| 1 | /* Copyright (C) 1995,1996,1997,1999,2000,2001 Free Software Foundation, Inc. |
| 2 | * |
| 3 | * This program is free software; you can redistribute it and/or modify |
| 4 | * it under the terms of the GNU General Public License as published by |
| 5 | * the Free Software Foundation; either version 2, or (at your option) |
| 6 | * any later version. |
| 7 | * |
| 8 | * This program is distributed in the hope that it will be useful, |
| 9 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 10 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 11 | * GNU General Public License for more details. |
| 12 | * |
| 13 | * You should have received a copy of the GNU General Public License |
| 14 | * along with this software; see the file COPYING. If not, write to |
| 15 | * the Free Software Foundation, Inc., 59 Temple Place, Suite 330, |
| 16 | * Boston, MA 02111-1307 USA |
| 17 | * |
| 18 | * As a special exception, the Free Software Foundation gives permission |
| 19 | * for additional uses of the text contained in its release of GUILE. |
| 20 | * |
| 21 | * The exception is that, if you link the GUILE library with other files |
| 22 | * to produce an executable, this does not by itself cause the |
| 23 | * resulting executable to be covered by the GNU General Public License. |
| 24 | * Your use of that executable is in no way restricted on account of |
| 25 | * linking the GUILE library code into it. |
| 26 | * |
| 27 | * This exception does not however invalidate any other reasons why |
| 28 | * the executable file might be covered by the GNU General Public License. |
| 29 | * |
| 30 | * This exception applies only to the code released by the |
| 31 | * Free Software Foundation under the name GUILE. If you copy |
| 32 | * code from other Free Software Foundation releases into a copy of |
| 33 | * GUILE, as the General Public License permits, the exception does |
| 34 | * not apply to the code that you add in this way. To avoid misleading |
| 35 | * anyone as to the status of such modified files, you must delete |
| 36 | * this exception notice from them. |
| 37 | * |
| 38 | * If you write modifications of your own for GUILE, it is your choice |
| 39 | * whether to permit this exception to apply to your modifications. |
| 40 | * If you do not wish that, delete this exception notice. */ |
| 41 | |
| 42 | |
| 43 | \f |
| 44 | |
| 45 | #include <stdio.h> |
| 46 | #include "libguile/_scm.h" |
| 47 | #include "libguile/chars.h" |
| 48 | #include "libguile/eval.h" |
| 49 | #include "libguile/unif.h" |
| 50 | #include "libguile/keywords.h" |
| 51 | #include "libguile/alist.h" |
| 52 | #include "libguile/srcprop.h" |
| 53 | #include "libguile/hashtab.h" |
| 54 | #include "libguile/hash.h" |
| 55 | #include "libguile/ports.h" |
| 56 | #include "libguile/root.h" |
| 57 | #include "libguile/strings.h" |
| 58 | #include "libguile/vectors.h" |
| 59 | |
| 60 | #include "libguile/validate.h" |
| 61 | #include "libguile/read.h" |
| 62 | |
| 63 | \f |
| 64 | |
| 65 | SCM_SYMBOL (scm_keyword_prefix, "prefix"); |
| 66 | |
| 67 | scm_t_option scm_read_opts[] = { |
| 68 | { SCM_OPTION_BOOLEAN, "copy", 0, |
| 69 | "Copy source code expressions." }, |
| 70 | { SCM_OPTION_BOOLEAN, "positions", 0, |
| 71 | "Record positions of source code expressions." }, |
| 72 | { SCM_OPTION_BOOLEAN, "case-insensitive", 0, |
| 73 | "Convert symbols to lower case."}, |
| 74 | { SCM_OPTION_SCM, "keywords", SCM_UNPACK (SCM_BOOL_F), |
| 75 | "Style of keyword recognition: #f or 'prefix."} |
| 76 | }; |
| 77 | |
| 78 | SCM_DEFINE (scm_read_options, "read-options-interface", 0, 1, 0, |
| 79 | (SCM setting), |
| 80 | "Option interface for the read options. Instead of using\n" |
| 81 | "this procedure directly, use the procedures @code{read-enable},\n" |
| 82 | "@code{read-disable}, @code{read-set!} and @var{read-options}.") |
| 83 | #define FUNC_NAME s_scm_read_options |
| 84 | { |
| 85 | SCM ans = scm_options (setting, |
| 86 | scm_read_opts, |
| 87 | SCM_N_READ_OPTIONS, |
| 88 | FUNC_NAME); |
| 89 | if (SCM_COPY_SOURCE_P) |
| 90 | SCM_RECORD_POSITIONS_P = 1; |
| 91 | return ans; |
| 92 | } |
| 93 | #undef FUNC_NAME |
| 94 | |
| 95 | /* An association list mapping extra hash characters to procedures. */ |
| 96 | static SCM *scm_read_hash_procedures; |
| 97 | |
| 98 | SCM_DEFINE (scm_read, "read", 0, 1, 0, |
| 99 | (SCM port), |
| 100 | "Read an s-expression from the input port @var{port}, or from\n" |
| 101 | "the current input port if @var{port} is not specified.\n" |
| 102 | "Any whitespace before the next token is discarded.") |
| 103 | #define FUNC_NAME s_scm_read |
| 104 | { |
| 105 | int c; |
| 106 | SCM tok_buf, copy; |
| 107 | |
| 108 | if (SCM_UNBNDP (port)) |
| 109 | port = scm_cur_inp; |
| 110 | SCM_VALIDATE_OPINPORT (1,port); |
| 111 | |
| 112 | c = scm_flush_ws (port, (char *) NULL); |
| 113 | if (EOF == c) |
| 114 | return SCM_EOF_VAL; |
| 115 | scm_ungetc (c, port); |
| 116 | |
| 117 | tok_buf = scm_allocate_string (30); |
| 118 | return scm_lreadr (&tok_buf, port, ©); |
| 119 | } |
| 120 | #undef FUNC_NAME |
| 121 | |
| 122 | |
| 123 | |
| 124 | char * |
| 125 | scm_grow_tok_buf (SCM *tok_buf) |
| 126 | { |
| 127 | size_t oldlen = SCM_STRING_LENGTH (*tok_buf); |
| 128 | SCM newstr = scm_allocate_string (2 * oldlen); |
| 129 | size_t i; |
| 130 | |
| 131 | for (i = 0; i != oldlen; ++i) |
| 132 | SCM_STRING_CHARS (newstr) [i] = SCM_STRING_CHARS (*tok_buf) [i]; |
| 133 | |
| 134 | *tok_buf = newstr; |
| 135 | return SCM_STRING_CHARS (newstr); |
| 136 | } |
| 137 | |
| 138 | |
| 139 | |
| 140 | int |
| 141 | scm_flush_ws (SCM port, const char *eoferr) |
| 142 | { |
| 143 | register int c; |
| 144 | while (1) |
| 145 | switch (c = scm_getc (port)) |
| 146 | { |
| 147 | case EOF: |
| 148 | goteof: |
| 149 | if (eoferr) |
| 150 | { |
| 151 | if (!SCM_FALSEP (SCM_FILENAME (port))) |
| 152 | scm_misc_error (eoferr, |
| 153 | "end of file in ~A", |
| 154 | scm_list_1 (SCM_FILENAME (port))); |
| 155 | else |
| 156 | scm_misc_error (eoferr, "end of file", SCM_EOL); |
| 157 | } |
| 158 | return c; |
| 159 | case ';': |
| 160 | lp: |
| 161 | switch (c = scm_getc (port)) |
| 162 | { |
| 163 | case EOF: |
| 164 | goto goteof; |
| 165 | default: |
| 166 | goto lp; |
| 167 | case SCM_LINE_INCREMENTORS: |
| 168 | break; |
| 169 | } |
| 170 | break; |
| 171 | case SCM_LINE_INCREMENTORS: |
| 172 | case SCM_SINGLE_SPACES: |
| 173 | case '\t': |
| 174 | break; |
| 175 | default: |
| 176 | return c; |
| 177 | } |
| 178 | } |
| 179 | |
| 180 | |
| 181 | |
| 182 | int |
| 183 | scm_casei_streq (char *s1, char *s2) |
| 184 | { |
| 185 | while (*s1 && *s2) |
| 186 | if (scm_downcase((int)*s1) != scm_downcase((int)*s2)) |
| 187 | return 0; |
| 188 | else |
| 189 | { |
| 190 | ++s1; |
| 191 | ++s2; |
| 192 | } |
| 193 | return !(*s1 || *s2); |
| 194 | } |
| 195 | |
| 196 | |
| 197 | /* recsexpr is used when recording expressions |
| 198 | * constructed by read:sharp. |
| 199 | */ |
| 200 | #ifndef DEBUG_EXTENSIONS |
| 201 | #define recsexpr(obj, line, column, filename) (obj) |
| 202 | #else |
| 203 | static SCM |
| 204 | recsexpr (SCM obj, long line, int column, SCM filename) |
| 205 | { |
| 206 | if (!SCM_CONSP(obj)) { |
| 207 | return obj; |
| 208 | } else { |
| 209 | SCM tmp = obj, copy; |
| 210 | /* If this sexpr is visible in the read:sharp source, we want to |
| 211 | keep that information, so only record non-constant cons cells |
| 212 | which haven't previously been read by the reader. */ |
| 213 | if (SCM_FALSEP (scm_whash_lookup (scm_source_whash, obj))) |
| 214 | { |
| 215 | if (SCM_COPY_SOURCE_P) |
| 216 | { |
| 217 | copy = scm_cons (recsexpr (SCM_CAR (obj), line, column, filename), |
| 218 | SCM_UNDEFINED); |
| 219 | while ((tmp = SCM_CDR (tmp)) && SCM_CONSP (tmp)) |
| 220 | { |
| 221 | SCM_SETCDR (copy, scm_cons (recsexpr (SCM_CAR (tmp), |
| 222 | line, |
| 223 | column, |
| 224 | filename), |
| 225 | SCM_UNDEFINED)); |
| 226 | copy = SCM_CDR (copy); |
| 227 | } |
| 228 | SCM_SETCDR (copy, tmp); |
| 229 | } |
| 230 | else |
| 231 | { |
| 232 | recsexpr (SCM_CAR (obj), line, column, filename); |
| 233 | while ((tmp = SCM_CDR (tmp)) && SCM_CONSP (tmp)) |
| 234 | recsexpr (SCM_CAR (tmp), line, column, filename); |
| 235 | copy = SCM_UNDEFINED; |
| 236 | } |
| 237 | scm_whash_insert (scm_source_whash, |
| 238 | obj, |
| 239 | scm_make_srcprops (line, |
| 240 | column, |
| 241 | filename, |
| 242 | copy, |
| 243 | SCM_EOL)); |
| 244 | } |
| 245 | return obj; |
| 246 | } |
| 247 | } |
| 248 | #endif |
| 249 | |
| 250 | /* Consume an SCSH-style block comment. Assume that we've already |
| 251 | read the initial `#!', and eat characters until we get a |
| 252 | newline/exclamation-point/sharp-sign/newline sequence. */ |
| 253 | |
| 254 | static void |
| 255 | skip_scsh_block_comment (SCM port) |
| 256 | #define FUNC_NAME "skip_scsh_block_comment" |
| 257 | { |
| 258 | /* Is this portable? Dear God, spare me from the non-eight-bit |
| 259 | characters. But is it tasteful? */ |
| 260 | long history = 0; |
| 261 | |
| 262 | for (;;) |
| 263 | { |
| 264 | int c = scm_getc (port); |
| 265 | |
| 266 | if (c == EOF) |
| 267 | SCM_MISC_ERROR ("unterminated `#! ... !#' comment", SCM_EOL); |
| 268 | history = ((history << 8) | (c & 0xff)) & 0xffffffff; |
| 269 | |
| 270 | /* Were the last four characters read "\n!#\n"? */ |
| 271 | if (history == (('\n' << 24) | ('!' << 16) | ('#' << 8) | '\n')) |
| 272 | return; |
| 273 | } |
| 274 | } |
| 275 | #undef FUNC_NAME |
| 276 | |
| 277 | |
| 278 | static SCM scm_get_hash_procedure(int c); |
| 279 | |
| 280 | static char s_list[]="list"; |
| 281 | |
| 282 | SCM |
| 283 | scm_lreadr (SCM *tok_buf,SCM port,SCM *copy) |
| 284 | #define FUNC_NAME "scm_lreadr" |
| 285 | { |
| 286 | int c; |
| 287 | size_t j; |
| 288 | SCM p; |
| 289 | |
| 290 | tryagain: |
| 291 | c = scm_flush_ws (port, s_scm_read); |
| 292 | tryagain_no_flush_ws: |
| 293 | switch (c) |
| 294 | { |
| 295 | case EOF: |
| 296 | return SCM_EOF_VAL; |
| 297 | |
| 298 | case '(': |
| 299 | return SCM_RECORD_POSITIONS_P |
| 300 | ? scm_lreadrecparen (tok_buf, port, s_list, copy) |
| 301 | : scm_lreadparen (tok_buf, port, s_list, copy); |
| 302 | case ')': |
| 303 | SCM_MISC_ERROR ("unexpected \")\"", SCM_EOL); |
| 304 | goto tryagain; |
| 305 | |
| 306 | case '\'': |
| 307 | p = scm_sym_quote; |
| 308 | goto recquote; |
| 309 | case '`': |
| 310 | p = scm_sym_quasiquote; |
| 311 | goto recquote; |
| 312 | case ',': |
| 313 | c = scm_getc (port); |
| 314 | if ('@' == c) |
| 315 | p = scm_sym_uq_splicing; |
| 316 | else |
| 317 | { |
| 318 | scm_ungetc (c, port); |
| 319 | p = scm_sym_unquote; |
| 320 | } |
| 321 | recquote: |
| 322 | p = scm_cons2 (p, |
| 323 | scm_lreadr (tok_buf, port, copy), |
| 324 | SCM_EOL); |
| 325 | if (SCM_RECORD_POSITIONS_P) |
| 326 | scm_whash_insert (scm_source_whash, |
| 327 | p, |
| 328 | scm_make_srcprops (SCM_LINUM (port), |
| 329 | SCM_COL (port) - 1, |
| 330 | SCM_FILENAME (port), |
| 331 | SCM_COPY_SOURCE_P |
| 332 | ? (*copy = scm_cons2 (SCM_CAR (p), |
| 333 | SCM_CAR (SCM_CDR (p)), |
| 334 | SCM_EOL)) |
| 335 | : SCM_UNDEFINED, |
| 336 | SCM_EOL)); |
| 337 | return p; |
| 338 | case '#': |
| 339 | c = scm_getc (port); |
| 340 | |
| 341 | { |
| 342 | /* Check for user-defined hash procedure first, to allow |
| 343 | overriding of builtin hash read syntaxes. */ |
| 344 | SCM sharp = scm_get_hash_procedure (c); |
| 345 | if (!SCM_FALSEP (sharp)) |
| 346 | { |
| 347 | int line = SCM_LINUM (port); |
| 348 | int column = SCM_COL (port) - 2; |
| 349 | SCM got; |
| 350 | |
| 351 | got = scm_call_2 (sharp, SCM_MAKE_CHAR (c), port); |
| 352 | if (SCM_EQ_P (got, SCM_UNSPECIFIED)) |
| 353 | goto handle_sharp; |
| 354 | if (SCM_RECORD_POSITIONS_P) |
| 355 | return *copy = recsexpr (got, line, column, |
| 356 | SCM_FILENAME (port)); |
| 357 | else |
| 358 | return got; |
| 359 | } |
| 360 | } |
| 361 | handle_sharp: |
| 362 | switch (c) |
| 363 | { |
| 364 | case '(': |
| 365 | p = scm_lreadparen (tok_buf, port, "vector", copy); |
| 366 | return SCM_NULLP (p) ? scm_nullvect : scm_vector (p); |
| 367 | |
| 368 | case 't': |
| 369 | case 'T': |
| 370 | return SCM_BOOL_T; |
| 371 | case 'f': |
| 372 | case 'F': |
| 373 | return SCM_BOOL_F; |
| 374 | |
| 375 | case 'b': |
| 376 | case 'B': |
| 377 | case 'o': |
| 378 | case 'O': |
| 379 | case 'd': |
| 380 | case 'D': |
| 381 | case 'x': |
| 382 | case 'X': |
| 383 | case 'i': |
| 384 | case 'I': |
| 385 | case 'e': |
| 386 | case 'E': |
| 387 | scm_ungetc (c, port); |
| 388 | c = '#'; |
| 389 | goto num; |
| 390 | |
| 391 | case '!': |
| 392 | /* start of a shell script. Parse as a block comment, |
| 393 | terminated by !#, just like SCSH. */ |
| 394 | skip_scsh_block_comment (port); |
| 395 | /* EOF is not an error here */ |
| 396 | c = scm_flush_ws (port, (char *)NULL); |
| 397 | goto tryagain_no_flush_ws; |
| 398 | |
| 399 | #ifdef HAVE_ARRAYS |
| 400 | case '*': |
| 401 | j = scm_read_token (c, tok_buf, port, 0); |
| 402 | p = scm_istr2bve (SCM_STRING_CHARS (*tok_buf) + 1, (long) (j - 1)); |
| 403 | if (!SCM_FALSEP (p)) |
| 404 | return p; |
| 405 | else |
| 406 | goto unkshrp; |
| 407 | #endif |
| 408 | |
| 409 | case '{': |
| 410 | j = scm_read_token (c, tok_buf, port, 1); |
| 411 | return scm_mem2symbol (SCM_STRING_CHARS (*tok_buf), j); |
| 412 | |
| 413 | case '\\': |
| 414 | c = scm_getc (port); |
| 415 | j = scm_read_token (c, tok_buf, port, 0); |
| 416 | if (j == 1) |
| 417 | return SCM_MAKE_CHAR (c); |
| 418 | if (c >= '0' && c < '8') |
| 419 | { |
| 420 | /* Dirk:FIXME:: This type of character syntax is not R5RS |
| 421 | * compliant. Further, it should be verified that the constant |
| 422 | * does only consist of octal digits. Finally, it should be |
| 423 | * checked whether the resulting fixnum is in the range of |
| 424 | * characters. */ |
| 425 | p = scm_i_mem2number (SCM_STRING_CHARS (*tok_buf), j, 8); |
| 426 | if (SCM_INUMP (p)) |
| 427 | return SCM_MAKE_CHAR (SCM_INUM (p)); |
| 428 | } |
| 429 | for (c = 0; c < scm_n_charnames; c++) |
| 430 | if (scm_charnames[c] |
| 431 | && (scm_casei_streq (scm_charnames[c], SCM_STRING_CHARS (*tok_buf)))) |
| 432 | return SCM_MAKE_CHAR (scm_charnums[c]); |
| 433 | SCM_MISC_ERROR ("unknown # object", SCM_EOL); |
| 434 | |
| 435 | /* #:SYMBOL is a syntax for keywords supported in all contexts. */ |
| 436 | case ':': |
| 437 | j = scm_read_token ('-', tok_buf, port, 0); |
| 438 | p = scm_mem2symbol (SCM_STRING_CHARS (*tok_buf), j); |
| 439 | return scm_make_keyword_from_dash_symbol (p); |
| 440 | |
| 441 | default: |
| 442 | callshrp: |
| 443 | { |
| 444 | SCM sharp = scm_get_hash_procedure (c); |
| 445 | |
| 446 | if (!SCM_FALSEP (sharp)) |
| 447 | { |
| 448 | int line = SCM_LINUM (port); |
| 449 | int column = SCM_COL (port) - 2; |
| 450 | SCM got; |
| 451 | |
| 452 | got = scm_call_2 (sharp, SCM_MAKE_CHAR (c), port); |
| 453 | if (SCM_EQ_P (got, SCM_UNSPECIFIED)) |
| 454 | goto unkshrp; |
| 455 | if (SCM_RECORD_POSITIONS_P) |
| 456 | return *copy = recsexpr (got, line, column, |
| 457 | SCM_FILENAME (port)); |
| 458 | else |
| 459 | return got; |
| 460 | } |
| 461 | } |
| 462 | unkshrp: |
| 463 | scm_misc_error (s_scm_read, "Unknown # object: ~S", |
| 464 | scm_list_1 (SCM_MAKE_CHAR (c))); |
| 465 | } |
| 466 | |
| 467 | case '"': |
| 468 | j = 0; |
| 469 | while ('"' != (c = scm_getc (port))) |
| 470 | { |
| 471 | if (c == EOF) |
| 472 | SCM_MISC_ERROR ("end of file in string constant", SCM_EOL); |
| 473 | |
| 474 | while (j + 2 >= SCM_STRING_LENGTH (*tok_buf)) |
| 475 | scm_grow_tok_buf (tok_buf); |
| 476 | |
| 477 | if (c == '\\') |
| 478 | switch (c = scm_getc (port)) |
| 479 | { |
| 480 | case '\n': |
| 481 | continue; |
| 482 | case '0': |
| 483 | c = '\0'; |
| 484 | break; |
| 485 | case 'f': |
| 486 | c = '\f'; |
| 487 | break; |
| 488 | case 'n': |
| 489 | c = '\n'; |
| 490 | break; |
| 491 | case 'r': |
| 492 | c = '\r'; |
| 493 | break; |
| 494 | case 't': |
| 495 | c = '\t'; |
| 496 | break; |
| 497 | case 'a': |
| 498 | c = '\007'; |
| 499 | break; |
| 500 | case 'v': |
| 501 | c = '\v'; |
| 502 | break; |
| 503 | } |
| 504 | SCM_STRING_CHARS (*tok_buf)[j] = c; |
| 505 | ++j; |
| 506 | } |
| 507 | if (j == 0) |
| 508 | return scm_nullstr; |
| 509 | SCM_STRING_CHARS (*tok_buf)[j] = 0; |
| 510 | return scm_mem2string (SCM_STRING_CHARS (*tok_buf), j); |
| 511 | |
| 512 | case '0': case '1': case '2': case '3': case '4': |
| 513 | case '5': case '6': case '7': case '8': case '9': |
| 514 | case '.': |
| 515 | case '-': |
| 516 | case '+': |
| 517 | num: |
| 518 | j = scm_read_token (c, tok_buf, port, 0); |
| 519 | if (j == 1 && (c == '+' || c == '-')) |
| 520 | /* Shortcut: Detected symbol '+ or '- */ |
| 521 | goto tok; |
| 522 | |
| 523 | p = scm_i_mem2number (SCM_STRING_CHARS (*tok_buf), j, 10); |
| 524 | if (!SCM_FALSEP (p)) |
| 525 | return p; |
| 526 | if (c == '#') |
| 527 | { |
| 528 | if ((j == 2) && (scm_getc (port) == '(')) |
| 529 | { |
| 530 | scm_ungetc ('(', port); |
| 531 | c = SCM_STRING_CHARS (*tok_buf)[1]; |
| 532 | goto callshrp; |
| 533 | } |
| 534 | SCM_MISC_ERROR ("unknown # object", SCM_EOL); |
| 535 | } |
| 536 | goto tok; |
| 537 | |
| 538 | case ':': |
| 539 | if (SCM_EQ_P (SCM_PACK (SCM_KEYWORD_STYLE), scm_keyword_prefix)) |
| 540 | { |
| 541 | j = scm_read_token ('-', tok_buf, port, 0); |
| 542 | p = scm_mem2symbol (SCM_STRING_CHARS (*tok_buf), j); |
| 543 | return scm_make_keyword_from_dash_symbol (p); |
| 544 | } |
| 545 | /* fallthrough */ |
| 546 | default: |
| 547 | j = scm_read_token (c, tok_buf, port, 0); |
| 548 | /* fallthrough */ |
| 549 | |
| 550 | tok: |
| 551 | return scm_mem2symbol (SCM_STRING_CHARS (*tok_buf), j); |
| 552 | } |
| 553 | } |
| 554 | #undef FUNC_NAME |
| 555 | |
| 556 | |
| 557 | #ifdef _UNICOS |
| 558 | _Pragma ("noopt"); /* # pragma _CRI noopt */ |
| 559 | #endif |
| 560 | |
| 561 | size_t |
| 562 | scm_read_token (int ic, SCM *tok_buf, SCM port, int weird) |
| 563 | { |
| 564 | register size_t j; |
| 565 | register int c; |
| 566 | register char *p; |
| 567 | |
| 568 | c = (SCM_CASE_INSENSITIVE_P ? scm_downcase(ic) : ic); |
| 569 | p = SCM_STRING_CHARS (*tok_buf); |
| 570 | |
| 571 | if (weird) |
| 572 | j = 0; |
| 573 | else |
| 574 | { |
| 575 | j = 0; |
| 576 | while (j + 2 >= SCM_STRING_LENGTH (*tok_buf)) |
| 577 | p = scm_grow_tok_buf (tok_buf); |
| 578 | p[j] = c; |
| 579 | ++j; |
| 580 | } |
| 581 | |
| 582 | while (1) |
| 583 | { |
| 584 | while (j + 2 >= SCM_STRING_LENGTH (*tok_buf)) |
| 585 | p = scm_grow_tok_buf (tok_buf); |
| 586 | c = scm_getc (port); |
| 587 | switch (c) |
| 588 | { |
| 589 | case '(': |
| 590 | case ')': |
| 591 | case '"': |
| 592 | case ';': |
| 593 | case SCM_WHITE_SPACES: |
| 594 | case SCM_LINE_INCREMENTORS: |
| 595 | if (weird) |
| 596 | goto default_case; |
| 597 | |
| 598 | scm_ungetc (c, port); |
| 599 | case EOF: |
| 600 | eof_case: |
| 601 | p[j] = 0; |
| 602 | return j; |
| 603 | case '\\': |
| 604 | if (!weird) |
| 605 | goto default_case; |
| 606 | else |
| 607 | { |
| 608 | c = scm_getc (port); |
| 609 | if (c == EOF) |
| 610 | goto eof_case; |
| 611 | else |
| 612 | goto default_case; |
| 613 | } |
| 614 | case '}': |
| 615 | if (!weird) |
| 616 | goto default_case; |
| 617 | |
| 618 | c = scm_getc (port); |
| 619 | if (c == '#') |
| 620 | { |
| 621 | p[j] = 0; |
| 622 | return j; |
| 623 | } |
| 624 | else |
| 625 | { |
| 626 | scm_ungetc (c, port); |
| 627 | c = '}'; |
| 628 | goto default_case; |
| 629 | } |
| 630 | |
| 631 | default: |
| 632 | default_case: |
| 633 | { |
| 634 | c = (SCM_CASE_INSENSITIVE_P ? scm_downcase(c) : c); |
| 635 | p[j] = c; |
| 636 | ++j; |
| 637 | } |
| 638 | |
| 639 | } |
| 640 | } |
| 641 | } |
| 642 | |
| 643 | #ifdef _UNICOS |
| 644 | _Pragma ("opt"); /* # pragma _CRI opt */ |
| 645 | #endif |
| 646 | |
| 647 | SCM |
| 648 | scm_lreadparen (SCM *tok_buf, SCM port, char *name, SCM *copy) |
| 649 | #define FUNC_NAME "scm_lreadparen" |
| 650 | { |
| 651 | SCM tmp; |
| 652 | SCM tl; |
| 653 | SCM ans; |
| 654 | int c; |
| 655 | |
| 656 | c = scm_flush_ws (port, name); |
| 657 | if (')' == c) |
| 658 | return SCM_EOL; |
| 659 | scm_ungetc (c, port); |
| 660 | if (SCM_EQ_P (scm_sym_dot, (tmp = scm_lreadr (tok_buf, port, copy)))) |
| 661 | { |
| 662 | ans = scm_lreadr (tok_buf, port, copy); |
| 663 | closeit: |
| 664 | if (')' != (c = scm_flush_ws (port, name))) |
| 665 | SCM_MISC_ERROR ("missing close paren", SCM_EOL); |
| 666 | return ans; |
| 667 | } |
| 668 | ans = tl = scm_cons (tmp, SCM_EOL); |
| 669 | while (')' != (c = scm_flush_ws (port, name))) |
| 670 | { |
| 671 | scm_ungetc (c, port); |
| 672 | if (SCM_EQ_P (scm_sym_dot, (tmp = scm_lreadr (tok_buf, port, copy)))) |
| 673 | { |
| 674 | SCM_SETCDR (tl, scm_lreadr (tok_buf, port, copy)); |
| 675 | goto closeit; |
| 676 | } |
| 677 | SCM_SETCDR (tl, scm_cons (tmp, SCM_EOL)); |
| 678 | tl = SCM_CDR (tl); |
| 679 | } |
| 680 | return ans; |
| 681 | } |
| 682 | #undef FUNC_NAME |
| 683 | |
| 684 | |
| 685 | SCM |
| 686 | scm_lreadrecparen (SCM *tok_buf, SCM port, char *name, SCM *copy) |
| 687 | #define FUNC_NAME "scm_lreadrecparen" |
| 688 | { |
| 689 | register int c; |
| 690 | register SCM tmp; |
| 691 | register SCM tl, tl2 = SCM_EOL; |
| 692 | SCM ans, ans2 = SCM_EOL; |
| 693 | /* Need to capture line and column numbers here. */ |
| 694 | int line = SCM_LINUM (port); |
| 695 | int column = SCM_COL (port) - 1; |
| 696 | |
| 697 | c = scm_flush_ws (port, name); |
| 698 | if (')' == c) |
| 699 | return SCM_EOL; |
| 700 | scm_ungetc (c, port); |
| 701 | if (SCM_EQ_P (scm_sym_dot, (tmp = scm_lreadr (tok_buf, port, copy)))) |
| 702 | { |
| 703 | ans = scm_lreadr (tok_buf, port, copy); |
| 704 | if (')' != (c = scm_flush_ws (port, name))) |
| 705 | SCM_MISC_ERROR ("missing close paren", SCM_EOL); |
| 706 | return ans; |
| 707 | } |
| 708 | /* Build the head of the list structure. */ |
| 709 | ans = tl = scm_cons (tmp, SCM_EOL); |
| 710 | if (SCM_COPY_SOURCE_P) |
| 711 | ans2 = tl2 = scm_cons (SCM_CONSP (tmp) |
| 712 | ? *copy |
| 713 | : tmp, |
| 714 | SCM_EOL); |
| 715 | while (')' != (c = scm_flush_ws (port, name))) |
| 716 | { |
| 717 | SCM new_tail; |
| 718 | |
| 719 | scm_ungetc (c, port); |
| 720 | if (SCM_EQ_P (scm_sym_dot, (tmp = scm_lreadr (tok_buf, port, copy)))) |
| 721 | { |
| 722 | SCM_SETCDR (tl, tmp = scm_lreadr (tok_buf, port, copy)); |
| 723 | if (SCM_COPY_SOURCE_P) |
| 724 | SCM_SETCDR (tl2, scm_cons (SCM_CONSP (tmp) |
| 725 | ? *copy |
| 726 | : tmp, |
| 727 | SCM_EOL)); |
| 728 | if (')' != (c = scm_flush_ws (port, name))) |
| 729 | SCM_MISC_ERROR ("missing close paren", SCM_EOL); |
| 730 | goto exit; |
| 731 | } |
| 732 | |
| 733 | new_tail = scm_cons (tmp, SCM_EOL); |
| 734 | SCM_SETCDR (tl, new_tail); |
| 735 | tl = new_tail; |
| 736 | |
| 737 | if (SCM_COPY_SOURCE_P) |
| 738 | { |
| 739 | SCM new_tail2 = scm_cons (SCM_CONSP (tmp) ? *copy : tmp, SCM_EOL); |
| 740 | SCM_SETCDR (tl2, new_tail2); |
| 741 | tl2 = new_tail2; |
| 742 | } |
| 743 | } |
| 744 | exit: |
| 745 | scm_whash_insert (scm_source_whash, |
| 746 | ans, |
| 747 | scm_make_srcprops (line, |
| 748 | column, |
| 749 | SCM_FILENAME (port), |
| 750 | SCM_COPY_SOURCE_P |
| 751 | ? *copy = ans2 |
| 752 | : SCM_UNDEFINED, |
| 753 | SCM_EOL)); |
| 754 | return ans; |
| 755 | } |
| 756 | #undef FUNC_NAME |
| 757 | |
| 758 | |
| 759 | \f |
| 760 | |
| 761 | /* Manipulate the read-hash-procedures alist. This could be written in |
| 762 | Scheme, but maybe it will also be used by C code during initialisation. */ |
| 763 | SCM_DEFINE (scm_read_hash_extend, "read-hash-extend", 2, 0, 0, |
| 764 | (SCM chr, SCM proc), |
| 765 | "Install the procedure @var{proc} for reading expressions\n" |
| 766 | "starting with the character sequence @code{#} and @var{chr}.\n" |
| 767 | "@var{proc} will be called with two arguments: the character\n" |
| 768 | "@var{chr} and the port to read further data from. The object\n" |
| 769 | "returned will be the return value of @code{read}.") |
| 770 | #define FUNC_NAME s_scm_read_hash_extend |
| 771 | { |
| 772 | SCM this; |
| 773 | SCM prev; |
| 774 | |
| 775 | SCM_VALIDATE_CHAR (1, chr); |
| 776 | SCM_ASSERT (SCM_FALSEP (proc) |
| 777 | || SCM_EQ_P (scm_procedure_p (proc), SCM_BOOL_T), |
| 778 | proc, SCM_ARG2, FUNC_NAME); |
| 779 | |
| 780 | /* Check if chr is already in the alist. */ |
| 781 | this = *scm_read_hash_procedures; |
| 782 | prev = SCM_BOOL_F; |
| 783 | while (1) |
| 784 | { |
| 785 | if (SCM_NULLP (this)) |
| 786 | { |
| 787 | /* not found, so add it to the beginning. */ |
| 788 | if (!SCM_FALSEP (proc)) |
| 789 | { |
| 790 | *scm_read_hash_procedures = |
| 791 | scm_cons (scm_cons (chr, proc), *scm_read_hash_procedures); |
| 792 | } |
| 793 | break; |
| 794 | } |
| 795 | if (SCM_EQ_P (chr, SCM_CAAR (this))) |
| 796 | { |
| 797 | /* already in the alist. */ |
| 798 | if (SCM_FALSEP (proc)) |
| 799 | { |
| 800 | /* remove it. */ |
| 801 | if (SCM_FALSEP (prev)) |
| 802 | { |
| 803 | *scm_read_hash_procedures = |
| 804 | SCM_CDR (*scm_read_hash_procedures); |
| 805 | } |
| 806 | else |
| 807 | scm_set_cdr_x (prev, SCM_CDR (this)); |
| 808 | } |
| 809 | else |
| 810 | { |
| 811 | /* replace it. */ |
| 812 | scm_set_cdr_x (SCM_CAR (this), proc); |
| 813 | } |
| 814 | break; |
| 815 | } |
| 816 | prev = this; |
| 817 | this = SCM_CDR (this); |
| 818 | } |
| 819 | |
| 820 | return SCM_UNSPECIFIED; |
| 821 | } |
| 822 | #undef FUNC_NAME |
| 823 | |
| 824 | /* Recover the read-hash procedure corresponding to char c. */ |
| 825 | static SCM |
| 826 | scm_get_hash_procedure (int c) |
| 827 | { |
| 828 | SCM rest = *scm_read_hash_procedures; |
| 829 | |
| 830 | while (1) |
| 831 | { |
| 832 | if (SCM_NULLP (rest)) |
| 833 | return SCM_BOOL_F; |
| 834 | |
| 835 | if (SCM_CHAR (SCM_CAAR (rest)) == c) |
| 836 | return SCM_CDAR (rest); |
| 837 | |
| 838 | rest = SCM_CDR (rest); |
| 839 | } |
| 840 | } |
| 841 | |
| 842 | void |
| 843 | scm_init_read () |
| 844 | { |
| 845 | scm_read_hash_procedures = |
| 846 | SCM_VARIABLE_LOC (scm_c_define ("read-hash-procedures", SCM_EOL)); |
| 847 | |
| 848 | scm_init_opts (scm_read_options, scm_read_opts, SCM_N_READ_OPTIONS); |
| 849 | #include "libguile/read.x" |
| 850 | } |
| 851 | |
| 852 | /* |
| 853 | Local Variables: |
| 854 | c-file-style: "gnu" |
| 855 | End: |
| 856 | */ |