global' |
| 10379 | will return after processing a single match. |
| 10380 | |
| 10381 | *** Match Structures |
| 10382 | |
| 10383 | A "match structure" is the object returned by `string-match' and |
| 10384 | `regexp-exec'. It describes which portion of a string, if any, matched |
| 10385 | the given regular expression. Match structures include: a reference to |
| 10386 | the string that was checked for matches; the starting and ending |
| 10387 | positions of the regexp match; and, if the regexp included any |
| 10388 | parenthesized subexpressions, the starting and ending positions of each |
| 10389 | submatch. |
| 10390 | |
| 10391 | In each of the regexp match functions described below, the `match' |
| 10392 | argument must be a match structure returned by a previous call to |
| 10393 | `string-match' or `regexp-exec'. Most of these functions return some |
| 10394 | information about the original target string that was matched against a |
| 10395 | regular expression; we will call that string TARGET for easy reference. |
| 10396 | |
| 10397 | **** Function: regexp-match? OBJ |
| 10398 | Return `#t' if OBJ is a match structure returned by a previous |
| 10399 | call to `regexp-exec', or `#f' otherwise. |
| 10400 | |
| 10401 | **** Function: match:substring MATCH [N] |
| 10402 | Return the portion of TARGET matched by subexpression number N. |
| 10403 | Submatch 0 (the default) represents the entire regexp match. If |
| 10404 | the regular expression as a whole matched, but the subexpression |
| 10405 | number N did not match, return `#f'. |
| 10406 | |
| 10407 | **** Function: match:start MATCH [N] |
| 10408 | Return the starting position of submatch number N. |
| 10409 | |
| 10410 | **** Function: match:end MATCH [N] |
| 10411 | Return the ending position of submatch number N. |
| 10412 | |
| 10413 | **** Function: match:prefix MATCH |
| 10414 | Return the unmatched portion of TARGET preceding the regexp match. |
| 10415 | |
| 10416 | **** Function: match:suffix MATCH |
| 10417 | Return the unmatched portion of TARGET following the regexp match. |
| 10418 | |
| 10419 | **** Function: match:count MATCH |
| 10420 | Return the number of parenthesized subexpressions from MATCH. |
| 10421 | Note that the entire regular expression match itself counts as a |
| 10422 | subexpression, and failed submatches are included in the count. |
| 10423 | |
| 10424 | **** Function: match:string MATCH |
| 10425 | Return the original TARGET string. |
| 10426 | |
| 10427 | *** Backslash Escapes |
| 10428 | |
| 10429 | Sometimes you will want a regexp to match characters like `*' or `$' |
| 10430 | exactly. For example, to check whether a particular string represents |
| 10431 | a menu entry from an Info node, it would be useful to match it against |
| 10432 | a regexp like `^* [^:]*::'. However, this won't work; because the |
| 10433 | asterisk is a metacharacter, it won't match the `*' at the beginning of |
| 10434 | the string. In this case, we want to make the first asterisk un-magic. |
| 10435 | |
| 10436 | You can do this by preceding the metacharacter with a backslash |
| 10437 | character `\'. (This is also called "quoting" the metacharacter, and |
| 10438 | is known as a "backslash escape".) When Guile sees a backslash in a |
| 10439 | regular expression, it considers the following glyph to be an ordinary |
| 10440 | character, no matter what special meaning it would ordinarily have. |
| 10441 | Therefore, we can make the above example work by changing the regexp to |
| 10442 | `^\* [^:]*::'. The `\*' sequence tells the regular expression engine |
| 10443 | to match only a single asterisk in the target string. |
| 10444 | |
| 10445 | Since the backslash is itself a metacharacter, you may force a |
| 10446 | regexp to match a backslash in the target string by preceding the |
| 10447 | backslash with itself. For example, to find variable references in a |
| 10448 | TeX program, you might want to find occurrences of the string `\let\' |
| 10449 | followed by any number of alphabetic characters. The regular expression |
| 10450 | `\\let\\[A-Za-z]*' would do this: the double backslashes in the regexp |
| 10451 | each match a single backslash in the target string. |
| 10452 | |
| 10453 | **** Function: regexp-quote STR |
| 10454 | Quote each special character found in STR with a backslash, and |
| 10455 | return the resulting string. |
| 10456 | |
| 10457 | *Very important:* Using backslash escapes in Guile source code (as |
| 10458 | in Emacs Lisp or C) can be tricky, because the backslash character has |
| 10459 | special meaning for the Guile reader. For example, if Guile encounters |
| 10460 | the character sequence `\n' in the middle of a string while processing |
| 10461 | Scheme code, it replaces those characters with a newline character. |
| 10462 | Similarly, the character sequence `\t' is replaced by a horizontal tab. |
| 10463 | Several of these "escape sequences" are processed by the Guile reader |
| 10464 | before your code is executed. Unrecognized escape sequences are |
| 10465 | ignored: if the characters `\*' appear in a string, they will be |
| 10466 | translated to the single character `*'. |
| 10467 | |
| 10468 | This translation is obviously undesirable for regular expressions, |
| 10469 | since we want to be able to include backslashes in a string in order to |
| 10470 | escape regexp metacharacters. Therefore, to make sure that a backslash |
| 10471 | is preserved in a string in your Guile program, you must use *two* |
| 10472 | consecutive backslashes: |
| 10473 | |
| 10474 | (define Info-menu-entry-pattern (make-regexp "^\\* [^:]*")) |
| 10475 | |
| 10476 | The string in this example is preprocessed by the Guile reader before |
| 10477 | any code is executed. The resulting argument to `make-regexp' is the |
| 10478 | string `^\* [^:]*', which is what we really want. |
| 10479 | |
| 10480 | This also means that in order to write a regular expression that |
| 10481 | matches a single backslash character, the regular expression string in |
| 10482 | the source code must include *four* backslashes. Each consecutive pair |
| 10483 | of backslashes gets translated by the Guile reader to a single |
| 10484 | backslash, and the resulting double-backslash is interpreted by the |
| 10485 | regexp engine as matching a single backslash character. Hence: |
| 10486 | |
| 10487 | (define tex-variable-pattern (make-regexp "\\\\let\\\\=[A-Za-z]*")) |
| 10488 | |
| 10489 | The reason for the unwieldiness of this syntax is historical. Both |
| 10490 | regular expression pattern matchers and Unix string processing systems |
| 10491 | have traditionally used backslashes with the special meanings described |
| 10492 | above. The POSIX regular expression specification and ANSI C standard |
| 10493 | both require these semantics. Attempting to abandon either convention |
| 10494 | would cause other kinds of compatibility problems, possibly more severe |
| 10495 | ones. Therefore, without extending the Scheme reader to support |
| 10496 | strings with different quoting conventions (an ungainly and confusing |
| 10497 | extension when implemented in other languages), we must adhere to this |
| 10498 | cumbersome escape syntax. |
| 10499 | |
| 10500 | * Changes to the gh_ interface |
| 10501 | |
| 10502 | * Changes to the scm_ interface |
| 10503 | |
| 10504 | * Changes to system call interfaces: |
| 10505 | |
| 10506 | ** The value returned by `raise' is now unspecified. It throws an exception |
| 10507 | if an error occurs. |
| 10508 | |
| 10509 | *** A new procedure `sigaction' can be used to install signal handlers |
| 10510 | |
| 10511 | (sigaction signum [action] [flags]) |
| 10512 | |
| 10513 | signum is the signal number, which can be specified using the value |
| 10514 | of SIGINT etc. |
| 10515 | |
| 10516 | If action is omitted, sigaction returns a pair: the CAR is the current |
| 10517 | signal hander, which will be either an integer with the value SIG_DFL |
| 10518 | (default action) or SIG_IGN (ignore), or the Scheme procedure which |
| 10519 | handles the signal, or #f if a non-Scheme procedure handles the |
| 10520 | signal. The CDR contains the current sigaction flags for the handler. |
| 10521 | |
| 10522 | If action is provided, it is installed as the new handler for signum. |
| 10523 | action can be a Scheme procedure taking one argument, or the value of |
| 10524 | SIG_DFL (default action) or SIG_IGN (ignore), or #f to restore |
| 10525 | whatever signal handler was installed before sigaction was first used. |
| 10526 | Flags can optionally be specified for the new handler (SA_RESTART is |
| 10527 | always used if the system provides it, so need not be specified.) The |
| 10528 | return value is a pair with information about the old handler as |
| 10529 | described above. |
| 10530 | |
| 10531 | This interface does not provide access to the "signal blocking" |
| 10532 | facility. Maybe this is not needed, since the thread support may |
| 10533 | provide solutions to the problem of consistent access to data |
| 10534 | structures. |
| 10535 | |
| 10536 | *** A new procedure `flush-all-ports' is equivalent to running |
| 10537 | `force-output' on every port open for output. |
| 10538 | |
| 10539 | ** Guile now provides information on how it was built, via the new |
| 10540 | global variable, %guile-build-info. This variable records the values |
| 10541 | of the standard GNU makefile directory variables as an assocation |
| 10542 | list, mapping variable names (symbols) onto directory paths (strings). |
| 10543 | For example, to find out where the Guile link libraries were |
| 10544 | installed, you can say: |
| 10545 | |
| 10546 | guile -c "(display (assq-ref %guile-build-info 'libdir)) (newline)" |
| 10547 | |
| 10548 | |
| 10549 | * Changes to the scm_ interface |
| 10550 | |
| 10551 | ** The new function scm_handle_by_message_noexit is just like the |
| 10552 | existing scm_handle_by_message function, except that it doesn't call |
| 10553 | exit to terminate the process. Instead, it prints a message and just |
| 10554 | returns #f. This might be a more appropriate catch-all handler for |
| 10555 | new dynamic roots and threads. |
| 10556 | |
| 10557 | \f |
| 10558 | Changes in Guile 1.1 (released Friday, May 16 1997): |
| 10559 | |
| 10560 | * Changes to the distribution. |
| 10561 | |
| 10562 | The Guile 1.0 distribution has been split up into several smaller |
| 10563 | pieces: |
| 10564 | guile-core --- the Guile interpreter itself. |
| 10565 | guile-tcltk --- the interface between the Guile interpreter and |
| 10566 | Tcl/Tk; Tcl is an interpreter for a stringy language, and Tk |
| 10567 | is a toolkit for building graphical user interfaces. |
| 10568 | guile-rgx-ctax --- the interface between Guile and the Rx regular |
| 10569 | expression matcher, and the translator for the Ctax |
| 10570 | programming language. These are packaged together because the |
| 10571 | Ctax translator uses Rx to parse Ctax source code. |
| 10572 | |
| 10573 | This NEWS file describes the changes made to guile-core since the 1.0 |
| 10574 | release. |
| 10575 | |
| 10576 | We no longer distribute the documentation, since it was either out of |
| 10577 | date, or incomplete. As soon as we have current documentation, we |
| 10578 | will distribute it. |
| 10579 | |
| 10580 | |
| 10581 | |
| 10582 | * Changes to the stand-alone interpreter |
| 10583 | |
| 10584 | ** guile now accepts command-line arguments compatible with SCSH, Olin |
| 10585 | Shivers' Scheme Shell. |
| 10586 | |
| 10587 | In general, arguments are evaluated from left to right, but there are |
| 10588 | exceptions. The following switches stop argument processing, and |
| 10589 | stash all remaining command-line arguments as the value returned by |
| 10590 | the (command-line) function. |
| 10591 | -s SCRIPT load Scheme source code from FILE, and exit |
| 10592 | -c EXPR evalute Scheme expression EXPR, and exit |
| 10593 | -- stop scanning arguments; run interactively |
| 10594 | |
| 10595 | The switches below are processed as they are encountered. |
| 10596 | -l FILE load Scheme source code from FILE |
| 10597 | -e FUNCTION after reading script, apply FUNCTION to |
| 10598 | command line arguments |
| 10599 | -ds do -s script at this point |
| 10600 | --emacs enable Emacs protocol (experimental) |
| 10601 | -h, --help display this help and exit |
| 10602 | -v, --version display version information and exit |
| 10603 | \ read arguments from following script lines |
| 10604 | |
| 10605 | So, for example, here is a Guile script named `ekko' (thanks, Olin) |
| 10606 | which re-implements the traditional "echo" command: |
| 10607 | |
| 10608 | #!/usr/local/bin/guile -s |
| 10609 | !# |
| 10610 | (define (main args) |
| 10611 | (map (lambda (arg) (display arg) (display " ")) |
| 10612 | (cdr args)) |
| 10613 | (newline)) |
| 10614 | |
| 10615 | (main (command-line)) |
| 10616 | |
| 10617 | Suppose we invoke this script as follows: |
| 10618 | |
| 10619 | ekko a speckled gecko |
| 10620 | |
| 10621 | Through the magic of Unix script processing (triggered by the `#!' |
| 10622 | token at the top of the file), /usr/local/bin/guile receives the |
| 10623 | following list of command-line arguments: |
| 10624 | |
| 10625 | ("-s" "./ekko" "a" "speckled" "gecko") |
| 10626 | |
| 10627 | Unix inserts the name of the script after the argument specified on |
| 10628 | the first line of the file (in this case, "-s"), and then follows that |
| 10629 | with the arguments given to the script. Guile loads the script, which |
| 10630 | defines the `main' function, and then applies it to the list of |
| 10631 | remaining command-line arguments, ("a" "speckled" "gecko"). |
| 10632 | |
| 10633 | In Unix, the first line of a script file must take the following form: |
| 10634 | |
| 10635 | #!INTERPRETER ARGUMENT |
| 10636 | |
| 10637 | where INTERPRETER is the absolute filename of the interpreter |
| 10638 | executable, and ARGUMENT is a single command-line argument to pass to |
| 10639 | the interpreter. |
| 10640 | |
| 10641 | You may only pass one argument to the interpreter, and its length is |
| 10642 | limited. These restrictions can be annoying to work around, so Guile |
| 10643 | provides a general mechanism (borrowed from, and compatible with, |
| 10644 | SCSH) for circumventing them. |
| 10645 | |
| 10646 | If the ARGUMENT in a Guile script is a single backslash character, |
| 10647 | `\', Guile will open the script file, parse arguments from its second |
| 10648 | and subsequent lines, and replace the `\' with them. So, for example, |
| 10649 | here is another implementation of the `ekko' script: |
| 10650 | |
| 10651 | #!/usr/local/bin/guile \ |
| 10652 | -e main -s |
| 10653 | !# |
| 10654 | (define (main args) |
| 10655 | (for-each (lambda (arg) (display arg) (display " ")) |
| 10656 | (cdr args)) |
| 10657 | (newline)) |
| 10658 | |
| 10659 | If the user invokes this script as follows: |
| 10660 | |
| 10661 | ekko a speckled gecko |
| 10662 | |
| 10663 | Unix expands this into |
| 10664 | |
| 10665 | /usr/local/bin/guile \ ekko a speckled gecko |
| 10666 | |
| 10667 | When Guile sees the `\' argument, it replaces it with the arguments |
| 10668 | read from the second line of the script, producing: |
| 10669 | |
| 10670 | /usr/local/bin/guile -e main -s ekko a speckled gecko |
| 10671 | |
| 10672 | This tells Guile to load the `ekko' script, and apply the function |
| 10673 | `main' to the argument list ("a" "speckled" "gecko"). |
| 10674 | |
| 10675 | Here is how Guile parses the command-line arguments: |
| 10676 | - Each space character terminates an argument. This means that two |
| 10677 | spaces in a row introduce an empty-string argument. |
| 10678 | - The tab character is not permitted (unless you quote it with the |
| 10679 | backslash character, as described below), to avoid confusion. |
| 10680 | - The newline character terminates the sequence of arguments, and will |
| 10681 | also terminate a final non-empty argument. (However, a newline |
| 10682 | following a space will not introduce a final empty-string argument; |
| 10683 | it only terminates the argument list.) |
| 10684 | - The backslash character is the escape character. It escapes |
| 10685 | backslash, space, tab, and newline. The ANSI C escape sequences |
| 10686 | like \n and \t are also supported. These produce argument |
| 10687 | constituents; the two-character combination \n doesn't act like a |
| 10688 | terminating newline. The escape sequence \NNN for exactly three |
| 10689 | octal digits reads as the character whose ASCII code is NNN. As |
| 10690 | above, characters produced this way are argument constituents. |
| 10691 | Backslash followed by other characters is not allowed. |
| 10692 | |
| 10693 | * Changes to the procedure for linking libguile with your programs |
| 10694 | |
| 10695 | ** Guile now builds and installs a shared guile library, if your |
| 10696 | system support shared libraries. (It still builds a static library on |
| 10697 | all systems.) Guile automatically detects whether your system |
| 10698 | supports shared libraries. To prevent Guile from buildisg shared |
| 10699 | libraries, pass the `--disable-shared' flag to the configure script. |
| 10700 | |
| 10701 | Guile takes longer to compile when it builds shared libraries, because |
| 10702 | it must compile every file twice --- once to produce position- |
| 10703 | independent object code, and once to produce normal object code. |
| 10704 | |
| 10705 | ** The libthreads library has been merged into libguile. |
| 10706 | |
| 10707 | To link a program against Guile, you now need only link against |
| 10708 | -lguile and -lqt; -lthreads is no longer needed. If you are using |
| 10709 | autoconf to generate configuration scripts for your application, the |
| 10710 | following lines should suffice to add the appropriate libraries to |
| 10711 | your link command: |
| 10712 | |
| 10713 | ### Find quickthreads and libguile. |
| 10714 | AC_CHECK_LIB(qt, main) |
| 10715 | AC_CHECK_LIB(guile, scm_shell) |
| 10716 | |
| 10717 | * Changes to Scheme functions |
| 10718 | |
| 10719 | ** Guile Scheme's special syntax for keyword objects is now optional, |
| 10720 | and disabled by default. |
| 10721 | |
| 10722 | The syntax variation from R4RS made it difficult to port some |
| 10723 | interesting packages to Guile. The routines which accepted keyword |
| 10724 | arguments (mostly in the module system) have been modified to also |
| 10725 | accept symbols whose names begin with `:'. |
| 10726 | |
| 10727 | To change the keyword syntax, you must first import the (ice-9 debug) |
| 10728 | module: |
| 10729 | (use-modules (ice-9 debug)) |
| 10730 | |
| 10731 | Then you can enable the keyword syntax as follows: |
| 10732 | (read-set! keywords 'prefix) |
| 10733 | |
| 10734 | To disable keyword syntax, do this: |
| 10735 | (read-set! keywords #f) |
| 10736 | |
| 10737 | ** Many more primitive functions accept shared substrings as |
| 10738 | arguments. In the past, these functions required normal, mutable |
| 10739 | strings as arguments, although they never made use of this |
| 10740 | restriction. |
| 10741 | |
| 10742 | ** The uniform array functions now operate on byte vectors. These |
| 10743 | functions are `array-fill!', `serial-array-copy!', `array-copy!', |
| 10744 | `serial-array-map', `array-map', `array-for-each', and |
| 10745 | `array-index-map!'. |
| 10746 | |
| 10747 | ** The new functions `trace' and `untrace' implement simple debugging |
| 10748 | support for Scheme functions. |
| 10749 | |
| 10750 | The `trace' function accepts any number of procedures as arguments, |
| 10751 | and tells the Guile interpreter to display each procedure's name and |
| 10752 | arguments each time the procedure is invoked. When invoked with no |
| 10753 | arguments, `trace' returns the list of procedures currently being |
| 10754 | traced. |
| 10755 | |
| 10756 | The `untrace' function accepts any number of procedures as arguments, |
| 10757 | and tells the Guile interpreter not to trace them any more. When |
| 10758 | invoked with no arguments, `untrace' untraces all curretly traced |
| 10759 | procedures. |
| 10760 | |
| 10761 | The tracing in Guile has an advantage over most other systems: we |
| 10762 | don't create new procedure objects, but mark the procedure objects |
| 10763 | themselves. This means that anonymous and internal procedures can be |
| 10764 | traced. |
| 10765 | |
| 10766 | ** The function `assert-repl-prompt' has been renamed to |
| 10767 | `set-repl-prompt!'. It takes one argument, PROMPT. |
| 10768 | - If PROMPT is #f, the Guile read-eval-print loop will not prompt. |
| 10769 | - If PROMPT is a string, we use it as a prompt. |
| 10770 | - If PROMPT is a procedure accepting no arguments, we call it, and |
| 10771 | display the result as a prompt. |
| 10772 | - Otherwise, we display "> ". |
| 10773 | |
| 10774 | ** The new function `eval-string' reads Scheme expressions from a |
| 10775 | string and evaluates them, returning the value of the last expression |
| 10776 | in the string. If the string contains no expressions, it returns an |
| 10777 | unspecified value. |
| 10778 | |
| 10779 | ** The new function `thunk?' returns true iff its argument is a |
| 10780 | procedure of zero arguments. |
| 10781 | |
| 10782 | ** `defined?' is now a builtin function, instead of syntax. This |
| 10783 | means that its argument should be quoted. It returns #t iff its |
| 10784 | argument is bound in the current module. |
| 10785 | |
| 10786 | ** The new syntax `use-modules' allows you to add new modules to your |
| 10787 | environment without re-typing a complete `define-module' form. It |
| 10788 | accepts any number of module names as arguments, and imports their |
| 10789 | public bindings into the current module. |
| 10790 | |
| 10791 | ** The new function (module-defined? NAME MODULE) returns true iff |
| 10792 | NAME, a symbol, is defined in MODULE, a module object. |
| 10793 | |
| 10794 | ** The new function `builtin-bindings' creates and returns a hash |
| 10795 | table containing copies of all the root module's bindings. |
| 10796 | |
| 10797 | ** The new function `builtin-weak-bindings' does the same as |
| 10798 | `builtin-bindings', but creates a doubly-weak hash table. |
| 10799 | |
| 10800 | ** The `equal?' function now considers variable objects to be |
| 10801 | equivalent if they have the same name and the same value. |
| 10802 | |
| 10803 | ** The new function `command-line' returns the command-line arguments |
| 10804 | given to Guile, as a list of strings. |
| 10805 | |
| 10806 | When using guile as a script interpreter, `command-line' returns the |
| 10807 | script's arguments; those processed by the interpreter (like `-s' or |
| 10808 | `-c') are omitted. (In other words, you get the normal, expected |
| 10809 | behavior.) Any application that uses scm_shell to process its |
| 10810 | command-line arguments gets this behavior as well. |
| 10811 | |
| 10812 | ** The new function `load-user-init' looks for a file called `.guile' |
| 10813 | in the user's home directory, and loads it if it exists. This is |
| 10814 | mostly for use by the code generated by scm_compile_shell_switches, |
| 10815 | but we thought it might also be useful in other circumstances. |
| 10816 | |
| 10817 | ** The new function `log10' returns the base-10 logarithm of its |
| 10818 | argument. |
| 10819 | |
| 10820 | ** Changes to I/O functions |
| 10821 | |
| 10822 | *** The functions `read', `primitive-load', `read-and-eval!', and |
| 10823 | `primitive-load-path' no longer take optional arguments controlling |
| 10824 | case insensitivity and a `#' parser. |
| 10825 | |
| 10826 | Case sensitivity is now controlled by a read option called |
| 10827 | `case-insensitive'. The user can add new `#' syntaxes with the |
| 10828 | `read-hash-extend' function (see below). |
| 10829 | |
| 10830 | *** The new function `read-hash-extend' allows the user to change the |
| 10831 | syntax of Guile Scheme in a somewhat controlled way. |
| 10832 | |
| 10833 | (read-hash-extend CHAR PROC) |
| 10834 | When parsing S-expressions, if we read a `#' character followed by |
| 10835 | the character CHAR, use PROC to parse an object from the stream. |
| 10836 | If PROC is #f, remove any parsing procedure registered for CHAR. |
| 10837 | |
| 10838 | The reader applies PROC to two arguments: CHAR and an input port. |
| 10839 | |
| 10840 | *** The new functions read-delimited and read-delimited! provide a |
| 10841 | general mechanism for doing delimited input on streams. |
| 10842 | |
| 10843 | (read-delimited DELIMS [PORT HANDLE-DELIM]) |
| 10844 | Read until we encounter one of the characters in DELIMS (a string), |
| 10845 | or end-of-file. PORT is the input port to read from; it defaults to |
| 10846 | the current input port. The HANDLE-DELIM parameter determines how |
| 10847 | the terminating character is handled; it should be one of the |
| 10848 | following symbols: |
| 10849 | |
| 10850 | 'trim omit delimiter from result |
| 10851 | 'peek leave delimiter character in input stream |
| 10852 | 'concat append delimiter character to returned value |
| 10853 | 'split return a pair: (RESULT . TERMINATOR) |
| 10854 | |
| 10855 | HANDLE-DELIM defaults to 'peek. |
| 10856 | |
| 10857 | (read-delimited! DELIMS BUF [PORT HANDLE-DELIM START END]) |
| 10858 | A side-effecting variant of `read-delimited'. |
| 10859 | |
| 10860 | The data is written into the string BUF at the indices in the |
| 10861 | half-open interval [START, END); the default interval is the whole |
| 10862 | string: START = 0 and END = (string-length BUF). The values of |
| 10863 | START and END must specify a well-defined interval in BUF, i.e. |
| 10864 | 0 <= START <= END <= (string-length BUF). |
| 10865 | |
| 10866 | It returns NBYTES, the number of bytes read. If the buffer filled |
| 10867 | up without a delimiter character being found, it returns #f. If the |
| 10868 | port is at EOF when the read starts, it returns the EOF object. |
| 10869 | |
| 10870 | If an integer is returned (i.e., the read is successfully terminated |
| 10871 | by reading a delimiter character), then the HANDLE-DELIM parameter |
| 10872 | determines how to handle the terminating character. It is described |
| 10873 | above, and defaults to 'peek. |
| 10874 | |
| 10875 | (The descriptions of these functions were borrowed from the SCSH |
| 10876 | manual, by Olin Shivers and Brian Carlstrom.) |
| 10877 | |
| 10878 | *** The `%read-delimited!' function is the primitive used to implement |
| 10879 | `read-delimited' and `read-delimited!'. |
| 10880 | |
| 10881 | (%read-delimited! DELIMS BUF GOBBLE? [PORT START END]) |
| 10882 | |
| 10883 | This returns a pair of values: (TERMINATOR . NUM-READ). |
| 10884 | - TERMINATOR describes why the read was terminated. If it is a |
| 10885 | character or the eof object, then that is the value that terminated |
| 10886 | the read. If it is #f, the function filled the buffer without finding |
| 10887 | a delimiting character. |
| 10888 | - NUM-READ is the number of characters read into BUF. |
| 10889 | |
| 10890 | If the read is successfully terminated by reading a delimiter |
| 10891 | character, then the gobble? parameter determines what to do with the |
| 10892 | terminating character. If true, the character is removed from the |
| 10893 | input stream; if false, the character is left in the input stream |
| 10894 | where a subsequent read operation will retrieve it. In either case, |
| 10895 | the character is also the first value returned by the procedure call. |
| 10896 | |
| 10897 | (The descriptions of this function was borrowed from the SCSH manual, |
| 10898 | by Olin Shivers and Brian Carlstrom.) |
| 10899 | |
| 10900 | *** The `read-line' and `read-line!' functions have changed; they now |
| 10901 | trim the terminator by default; previously they appended it to the |
| 10902 | returned string. For the old behavior, use (read-line PORT 'concat). |
| 10903 | |
| 10904 | *** The functions `uniform-array-read!' and `uniform-array-write!' now |
| 10905 | take new optional START and END arguments, specifying the region of |
| 10906 | the array to read and write. |
| 10907 | |
| 10908 | *** The `ungetc-char-ready?' function has been removed. We feel it's |
| 10909 | inappropriate for an interface to expose implementation details this |
| 10910 | way. |
| 10911 | |
| 10912 | ** Changes to the Unix library and system call interface |
| 10913 | |
| 10914 | *** The new fcntl function provides access to the Unix `fcntl' system |
| 10915 | call. |
| 10916 | |
| 10917 | (fcntl PORT COMMAND VALUE) |
| 10918 | Apply COMMAND to PORT's file descriptor, with VALUE as an argument. |
| 10919 | Values for COMMAND are: |
| 10920 | |
| 10921 | F_DUPFD duplicate a file descriptor |
| 10922 | F_GETFD read the descriptor's close-on-exec flag |
| 10923 | F_SETFD set the descriptor's close-on-exec flag to VALUE |
| 10924 | F_GETFL read the descriptor's flags, as set on open |
| 10925 | F_SETFL set the descriptor's flags, as set on open to VALUE |
| 10926 | F_GETOWN return the process ID of a socket's owner, for SIGIO |
| 10927 | F_SETOWN set the process that owns a socket to VALUE, for SIGIO |
| 10928 | FD_CLOEXEC not sure what this is |
| 10929 | |
| 10930 | For details, see the documentation for the fcntl system call. |
| 10931 | |
| 10932 | *** The arguments to `select' have changed, for compatibility with |
| 10933 | SCSH. The TIMEOUT parameter may now be non-integral, yielding the |
| 10934 | expected behavior. The MILLISECONDS parameter has been changed to |
| 10935 | MICROSECONDS, to more closely resemble the underlying system call. |
| 10936 | The RVEC, WVEC, and EVEC arguments can now be vectors; the type of the |
| 10937 | corresponding return set will be the same. |
| 10938 | |
| 10939 | *** The arguments to the `mknod' system call have changed. They are |
| 10940 | now: |
| 10941 | |
| 10942 | (mknod PATH TYPE PERMS DEV) |
| 10943 | Create a new file (`node') in the file system. PATH is the name of |
| 10944 | the file to create. TYPE is the kind of file to create; it should |
| 10945 | be 'fifo, 'block-special, or 'char-special. PERMS specifies the |
| 10946 | permission bits to give the newly created file. If TYPE is |
| 10947 | 'block-special or 'char-special, DEV specifies which device the |
| 10948 | special file refers to; its interpretation depends on the kind of |
| 10949 | special file being created. |
| 10950 | |
| 10951 | *** The `fork' function has been renamed to `primitive-fork', to avoid |
| 10952 | clashing with various SCSH forks. |
| 10953 | |
| 10954 | *** The `recv' and `recvfrom' functions have been renamed to `recv!' |
| 10955 | and `recvfrom!'. They no longer accept a size for a second argument; |
| 10956 | you must pass a string to hold the received value. They no longer |
| 10957 | return the buffer. Instead, `recv' returns the length of the message |
| 10958 | received, and `recvfrom' returns a pair containing the packet's length |
| 10959 | and originating address. |
| 10960 | |
| 10961 | *** The file descriptor datatype has been removed, as have the |
| 10962 | `read-fd', `write-fd', `close', `lseek', and `dup' functions. |
| 10963 | We plan to replace these functions with a SCSH-compatible interface. |
| 10964 | |
| 10965 | *** The `create' function has been removed; it's just a special case |
| 10966 | of `open'. |
| 10967 | |
| 10968 | *** There are new functions to break down process termination status |
| 10969 | values. In the descriptions below, STATUS is a value returned by |
| 10970 | `waitpid'. |
| 10971 | |
| 10972 | (status:exit-val STATUS) |
| 10973 | If the child process exited normally, this function returns the exit |
| 10974 | code for the child process (i.e., the value passed to exit, or |
| 10975 | returned from main). If the child process did not exit normally, |
| 10976 | this function returns #f. |
| 10977 | |
| 10978 | (status:stop-sig STATUS) |
| 10979 | If the child process was suspended by a signal, this function |
| 10980 | returns the signal that suspended the child. Otherwise, it returns |
| 10981 | #f. |
| 10982 | |
| 10983 | (status:term-sig STATUS) |
| 10984 | If the child process terminated abnormally, this function returns |
| 10985 | the signal that terminated the child. Otherwise, this function |
| 10986 | returns false. |
| 10987 | |
| 10988 | POSIX promises that exactly one of these functions will return true on |
| 10989 | a valid STATUS value. |
| 10990 | |
| 10991 | These functions are compatible with SCSH. |
| 10992 | |
| 10993 | *** There are new accessors and setters for the broken-out time vectors |
| 10994 | returned by `localtime', `gmtime', and that ilk. They are: |
| 10995 | |
| 10996 | Component Accessor Setter |
| 10997 | ========================= ============ ============ |
| 10998 | seconds tm:sec set-tm:sec |
| 10999 | minutes tm:min set-tm:min |
| 11000 | hours tm:hour set-tm:hour |
| 11001 | day of the month tm:mday set-tm:mday |
| 11002 | month tm:mon set-tm:mon |
| 11003 | year tm:year set-tm:year |
| 11004 | day of the week tm:wday set-tm:wday |
| 11005 | day in the year tm:yday set-tm:yday |
| 11006 | daylight saving time tm:isdst set-tm:isdst |
| 11007 | GMT offset, seconds tm:gmtoff set-tm:gmtoff |
| 11008 | name of time zone tm:zone set-tm:zone |
| 11009 | |
| 11010 | *** There are new accessors for the vectors returned by `uname', |
| 11011 | describing the host system: |
| 11012 | |
| 11013 | Component Accessor |
| 11014 | ============================================== ================ |
| 11015 | name of the operating system implementation utsname:sysname |
| 11016 | network name of this machine utsname:nodename |
| 11017 | release level of the operating system utsname:release |
| 11018 | version level of the operating system utsname:version |
| 11019 | machine hardware platform utsname:machine |
| 11020 | |
| 11021 | *** There are new accessors for the vectors returned by `getpw', |
| 11022 | `getpwnam', `getpwuid', and `getpwent', describing entries from the |
| 11023 | system's user database: |
| 11024 | |
| 11025 | Component Accessor |
| 11026 | ====================== ================= |
| 11027 | user name passwd:name |
| 11028 | user password passwd:passwd |
| 11029 | user id passwd:uid |
| 11030 | group id passwd:gid |
| 11031 | real name passwd:gecos |
| 11032 | home directory passwd:dir |
| 11033 | shell program passwd:shell |
| 11034 | |
| 11035 | *** There are new accessors for the vectors returned by `getgr', |
| 11036 | `getgrnam', `getgrgid', and `getgrent', describing entries from the |
| 11037 | system's group database: |
| 11038 | |
| 11039 | Component Accessor |
| 11040 | ======================= ============ |
| 11041 | group name group:name |
| 11042 | group password group:passwd |
| 11043 | group id group:gid |
| 11044 | group members group:mem |
| 11045 | |
| 11046 | *** There are new accessors for the vectors returned by `gethost', |
| 11047 | `gethostbyaddr', `gethostbyname', and `gethostent', describing |
| 11048 | internet hosts: |
| 11049 | |
| 11050 | Component Accessor |
| 11051 | ========================= =============== |
| 11052 | official name of host hostent:name |
| 11053 | alias list hostent:aliases |
| 11054 | host address type hostent:addrtype |
| 11055 | length of address hostent:length |
| 11056 | list of addresses hostent:addr-list |
| 11057 | |
| 11058 | *** There are new accessors for the vectors returned by `getnet', |
| 11059 | `getnetbyaddr', `getnetbyname', and `getnetent', describing internet |
| 11060 | networks: |
| 11061 | |
| 11062 | Component Accessor |
| 11063 | ========================= =============== |
| 11064 | official name of net netent:name |
| 11065 | alias list netent:aliases |
| 11066 | net number type netent:addrtype |
| 11067 | net number netent:net |
| 11068 | |
| 11069 | *** There are new accessors for the vectors returned by `getproto', |
| 11070 | `getprotobyname', `getprotobynumber', and `getprotoent', describing |
| 11071 | internet protocols: |
| 11072 | |
| 11073 | Component Accessor |
| 11074 | ========================= =============== |
| 11075 | official protocol name protoent:name |
| 11076 | alias list protoent:aliases |
| 11077 | protocol number protoent:proto |
| 11078 | |
| 11079 | *** There are new accessors for the vectors returned by `getserv', |
| 11080 | `getservbyname', `getservbyport', and `getservent', describing |
| 11081 | internet protocols: |
| 11082 | |
| 11083 | Component Accessor |
| 11084 | ========================= =============== |
| 11085 | official service name servent:name |
| 11086 | alias list servent:aliases |
| 11087 | port number servent:port |
| 11088 | protocol to use servent:proto |
| 11089 | |
| 11090 | *** There are new accessors for the sockaddr structures returned by |
| 11091 | `accept', `getsockname', `getpeername', `recvfrom!': |
| 11092 | |
| 11093 | Component Accessor |
| 11094 | ======================================== =============== |
| 11095 | address format (`family') sockaddr:fam |
| 11096 | path, for file domain addresses sockaddr:path |
| 11097 | address, for internet domain addresses sockaddr:addr |
| 11098 | TCP or UDP port, for internet sockaddr:port |
| 11099 | |
| 11100 | *** The `getpwent', `getgrent', `gethostent', `getnetent', |
| 11101 | `getprotoent', and `getservent' functions now return #f at the end of |
| 11102 | the user database. (They used to throw an exception.) |
| 11103 | |
| 11104 | Note that calling MUMBLEent function is equivalent to calling the |
| 11105 | corresponding MUMBLE function with no arguments. |
| 11106 | |
| 11107 | *** The `setpwent', `setgrent', `sethostent', `setnetent', |
| 11108 | `setprotoent', and `setservent' routines now take no arguments. |
| 11109 | |
| 11110 | *** The `gethost', `getproto', `getnet', and `getserv' functions now |
| 11111 | provide more useful information when they throw an exception. |
| 11112 | |
| 11113 | *** The `lnaof' function has been renamed to `inet-lnaof'. |
| 11114 | |
| 11115 | *** Guile now claims to have the `current-time' feature. |
| 11116 | |
| 11117 | *** The `mktime' function now takes an optional second argument ZONE, |
| 11118 | giving the time zone to use for the conversion. ZONE should be a |
| 11119 | string, in the same format as expected for the "TZ" environment variable. |
| 11120 | |
| 11121 | *** The `strptime' function now returns a pair (TIME . COUNT), where |
| 11122 | TIME is the parsed time as a vector, and COUNT is the number of |
| 11123 | characters from the string left unparsed. This function used to |
| 11124 | return the remaining characters as a string. |
| 11125 | |
| 11126 | *** The `gettimeofday' function has replaced the old `time+ticks' function. |
| 11127 | The return value is now (SECONDS . MICROSECONDS); the fractional |
| 11128 | component is no longer expressed in "ticks". |
| 11129 | |
| 11130 | *** The `ticks/sec' constant has been removed, in light of the above change. |
| 11131 | |
| 11132 | * Changes to the gh_ interface |
| 11133 | |
| 11134 | ** gh_eval_str() now returns an SCM object which is the result of the |
| 11135 | evaluation |
| 11136 | |
| 11137 | ** gh_scm2str() now copies the Scheme data to a caller-provided C |
| 11138 | array |
| 11139 | |
| 11140 | ** gh_scm2newstr() now makes a C array, copies the Scheme data to it, |
| 11141 | and returns the array |
| 11142 | |
| 11143 | ** gh_scm2str0() is gone: there is no need to distinguish |
| 11144 | null-terminated from non-null-terminated, since gh_scm2newstr() allows |
| 11145 | the user to interpret the data both ways. |
| 11146 | |
| 11147 | * Changes to the scm_ interface |
| 11148 | |
| 11149 | ** The new function scm_symbol_value0 provides an easy way to get a |
| 11150 | symbol's value from C code: |
| 11151 | |
| 11152 | SCM scm_symbol_value0 (char *NAME) |
| 11153 | Return the value of the symbol named by the null-terminated string |
| 11154 | NAME in the current module. If the symbol named NAME is unbound in |
| 11155 | the current module, return SCM_UNDEFINED. |
| 11156 | |
| 11157 | ** The new function scm_sysintern0 creates new top-level variables, |
| 11158 | without assigning them a value. |
| 11159 | |
| 11160 | SCM scm_sysintern0 (char *NAME) |
| 11161 | Create a new Scheme top-level variable named NAME. NAME is a |
| 11162 | null-terminated string. Return the variable's value cell. |
| 11163 | |
| 11164 | ** The function scm_internal_catch is the guts of catch. It handles |
| 11165 | all the mechanics of setting up a catch target, invoking the catch |
| 11166 | body, and perhaps invoking the handler if the body does a throw. |
| 11167 | |
| 11168 | The function is designed to be usable from C code, but is general |
| 11169 | enough to implement all the semantics Guile Scheme expects from throw. |
| 11170 | |
| 11171 | TAG is the catch tag. Typically, this is a symbol, but this function |
| 11172 | doesn't actually care about that. |
| 11173 | |
| 11174 | BODY is a pointer to a C function which runs the body of the catch; |
| 11175 | this is the code you can throw from. We call it like this: |
| 11176 | BODY (BODY_DATA, JMPBUF) |
| 11177 | where: |
| 11178 | BODY_DATA is just the BODY_DATA argument we received; we pass it |
| 11179 | through to BODY as its first argument. The caller can make |
| 11180 | BODY_DATA point to anything useful that BODY might need. |
| 11181 | JMPBUF is the Scheme jmpbuf object corresponding to this catch, |
| 11182 | which we have just created and initialized. |
| 11183 | |
| 11184 | HANDLER is a pointer to a C function to deal with a throw to TAG, |
| 11185 | should one occur. We call it like this: |
| 11186 | HANDLER (HANDLER_DATA, THROWN_TAG, THROW_ARGS) |
| 11187 | where |
| 11188 | HANDLER_DATA is the HANDLER_DATA argument we recevied; it's the |
| 11189 | same idea as BODY_DATA above. |
| 11190 | THROWN_TAG is the tag that the user threw to; usually this is |
| 11191 | TAG, but it could be something else if TAG was #t (i.e., a |
| 11192 | catch-all), or the user threw to a jmpbuf. |
| 11193 | THROW_ARGS is the list of arguments the user passed to the THROW |
| 11194 | function. |
| 11195 | |
| 11196 | BODY_DATA is just a pointer we pass through to BODY. HANDLER_DATA |
| 11197 | is just a pointer we pass through to HANDLER. We don't actually |
| 11198 | use either of those pointers otherwise ourselves. The idea is |
| 11199 | that, if our caller wants to communicate something to BODY or |
| 11200 | HANDLER, it can pass a pointer to it as MUMBLE_DATA, which BODY and |
| 11201 | HANDLER can then use. Think of it as a way to make BODY and |
| 11202 | HANDLER closures, not just functions; MUMBLE_DATA points to the |
| 11203 | enclosed variables. |
| 11204 | |
| 11205 | Of course, it's up to the caller to make sure that any data a |
| 11206 | MUMBLE_DATA needs is protected from GC. A common way to do this is |
| 11207 | to make MUMBLE_DATA a pointer to data stored in an automatic |
| 11208 | structure variable; since the collector must scan the stack for |
| 11209 | references anyway, this assures that any references in MUMBLE_DATA |
| 11210 | will be found. |
| 11211 | |
| 11212 | ** The new function scm_internal_lazy_catch is exactly like |
| 11213 | scm_internal_catch, except: |
| 11214 | |
| 11215 | - It does not unwind the stack (this is the major difference). |
| 11216 | - If handler returns, its value is returned from the throw. |
| 11217 | - BODY always receives #f as its JMPBUF argument (since there's no |
| 11218 | jmpbuf associated with a lazy catch, because we don't unwind the |
| 11219 | stack.) |
| 11220 | |
| 11221 | ** scm_body_thunk is a new body function you can pass to |
| 11222 | scm_internal_catch if you want the body to be like Scheme's `catch' |
| 11223 | --- a thunk, or a function of one argument if the tag is #f. |
| 11224 | |
| 11225 | BODY_DATA is a pointer to a scm_body_thunk_data structure, which |
| 11226 | contains the Scheme procedure to invoke as the body, and the tag |
| 11227 | we're catching. If the tag is #f, then we pass JMPBUF (created by |
| 11228 | scm_internal_catch) to the body procedure; otherwise, the body gets |
| 11229 | no arguments. |
| 11230 | |
| 11231 | ** scm_handle_by_proc is a new handler function you can pass to |
| 11232 | scm_internal_catch if you want the handler to act like Scheme's catch |
| 11233 | --- call a procedure with the tag and the throw arguments. |
| 11234 | |
| 11235 | If the user does a throw to this catch, this function runs a handler |
| 11236 | procedure written in Scheme. HANDLER_DATA is a pointer to an SCM |
| 11237 | variable holding the Scheme procedure object to invoke. It ought to |
| 11238 | be a pointer to an automatic variable (i.e., one living on the stack), |
| 11239 | or the procedure object should be otherwise protected from GC. |
| 11240 | |
| 11241 | ** scm_handle_by_message is a new handler function to use with |
| 11242 | `scm_internal_catch' if you want Guile to print a message and die. |
| 11243 | It's useful for dealing with throws to uncaught keys at the top level. |
| 11244 | |
| 11245 | HANDLER_DATA, if non-zero, is assumed to be a char * pointing to a |
| 11246 | message header to print; if zero, we use "guile" instead. That |
| 11247 | text is followed by a colon, then the message described by ARGS. |
| 11248 | |
| 11249 | ** The return type of scm_boot_guile is now void; the function does |
| 11250 | not return a value, and indeed, never returns at all. |
| 11251 | |
| 11252 | ** The new function scm_shell makes it easy for user applications to |
| 11253 | process command-line arguments in a way that is compatible with the |
| 11254 | stand-alone guile interpreter (which is in turn compatible with SCSH, |
| 11255 | the Scheme shell). |
| 11256 | |
| 11257 | To use the scm_shell function, first initialize any guile modules |
| 11258 | linked into your application, and then call scm_shell with the values |
| 11259 | of ARGC and ARGV your `main' function received. scm_shell will add |
| 11260 | any SCSH-style meta-arguments from the top of the script file to the |
| 11261 | argument vector, and then process the command-line arguments. This |
| 11262 | generally means loading a script file or starting up an interactive |
| 11263 | command interpreter. For details, see "Changes to the stand-alone |
| 11264 | interpreter" above. |
| 11265 | |
| 11266 | ** The new functions scm_get_meta_args and scm_count_argv help you |
| 11267 | implement the SCSH-style meta-argument, `\'. |
| 11268 | |
| 11269 | char **scm_get_meta_args (int ARGC, char **ARGV) |
| 11270 | If the second element of ARGV is a string consisting of a single |
| 11271 | backslash character (i.e. "\\" in Scheme notation), open the file |
| 11272 | named by the following argument, parse arguments from it, and return |
| 11273 | the spliced command line. The returned array is terminated by a |
| 11274 | null pointer. |
| 11275 | |
| 11276 | For details of argument parsing, see above, under "guile now accepts |
| 11277 | command-line arguments compatible with SCSH..." |
| 11278 | |
| 11279 | int scm_count_argv (char **ARGV) |
| 11280 | Count the arguments in ARGV, assuming it is terminated by a null |
| 11281 | pointer. |
| 11282 | |
| 11283 | For an example of how these functions might be used, see the source |
| 11284 | code for the function scm_shell in libguile/script.c. |
| 11285 | |
| 11286 | You will usually want to use scm_shell instead of calling this |
| 11287 | function yourself. |
| 11288 | |
| 11289 | ** The new function scm_compile_shell_switches turns an array of |
| 11290 | command-line arguments into Scheme code to carry out the actions they |
| 11291 | describe. Given ARGC and ARGV, it returns a Scheme expression to |
| 11292 | evaluate, and calls scm_set_program_arguments to make any remaining |
| 11293 | command-line arguments available to the Scheme code. For example, |
| 11294 | given the following arguments: |
| 11295 | |
| 11296 | -e main -s ekko a speckled gecko |
| 11297 | |
| 11298 | scm_set_program_arguments will return the following expression: |
| 11299 | |
| 11300 | (begin (load "ekko") (main (command-line)) (quit)) |
| 11301 | |
| 11302 | You will usually want to use scm_shell instead of calling this |
| 11303 | function yourself. |
| 11304 | |
| 11305 | ** The function scm_shell_usage prints a usage message appropriate for |
| 11306 | an interpreter that uses scm_compile_shell_switches to handle its |
| 11307 | command-line arguments. |
| 11308 | |
| 11309 | void scm_shell_usage (int FATAL, char *MESSAGE) |
| 11310 | Print a usage message to the standard error output. If MESSAGE is |
| 11311 | non-zero, write it before the usage message, followed by a newline. |
| 11312 | If FATAL is non-zero, exit the process, using FATAL as the |
| 11313 | termination status. (If you want to be compatible with Guile, |
| 11314 | always use 1 as the exit status when terminating due to command-line |
| 11315 | usage problems.) |
| 11316 | |
| 11317 | You will usually want to use scm_shell instead of calling this |
| 11318 | function yourself. |
| 11319 | |
| 11320 | ** scm_eval_0str now returns SCM_UNSPECIFIED if the string contains no |
| 11321 | expressions. It used to return SCM_EOL. Earth-shattering. |
| 11322 | |
| 11323 | ** The macros for declaring scheme objects in C code have been |
| 11324 | rearranged slightly. They are now: |
| 11325 | |
| 11326 | SCM_SYMBOL (C_NAME, SCHEME_NAME) |
| 11327 | Declare a static SCM variable named C_NAME, and initialize it to |
| 11328 | point to the Scheme symbol whose name is SCHEME_NAME. C_NAME should |
| 11329 | be a C identifier, and SCHEME_NAME should be a C string. |
| 11330 | |
| 11331 | SCM_GLOBAL_SYMBOL (C_NAME, SCHEME_NAME) |
| 11332 | Just like SCM_SYMBOL, but make C_NAME globally visible. |
| 11333 | |
| 11334 | SCM_VCELL (C_NAME, SCHEME_NAME) |
| 11335 | Create a global variable at the Scheme level named SCHEME_NAME. |
| 11336 | Declare a static SCM variable named C_NAME, and initialize it to |
| 11337 | point to the Scheme variable's value cell. |
| 11338 | |
| 11339 | SCM_GLOBAL_VCELL (C_NAME, SCHEME_NAME) |
| 11340 | Just like SCM_VCELL, but make C_NAME globally visible. |
| 11341 | |
| 11342 | The `guile-snarf' script writes initialization code for these macros |
| 11343 | to its standard output, given C source code as input. |
| 11344 | |
| 11345 | The SCM_GLOBAL macro is gone. |
| 11346 | |
| 11347 | ** The scm_read_line and scm_read_line_x functions have been replaced |
| 11348 | by Scheme code based on the %read-delimited! procedure (known to C |
| 11349 | code as scm_read_delimited_x). See its description above for more |
| 11350 | information. |
| 11351 | |
| 11352 | ** The function scm_sys_open has been renamed to scm_open. It now |
| 11353 | returns a port instead of an FD object. |
| 11354 | |
| 11355 | * The dynamic linking support has changed. For more information, see |
| 11356 | libguile/DYNAMIC-LINKING. |
| 11357 | |
| 11358 | \f |
| 11359 | Guile 1.0b3 |
| 11360 | |
| 11361 | User-visible changes from Thursday, September 5, 1996 until Guile 1.0 |
| 11362 | (Sun 5 Jan 1997): |
| 11363 | |
| 11364 | * Changes to the 'guile' program: |
| 11365 | |
| 11366 | ** Guile now loads some new files when it starts up. Guile first |
| 11367 | searches the load path for init.scm, and loads it if found. Then, if |
| 11368 | Guile is not being used to execute a script, and the user's home |
| 11369 | directory contains a file named `.guile', Guile loads that. |
| 11370 | |
| 11371 | ** You can now use Guile as a shell script interpreter. |
| 11372 | |
| 11373 | To paraphrase the SCSH manual: |
| 11374 | |
| 11375 | When Unix tries to execute an executable file whose first two |
| 11376 | characters are the `#!', it treats the file not as machine code to |
| 11377 | be directly executed by the native processor, but as source code |
| 11378 | to be executed by some interpreter. The interpreter to use is |
| 11379 | specified immediately after the #! sequence on the first line of |
| 11380 | the source file. The kernel reads in the name of the interpreter, |
| 11381 | and executes that instead. It passes the interpreter the source |
| 11382 | filename as its first argument, with the original arguments |
| 11383 | following. Consult the Unix man page for the `exec' system call |
| 11384 | for more information. |
| 11385 | |
| 11386 | Now you can use Guile as an interpreter, using a mechanism which is a |
| 11387 | compatible subset of that provided by SCSH. |
| 11388 | |
| 11389 | Guile now recognizes a '-s' command line switch, whose argument is the |
| 11390 | name of a file of Scheme code to load. It also treats the two |
| 11391 | characters `#!' as the start of a comment, terminated by `!#'. Thus, |
| 11392 | to make a file of Scheme code directly executable by Unix, insert the |
| 11393 | following two lines at the top of the file: |
| 11394 | |
| 11395 | #!/usr/local/bin/guile -s |
| 11396 | !# |
| 11397 | |
| 11398 | Guile treats the argument of the `-s' command-line switch as the name |
| 11399 | of a file of Scheme code to load, and treats the sequence `#!' as the |
| 11400 | start of a block comment, terminated by `!#'. |
| 11401 | |
| 11402 | For example, here's a version of 'echo' written in Scheme: |
| 11403 | |
| 11404 | #!/usr/local/bin/guile -s |
| 11405 | !# |
| 11406 | (let loop ((args (cdr (program-arguments)))) |
| 11407 | (if (pair? args) |
| 11408 | (begin |
| 11409 | (display (car args)) |
| 11410 | (if (pair? (cdr args)) |
| 11411 | (display " ")) |
| 11412 | (loop (cdr args))))) |
| 11413 | (newline) |
| 11414 | |
| 11415 | Why does `#!' start a block comment terminated by `!#', instead of the |
| 11416 | end of the line? That is the notation SCSH uses, and although we |
| 11417 | don't yet support the other SCSH features that motivate that choice, |
| 11418 | we would like to be backward-compatible with any existing Guile |
| 11419 | scripts once we do. Furthermore, if the path to Guile on your system |
| 11420 | is too long for your kernel, you can start the script with this |
| 11421 | horrible hack: |
| 11422 | |
| 11423 | #!/bin/sh |
| 11424 | exec /really/long/path/to/guile -s "$0" ${1+"$@"} |
| 11425 | !# |
| 11426 | |
| 11427 | Note that some very old Unix systems don't support the `#!' syntax. |
| 11428 | |
| 11429 | |
| 11430 | ** You can now run Guile without installing it. |
| 11431 | |
| 11432 | Previous versions of the interactive Guile interpreter (`guile') |
| 11433 | couldn't start up unless Guile's Scheme library had been installed; |
| 11434 | they used the value of the environment variable `SCHEME_LOAD_PATH' |
| 11435 | later on in the startup process, but not to find the startup code |
| 11436 | itself. Now Guile uses `SCHEME_LOAD_PATH' in all searches for Scheme |
| 11437 | code. |
| 11438 | |
| 11439 | To run Guile without installing it, build it in the normal way, and |
| 11440 | then set the environment variable `SCHEME_LOAD_PATH' to a |
| 11441 | colon-separated list of directories, including the top-level directory |
| 11442 | of the Guile sources. For example, if you unpacked Guile so that the |
| 11443 | full filename of this NEWS file is /home/jimb/guile-1.0b3/NEWS, then |
| 11444 | you might say |
| 11445 | |
| 11446 | export SCHEME_LOAD_PATH=/home/jimb/my-scheme:/home/jimb/guile-1.0b3 |
| 11447 | |
| 11448 | |
| 11449 | ** Guile's read-eval-print loop no longer prints #<unspecified> |
| 11450 | results. If the user wants to see this, she can evaluate the |
| 11451 | expression (assert-repl-print-unspecified #t), perhaps in her startup |
| 11452 | file. |
| 11453 | |
| 11454 | ** Guile no longer shows backtraces by default when an error occurs; |
| 11455 | however, it does display a message saying how to get one, and how to |
| 11456 | request that they be displayed by default. After an error, evaluate |
| 11457 | (backtrace) |
| 11458 | to see a backtrace, and |
| 11459 | (debug-enable 'backtrace) |
| 11460 | to see them by default. |
| 11461 | |
| 11462 | |
| 11463 | |
| 11464 | * Changes to Guile Scheme: |
| 11465 | |
| 11466 | ** Guile now distinguishes between #f and the empty list. |
| 11467 | |
| 11468 | This is for compatibility with the IEEE standard, the (possibly) |
| 11469 | upcoming Revised^5 Report on Scheme, and many extant Scheme |
| 11470 | implementations. |
| 11471 | |
| 11472 | Guile used to have #f and '() denote the same object, to make Scheme's |
| 11473 | type system more compatible with Emacs Lisp's. However, the change |
| 11474 | caused too much trouble for Scheme programmers, and we found another |
| 11475 | way to reconcile Emacs Lisp with Scheme that didn't require this. |
| 11476 | |
| 11477 | |
| 11478 | ** Guile's delq, delv, delete functions, and their destructive |
| 11479 | counterparts, delq!, delv!, and delete!, now remove all matching |
| 11480 | elements from the list, not just the first. This matches the behavior |
| 11481 | of the corresponding Emacs Lisp functions, and (I believe) the Maclisp |
| 11482 | functions which inspired them. |
| 11483 | |
| 11484 | I recognize that this change may break code in subtle ways, but it |
| 11485 | seems best to make the change before the FSF's first Guile release, |
| 11486 | rather than after. |
| 11487 | |
| 11488 | |
| 11489 | ** The compiled-library-path function has been deleted from libguile. |
| 11490 | |
| 11491 | ** The facilities for loading Scheme source files have changed. |
| 11492 | |
| 11493 | *** The variable %load-path now tells Guile which directories to search |
| 11494 | for Scheme code. Its value is a list of strings, each of which names |
| 11495 | a directory. |
| 11496 | |
| 11497 | *** The variable %load-extensions now tells Guile which extensions to |
| 11498 | try appending to a filename when searching the load path. Its value |
| 11499 | is a list of strings. Its default value is ("" ".scm"). |
| 11500 | |
| 11501 | *** (%search-load-path FILENAME) searches the directories listed in the |
| 11502 | value of the %load-path variable for a Scheme file named FILENAME, |
| 11503 | with all the extensions listed in %load-extensions. If it finds a |
| 11504 | match, then it returns its full filename. If FILENAME is absolute, it |
| 11505 | returns it unchanged. Otherwise, it returns #f. |
| 11506 | |
| 11507 | %search-load-path will not return matches that refer to directories. |
| 11508 | |
| 11509 | *** (primitive-load FILENAME :optional CASE-INSENSITIVE-P SHARP) |
| 11510 | uses %seach-load-path to find a file named FILENAME, and loads it if |
| 11511 | it finds it. If it can't read FILENAME for any reason, it throws an |
| 11512 | error. |
| 11513 | |
| 11514 | The arguments CASE-INSENSITIVE-P and SHARP are interpreted as by the |
| 11515 | `read' function. |
| 11516 | |
| 11517 | *** load uses the same searching semantics as primitive-load. |
| 11518 | |
| 11519 | *** The functions %try-load, try-load-with-path, %load, load-with-path, |
| 11520 | basic-try-load-with-path, basic-load-with-path, try-load-module-with- |
| 11521 | path, and load-module-with-path have been deleted. The functions |
| 11522 | above should serve their purposes. |
| 11523 | |
| 11524 | *** If the value of the variable %load-hook is a procedure, |
| 11525 | `primitive-load' applies its value to the name of the file being |
| 11526 | loaded (without the load path directory name prepended). If its value |
| 11527 | is #f, it is ignored. Otherwise, an error occurs. |
| 11528 | |
| 11529 | This is mostly useful for printing load notification messages. |
| 11530 | |
| 11531 | |
| 11532 | ** The function `eval!' is no longer accessible from the scheme level. |
| 11533 | We can't allow operations which introduce glocs into the scheme level, |
| 11534 | because Guile's type system can't handle these as data. Use `eval' or |
| 11535 | `read-and-eval!' (see below) as replacement. |
| 11536 | |
| 11537 | ** The new function read-and-eval! reads an expression from PORT, |
| 11538 | evaluates it, and returns the result. This is more efficient than |
| 11539 | simply calling `read' and `eval', since it is not necessary to make a |
| 11540 | copy of the expression for the evaluator to munge. |
| 11541 | |
| 11542 | Its optional arguments CASE_INSENSITIVE_P and SHARP are interpreted as |
| 11543 | for the `read' function. |
| 11544 | |
| 11545 | |
| 11546 | ** The function `int?' has been removed; its definition was identical |
| 11547 | to that of `integer?'. |
| 11548 | |
| 11549 | ** The functions `<?', `<?', `<=?', `=?', `>?', and `>=?'. Code should |
| 11550 | use the R4RS names for these functions. |
| 11551 | |
| 11552 | ** The function object-properties no longer returns the hash handle; |
| 11553 | it simply returns the object's property list. |
| 11554 | |
| 11555 | ** Many functions have been changed to throw errors, instead of |
| 11556 | returning #f on failure. The point of providing exception handling in |
| 11557 | the language is to simplify the logic of user code, but this is less |
| 11558 | useful if Guile's primitives don't throw exceptions. |
| 11559 | |
| 11560 | ** The function `fileno' has been renamed from `%fileno'. |
| 11561 | |
| 11562 | ** The function primitive-mode->fdes returns #t or #f now, not 1 or 0. |
| 11563 | |
| 11564 | |
| 11565 | * Changes to Guile's C interface: |
| 11566 | |
| 11567 | ** The library's initialization procedure has been simplified. |
| 11568 | scm_boot_guile now has the prototype: |
| 11569 | |
| 11570 | void scm_boot_guile (int ARGC, |
| 11571 | char **ARGV, |
| 11572 | void (*main_func) (), |
| 11573 | void *closure); |
| 11574 | |
| 11575 | scm_boot_guile calls MAIN_FUNC, passing it CLOSURE, ARGC, and ARGV. |
| 11576 | MAIN_FUNC should do all the work of the program (initializing other |
| 11577 | packages, reading user input, etc.) before returning. When MAIN_FUNC |
| 11578 | returns, call exit (0); this function never returns. If you want some |
| 11579 | other exit value, MAIN_FUNC may call exit itself. |
| 11580 | |
| 11581 | scm_boot_guile arranges for program-arguments to return the strings |
| 11582 | given by ARGC and ARGV. If MAIN_FUNC modifies ARGC/ARGV, should call |
| 11583 | scm_set_program_arguments with the final list, so Scheme code will |
| 11584 | know which arguments have been processed. |
| 11585 | |
| 11586 | scm_boot_guile establishes a catch-all catch handler which prints an |
| 11587 | error message and exits the process. This means that Guile exits in a |
| 11588 | coherent way when system errors occur and the user isn't prepared to |
| 11589 | handle it. If the user doesn't like this behavior, they can establish |
| 11590 | their own universal catcher in MAIN_FUNC to shadow this one. |
| 11591 | |
| 11592 | Why must the caller do all the real work from MAIN_FUNC? The garbage |
| 11593 | collector assumes that all local variables of type SCM will be above |
| 11594 | scm_boot_guile's stack frame on the stack. If you try to manipulate |
| 11595 | SCM values after this function returns, it's the luck of the draw |
| 11596 | whether the GC will be able to find the objects you allocate. So, |
| 11597 | scm_boot_guile function exits, rather than returning, to discourage |
| 11598 | people from making that mistake. |
| 11599 | |
| 11600 | The IN, OUT, and ERR arguments were removed; there are other |
| 11601 | convenient ways to override these when desired. |
| 11602 | |
| 11603 | The RESULT argument was deleted; this function should never return. |
| 11604 | |
| 11605 | The BOOT_CMD argument was deleted; the MAIN_FUNC argument is more |
| 11606 | general. |
| 11607 | |
| 11608 | |
| 11609 | ** Guile's header files should no longer conflict with your system's |
| 11610 | header files. |
| 11611 | |
| 11612 | In order to compile code which #included <libguile.h>, previous |
| 11613 | versions of Guile required you to add a directory containing all the |
| 11614 | Guile header files to your #include path. This was a problem, since |
| 11615 | Guile's header files have names which conflict with many systems' |
| 11616 | header files. |
| 11617 | |
| 11618 | Now only <libguile.h> need appear in your #include path; you must |
| 11619 | refer to all Guile's other header files as <libguile/mumble.h>. |
| 11620 | Guile's installation procedure puts libguile.h in $(includedir), and |
| 11621 | the rest in $(includedir)/libguile. |
| 11622 | |
| 11623 | |
| 11624 | ** Two new C functions, scm_protect_object and scm_unprotect_object, |
| 11625 | have been added to the Guile library. |
| 11626 | |
| 11627 | scm_protect_object (OBJ) protects OBJ from the garbage collector. |
| 11628 | OBJ will not be freed, even if all other references are dropped, |
| 11629 | until someone does scm_unprotect_object (OBJ). Both functions |
| 11630 | return OBJ. |
| 11631 | |
| 11632 | Note that calls to scm_protect_object do not nest. You can call |
| 11633 | scm_protect_object any number of times on a given object, and the |
| 11634 | next call to scm_unprotect_object will unprotect it completely. |
| 11635 | |
| 11636 | Basically, scm_protect_object and scm_unprotect_object just |
| 11637 | maintain a list of references to things. Since the GC knows about |
| 11638 | this list, all objects it mentions stay alive. scm_protect_object |
| 11639 | adds its argument to the list; scm_unprotect_object remove its |
| 11640 | argument from the list. |
| 11641 | |
| 11642 | |
| 11643 | ** scm_eval_0str now returns the value of the last expression |
| 11644 | evaluated. |
| 11645 | |
| 11646 | ** The new function scm_read_0str reads an s-expression from a |
| 11647 | null-terminated string, and returns it. |
| 11648 | |
| 11649 | ** The new function `scm_stdio_to_port' converts a STDIO file pointer |
| 11650 | to a Scheme port object. |
| 11651 | |
| 11652 | ** The new function `scm_set_program_arguments' allows C code to set |
| 11653 | the value returned by the Scheme `program-arguments' function. |
| 11654 | |
| 11655 | \f |
| 11656 | Older changes: |
| 11657 | |
| 11658 | * Guile no longer includes sophisticated Tcl/Tk support. |
| 11659 | |
| 11660 | The old Tcl/Tk support was unsatisfying to us, because it required the |
| 11661 | user to link against the Tcl library, as well as Tk and Guile. The |
| 11662 | interface was also un-lispy, in that it preserved Tcl/Tk's practice of |
| 11663 | referring to widgets by names, rather than exporting widgets to Scheme |
| 11664 | code as a special datatype. |
| 11665 | |
| 11666 | In the Usenix Tk Developer's Workshop held in July 1996, the Tcl/Tk |
| 11667 | maintainers described some very interesting changes in progress to the |
| 11668 | Tcl/Tk internals, which would facilitate clean interfaces between lone |
| 11669 | Tk and other interpreters --- even for garbage-collected languages |
| 11670 | like Scheme. They expected the new Tk to be publicly available in the |
| 11671 | fall of 1996. |
| 11672 | |
| 11673 | Since it seems that Guile might soon have a new, cleaner interface to |
| 11674 | lone Tk, and that the old Guile/Tk glue code would probably need to be |
| 11675 | completely rewritten, we (Jim Blandy and Richard Stallman) have |
| 11676 | decided not to support the old code. We'll spend the time instead on |
| 11677 | a good interface to the newer Tk, as soon as it is available. |
| 11678 | |
| 11679 | Until then, gtcltk-lib provides trivial, low-maintenance functionality. |
| 11680 | |
| 11681 | \f |
| 11682 | Copyright information: |
| 11683 | |
| 11684 | Copyright (C) 1996, 1997, 1998, 1999, 2000, 2001, 2006 Free Software Foundation, Inc. |
| 11685 | |
| 11686 | Permission is granted to anyone to make or distribute verbatim copies |
| 11687 | of this document as received, in any medium, provided that the |
| 11688 | copyright notice and this permission notice are preserved, |
| 11689 | thus giving the recipient permission to redistribute in turn. |
| 11690 | |
| 11691 | Permission is granted to distribute modified versions |
| 11692 | of this document, or of portions of it, |
| 11693 | under the above conditions, provided also that they |
| 11694 | carry prominent notices stating who last changed them. |
| 11695 | |
| 11696 | \f |
| 11697 | Local variables: |
| 11698 | mode: outline |
| 11699 | paragraph-separate: "[ \f]*$" |
| 11700 | end: |