2 @c This is part of the GNU Guile Reference Manual.
3 @c Copyright (C) 1996, 1997, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2009, 2010
4 @c Free Software Foundation, Inc.
5 @c See the file guile.texi for copying conditions.
7 @node Compound Data Types
8 @section Compound Data Types
10 This chapter describes Guile's compound data types. By @dfn{compound}
11 we mean that the primary purpose of these data types is to act as
12 containers for other kinds of data (including other compound objects).
13 For instance, a (non-uniform) vector with length 5 is a container that
14 can hold five arbitrary Scheme objects.
16 The various kinds of container object differ from each other in how
17 their memory is allocated, how they are indexed, and how particular
18 values can be looked up within them.
21 * Pairs:: Scheme's basic building block.
22 * Lists:: Special list functions supported by Guile.
23 * Vectors:: One-dimensional arrays of Scheme objects.
24 * Bit Vectors:: Vectors of bits.
25 * Generalized Vectors:: Treating all vector-like things uniformly.
26 * Arrays:: Matrices, etc.
27 * VLists:: Vector-like lists.
30 * Dictionary Types:: About dictionary types in general.
31 * Association Lists:: List-based dictionaries.
32 * VHashes:: VList-based dictionaries.
33 * Hash Tables:: Table-based dictionaries.
41 Pairs are used to combine two Scheme objects into one compound object.
42 Hence the name: A pair stores a pair of objects.
44 The data type @dfn{pair} is extremely important in Scheme, just like in
45 any other Lisp dialect. The reason is that pairs are not only used to
46 make two values available as one object, but that pairs are used for
47 constructing lists of values. Because lists are so important in Scheme,
48 they are described in a section of their own (@pxref{Lists}).
50 Pairs can literally get entered in source code or at the REPL, in the
51 so-called @dfn{dotted list} syntax. This syntax consists of an opening
52 parentheses, the first element of the pair, a dot, the second element
53 and a closing parentheses. The following example shows how a pair
54 consisting of the two numbers 1 and 2, and a pair containing the symbols
55 @code{foo} and @code{bar} can be entered. It is very important to write
56 the whitespace before and after the dot, because otherwise the Scheme
57 parser would not be able to figure out where to split the tokens.
64 But beware, if you want to try out these examples, you have to
65 @dfn{quote} the expressions. More information about quotation is
66 available in the section @ref{Expression Syntax}. The correct way
67 to try these examples is as follows.
78 A new pair is made by calling the procedure @code{cons} with two
79 arguments. Then the argument values are stored into a newly allocated
80 pair, and the pair is returned. The name @code{cons} stands for
81 "construct". Use the procedure @code{pair?} to test whether a
82 given Scheme object is a pair or not.
85 @deffn {Scheme Procedure} cons x y
86 @deffnx {C Function} scm_cons (x, y)
87 Return a newly allocated pair whose car is @var{x} and whose
88 cdr is @var{y}. The pair is guaranteed to be different (in the
89 sense of @code{eq?}) from every previously existing object.
93 @deffn {Scheme Procedure} pair? x
94 @deffnx {C Function} scm_pair_p (x)
95 Return @code{#t} if @var{x} is a pair; otherwise return
99 @deftypefn {C Function} int scm_is_pair (SCM x)
100 Return 1 when @var{x} is a pair; otherwise return 0.
103 The two parts of a pair are traditionally called @dfn{car} and
104 @dfn{cdr}. They can be retrieved with procedures of the same name
105 (@code{car} and @code{cdr}), and can be modified with the procedures
106 @code{set-car!} and @code{set-cdr!}. Since a very common operation in
107 Scheme programs is to access the car of a car of a pair, or the car of
108 the cdr of a pair, etc., the procedures called @code{caar},
109 @code{cadr} and so on are also predefined.
113 @deffn {Scheme Procedure} car pair
114 @deffnx {Scheme Procedure} cdr pair
115 @deffnx {C Function} scm_car (pair)
116 @deffnx {C Function} scm_cdr (pair)
117 Return the car or the cdr of @var{pair}, respectively.
120 @deftypefn {C Macro} SCM SCM_CAR (SCM pair)
121 @deftypefnx {C Macro} SCM SCM_CDR (SCM pair)
122 These two macros are the fastest way to access the car or cdr of a
123 pair; they can be thought of as compiling into a single memory
126 These macros do no checking at all. The argument @var{pair} must be a
130 @deffn {Scheme Procedure} cddr pair
131 @deffnx {Scheme Procedure} cdar pair
132 @deffnx {Scheme Procedure} cadr pair
133 @deffnx {Scheme Procedure} caar pair
134 @deffnx {Scheme Procedure} cdddr pair
135 @deffnx {Scheme Procedure} cddar pair
136 @deffnx {Scheme Procedure} cdadr pair
137 @deffnx {Scheme Procedure} cdaar pair
138 @deffnx {Scheme Procedure} caddr pair
139 @deffnx {Scheme Procedure} cadar pair
140 @deffnx {Scheme Procedure} caadr pair
141 @deffnx {Scheme Procedure} caaar pair
142 @deffnx {Scheme Procedure} cddddr pair
143 @deffnx {Scheme Procedure} cdddar pair
144 @deffnx {Scheme Procedure} cddadr pair
145 @deffnx {Scheme Procedure} cddaar pair
146 @deffnx {Scheme Procedure} cdaddr pair
147 @deffnx {Scheme Procedure} cdadar pair
148 @deffnx {Scheme Procedure} cdaadr pair
149 @deffnx {Scheme Procedure} cdaaar pair
150 @deffnx {Scheme Procedure} cadddr pair
151 @deffnx {Scheme Procedure} caddar pair
152 @deffnx {Scheme Procedure} cadadr pair
153 @deffnx {Scheme Procedure} cadaar pair
154 @deffnx {Scheme Procedure} caaddr pair
155 @deffnx {Scheme Procedure} caadar pair
156 @deffnx {Scheme Procedure} caaadr pair
157 @deffnx {Scheme Procedure} caaaar pair
158 @deffnx {C Function} scm_cddr (pair)
159 @deffnx {C Function} scm_cdar (pair)
160 @deffnx {C Function} scm_cadr (pair)
161 @deffnx {C Function} scm_caar (pair)
162 @deffnx {C Function} scm_cdddr (pair)
163 @deffnx {C Function} scm_cddar (pair)
164 @deffnx {C Function} scm_cdadr (pair)
165 @deffnx {C Function} scm_cdaar (pair)
166 @deffnx {C Function} scm_caddr (pair)
167 @deffnx {C Function} scm_cadar (pair)
168 @deffnx {C Function} scm_caadr (pair)
169 @deffnx {C Function} scm_caaar (pair)
170 @deffnx {C Function} scm_cddddr (pair)
171 @deffnx {C Function} scm_cdddar (pair)
172 @deffnx {C Function} scm_cddadr (pair)
173 @deffnx {C Function} scm_cddaar (pair)
174 @deffnx {C Function} scm_cdaddr (pair)
175 @deffnx {C Function} scm_cdadar (pair)
176 @deffnx {C Function} scm_cdaadr (pair)
177 @deffnx {C Function} scm_cdaaar (pair)
178 @deffnx {C Function} scm_cadddr (pair)
179 @deffnx {C Function} scm_caddar (pair)
180 @deffnx {C Function} scm_cadadr (pair)
181 @deffnx {C Function} scm_cadaar (pair)
182 @deffnx {C Function} scm_caaddr (pair)
183 @deffnx {C Function} scm_caadar (pair)
184 @deffnx {C Function} scm_caaadr (pair)
185 @deffnx {C Function} scm_caaaar (pair)
186 These procedures are compositions of @code{car} and @code{cdr}, where
187 for example @code{caddr} could be defined by
190 (define caddr (lambda (x) (car (cdr (cdr x)))))
193 @code{cadr}, @code{caddr} and @code{cadddr} pick out the second, third
194 or fourth elements of a list, respectively. SRFI-1 provides the same
195 under the names @code{second}, @code{third} and @code{fourth}
196 (@pxref{SRFI-1 Selectors}).
200 @deffn {Scheme Procedure} set-car! pair value
201 @deffnx {C Function} scm_set_car_x (pair, value)
202 Stores @var{value} in the car field of @var{pair}. The value returned
203 by @code{set-car!} is unspecified.
207 @deffn {Scheme Procedure} set-cdr! pair value
208 @deffnx {C Function} scm_set_cdr_x (pair, value)
209 Stores @var{value} in the cdr field of @var{pair}. The value returned
210 by @code{set-cdr!} is unspecified.
218 A very important data type in Scheme---as well as in all other Lisp
219 dialects---is the data type @dfn{list}.@footnote{Strictly speaking,
220 Scheme does not have a real datatype @dfn{list}. Lists are made up of
221 @dfn{chained pairs}, and only exist by definition---a list is a chain
222 of pairs which looks like a list.}
224 This is the short definition of what a list is:
228 Either the empty list @code{()},
231 or a pair which has a list in its cdr.
234 @c FIXME::martin: Describe the pair chaining in more detail.
236 @c FIXME::martin: What is a proper, what an improper list?
237 @c What is a circular list?
239 @c FIXME::martin: Maybe steal some graphics from the Elisp reference
243 * List Syntax:: Writing literal lists.
244 * List Predicates:: Testing lists.
245 * List Constructors:: Creating new lists.
246 * List Selection:: Selecting from lists, getting their length.
247 * Append/Reverse:: Appending and reversing lists.
248 * List Modification:: Modifying existing lists.
249 * List Searching:: Searching for list elements
250 * List Mapping:: Applying procedures to lists.
254 @subsubsection List Read Syntax
256 The syntax for lists is an opening parentheses, then all the elements of
257 the list (separated by whitespace) and finally a closing
258 parentheses.@footnote{Note that there is no separation character between
259 the list elements, like a comma or a semicolon.}.
262 (1 2 3) ; @r{a list of the numbers 1, 2 and 3}
263 ("foo" bar 3.1415) ; @r{a string, a symbol and a real number}
264 () ; @r{the empty list}
267 The last example needs a bit more explanation. A list with no elements,
268 called the @dfn{empty list}, is special in some ways. It is used for
269 terminating lists by storing it into the cdr of the last pair that makes
270 up a list. An example will clear that up:
281 This example also shows that lists have to be quoted when written
282 (@pxref{Expression Syntax}), because they would otherwise be
283 mistakingly taken as procedure applications (@pxref{Simple
287 @node List Predicates
288 @subsubsection List Predicates
290 Often it is useful to test whether a given Scheme object is a list or
291 not. List-processing procedures could use this information to test
292 whether their input is valid, or they could do different things
293 depending on the datatype of their arguments.
296 @deffn {Scheme Procedure} list? x
297 @deffnx {C Function} scm_list_p (x)
298 Return @code{#t} iff @var{x} is a proper list, else @code{#f}.
301 The predicate @code{null?} is often used in list-processing code to
302 tell whether a given list has run out of elements. That is, a loop
303 somehow deals with the elements of a list until the list satisfies
304 @code{null?}. Then, the algorithm terminates.
307 @deffn {Scheme Procedure} null? x
308 @deffnx {C Function} scm_null_p (x)
309 Return @code{#t} iff @var{x} is the empty list, else @code{#f}.
312 @deftypefn {C Function} int scm_is_null (SCM x)
313 Return 1 when @var{x} is the empty list; otherwise return 0.
317 @node List Constructors
318 @subsubsection List Constructors
320 This section describes the procedures for constructing new lists.
321 @code{list} simply returns a list where the elements are the arguments,
322 @code{cons*} is similar, but the last argument is stored in the cdr of
323 the last pair of the list.
325 @c C Function scm_list(rest) used to be documented here, but it's a
326 @c no-op since it does nothing but return the list the caller must
327 @c have already created.
329 @deffn {Scheme Procedure} list elem1 @dots{} elemN
330 @deffnx {C Function} scm_list_1 (elem1)
331 @deffnx {C Function} scm_list_2 (elem1, elem2)
332 @deffnx {C Function} scm_list_3 (elem1, elem2, elem3)
333 @deffnx {C Function} scm_list_4 (elem1, elem2, elem3, elem4)
334 @deffnx {C Function} scm_list_5 (elem1, elem2, elem3, elem4, elem5)
335 @deffnx {C Function} scm_list_n (elem1, @dots{}, elemN, @nicode{SCM_UNDEFINED})
337 Return a new list containing elements @var{elem1} to @var{elemN}.
339 @code{scm_list_n} takes a variable number of arguments, terminated by
340 the special @code{SCM_UNDEFINED}. That final @code{SCM_UNDEFINED} is
341 not included in the list. None of @var{elem1} to @var{elemN} can
342 themselves be @code{SCM_UNDEFINED}, or @code{scm_list_n} will
343 terminate at that point.
346 @c C Function scm_cons_star(arg1,rest) used to be documented here,
347 @c but it's not really a useful interface, since it expects the
348 @c caller to have already consed up all but the first argument
351 @deffn {Scheme Procedure} cons* arg1 arg2 @dots{}
352 Like @code{list}, but the last arg provides the tail of the
353 constructed list, returning @code{(cons @var{arg1} (cons
354 @var{arg2} (cons @dots{} @var{argn})))}. Requires at least one
355 argument. If given one argument, that argument is returned as
356 result. This function is called @code{list*} in some other
357 Schemes and in Common LISP.
360 @deffn {Scheme Procedure} list-copy lst
361 @deffnx {C Function} scm_list_copy (lst)
362 Return a (newly-created) copy of @var{lst}.
365 @deffn {Scheme Procedure} make-list n [init]
366 Create a list containing of @var{n} elements, where each element is
367 initialized to @var{init}. @var{init} defaults to the empty list
368 @code{()} if not given.
371 Note that @code{list-copy} only makes a copy of the pairs which make up
372 the spine of the lists. The list elements are not copied, which means
373 that modifying the elements of the new list also modifies the elements
374 of the old list. On the other hand, applying procedures like
375 @code{set-cdr!} or @code{delv!} to the new list will not alter the old
376 list. If you also need to copy the list elements (making a deep copy),
377 use the procedure @code{copy-tree} (@pxref{Copying}).
380 @subsubsection List Selection
382 These procedures are used to get some information about a list, or to
383 retrieve one or more elements of a list.
386 @deffn {Scheme Procedure} length lst
387 @deffnx {C Function} scm_length (lst)
388 Return the number of elements in list @var{lst}.
391 @deffn {Scheme Procedure} last-pair lst
392 @deffnx {C Function} scm_last_pair (lst)
393 Return the last pair in @var{lst}, signalling an error if
394 @var{lst} is circular.
398 @deffn {Scheme Procedure} list-ref list k
399 @deffnx {C Function} scm_list_ref (list, k)
400 Return the @var{k}th element from @var{list}.
404 @deffn {Scheme Procedure} list-tail lst k
405 @deffnx {Scheme Procedure} list-cdr-ref lst k
406 @deffnx {C Function} scm_list_tail (lst, k)
407 Return the "tail" of @var{lst} beginning with its @var{k}th element.
408 The first element of the list is considered to be element 0.
410 @code{list-tail} and @code{list-cdr-ref} are identical. It may help to
411 think of @code{list-cdr-ref} as accessing the @var{k}th cdr of the list,
412 or returning the results of cdring @var{k} times down @var{lst}.
415 @deffn {Scheme Procedure} list-head lst k
416 @deffnx {C Function} scm_list_head (lst, k)
417 Copy the first @var{k} elements from @var{lst} into a new list, and
422 @subsubsection Append and Reverse
424 @code{append} and @code{append!} are used to concatenate two or more
425 lists in order to form a new list. @code{reverse} and @code{reverse!}
426 return lists with the same elements as their arguments, but in reverse
427 order. The procedure variants with an @code{!} directly modify the
428 pairs which form the list, whereas the other procedures create new
429 pairs. This is why you should be careful when using the side-effecting
433 @deffn {Scheme Procedure} append lst1 @dots{} lstN
434 @deffnx {Scheme Procedure} append! lst1 @dots{} lstN
435 @deffnx {C Function} scm_append (lstlst)
436 @deffnx {C Function} scm_append_x (lstlst)
437 Return a list comprising all the elements of lists @var{lst1} to
441 (append '(x) '(y)) @result{} (x y)
442 (append '(a) '(b c d)) @result{} (a b c d)
443 (append '(a (b)) '((c))) @result{} (a (b) (c))
446 The last argument @var{lstN} may actually be any object; an improper
447 list results if the last argument is not a proper list.
450 (append '(a b) '(c . d)) @result{} (a b c . d)
451 (append '() 'a) @result{} a
454 @code{append} doesn't modify the given lists, but the return may share
455 structure with the final @var{lstN}. @code{append!} modifies the
456 given lists to form its return.
458 For @code{scm_append} and @code{scm_append_x}, @var{lstlst} is a list
459 of the list operands @var{lst1} @dots{} @var{lstN}. That @var{lstlst}
460 itself is not modified or used in the return.
464 @deffn {Scheme Procedure} reverse lst
465 @deffnx {Scheme Procedure} reverse! lst [newtail]
466 @deffnx {C Function} scm_reverse (lst)
467 @deffnx {C Function} scm_reverse_x (lst, newtail)
468 Return a list comprising the elements of @var{lst}, in reverse order.
470 @code{reverse} constructs a new list, @code{reverse!} modifies
471 @var{lst} in constructing its return.
473 For @code{reverse!}, the optional @var{newtail} is appended to the
474 result. @var{newtail} isn't reversed, it simply becomes the list
475 tail. For @code{scm_reverse_x}, the @var{newtail} parameter is
476 mandatory, but can be @code{SCM_EOL} if no further tail is required.
479 @node List Modification
480 @subsubsection List Modification
482 The following procedures modify an existing list, either by changing
483 elements of the list, or by changing the list structure itself.
485 @deffn {Scheme Procedure} list-set! list k val
486 @deffnx {C Function} scm_list_set_x (list, k, val)
487 Set the @var{k}th element of @var{list} to @var{val}.
490 @deffn {Scheme Procedure} list-cdr-set! list k val
491 @deffnx {C Function} scm_list_cdr_set_x (list, k, val)
492 Set the @var{k}th cdr of @var{list} to @var{val}.
495 @deffn {Scheme Procedure} delq item lst
496 @deffnx {C Function} scm_delq (item, lst)
497 Return a newly-created copy of @var{lst} with elements
498 @code{eq?} to @var{item} removed. This procedure mirrors
499 @code{memq}: @code{delq} compares elements of @var{lst} against
500 @var{item} with @code{eq?}.
503 @deffn {Scheme Procedure} delv item lst
504 @deffnx {C Function} scm_delv (item, lst)
505 Return a newly-created copy of @var{lst} with elements
506 @code{eqv?} to @var{item} removed. This procedure mirrors
507 @code{memv}: @code{delv} compares elements of @var{lst} against
508 @var{item} with @code{eqv?}.
511 @deffn {Scheme Procedure} delete item lst
512 @deffnx {C Function} scm_delete (item, lst)
513 Return a newly-created copy of @var{lst} with elements
514 @code{equal?} to @var{item} removed. This procedure mirrors
515 @code{member}: @code{delete} compares elements of @var{lst}
516 against @var{item} with @code{equal?}.
518 See also SRFI-1 which has an extended @code{delete} (@ref{SRFI-1
519 Deleting}), and also an @code{lset-difference} which can delete
520 multiple @var{item}s in one call (@ref{SRFI-1 Set Operations}).
523 @deffn {Scheme Procedure} delq! item lst
524 @deffnx {Scheme Procedure} delv! item lst
525 @deffnx {Scheme Procedure} delete! item lst
526 @deffnx {C Function} scm_delq_x (item, lst)
527 @deffnx {C Function} scm_delv_x (item, lst)
528 @deffnx {C Function} scm_delete_x (item, lst)
529 These procedures are destructive versions of @code{delq}, @code{delv}
530 and @code{delete}: they modify the pointers in the existing @var{lst}
531 rather than creating a new list. Caveat evaluator: Like other
532 destructive list functions, these functions cannot modify the binding of
533 @var{lst}, and so cannot be used to delete the first element of
534 @var{lst} destructively.
537 @deffn {Scheme Procedure} delq1! item lst
538 @deffnx {C Function} scm_delq1_x (item, lst)
539 Like @code{delq!}, but only deletes the first occurrence of
540 @var{item} from @var{lst}. Tests for equality using
541 @code{eq?}. See also @code{delv1!} and @code{delete1!}.
544 @deffn {Scheme Procedure} delv1! item lst
545 @deffnx {C Function} scm_delv1_x (item, lst)
546 Like @code{delv!}, but only deletes the first occurrence of
547 @var{item} from @var{lst}. Tests for equality using
548 @code{eqv?}. See also @code{delq1!} and @code{delete1!}.
551 @deffn {Scheme Procedure} delete1! item lst
552 @deffnx {C Function} scm_delete1_x (item, lst)
553 Like @code{delete!}, but only deletes the first occurrence of
554 @var{item} from @var{lst}. Tests for equality using
555 @code{equal?}. See also @code{delq1!} and @code{delv1!}.
558 @deffn {Scheme Procedure} filter pred lst
559 @deffnx {Scheme Procedure} filter! pred lst
560 Return a list containing all elements from @var{lst} which satisfy the
561 predicate @var{pred}. The elements in the result list have the same
562 order as in @var{lst}. The order in which @var{pred} is applied to
563 the list elements is not specified.
565 @code{filter} does not change @var{lst}, but the result may share a
566 tail with it. @code{filter!} may modify @var{lst} to construct its
571 @subsubsection List Searching
573 The following procedures search lists for particular elements. They use
574 different comparison predicates for comparing list elements with the
575 object to be searched. When they fail, they return @code{#f}, otherwise
576 they return the sublist whose car is equal to the search object, where
577 equality depends on the equality predicate used.
580 @deffn {Scheme Procedure} memq x lst
581 @deffnx {C Function} scm_memq (x, lst)
582 Return the first sublist of @var{lst} whose car is @code{eq?}
583 to @var{x} where the sublists of @var{lst} are the non-empty
584 lists returned by @code{(list-tail @var{lst} @var{k})} for
585 @var{k} less than the length of @var{lst}. If @var{x} does not
586 occur in @var{lst}, then @code{#f} (not the empty list) is
591 @deffn {Scheme Procedure} memv x lst
592 @deffnx {C Function} scm_memv (x, lst)
593 Return the first sublist of @var{lst} whose car is @code{eqv?}
594 to @var{x} where the sublists of @var{lst} are the non-empty
595 lists returned by @code{(list-tail @var{lst} @var{k})} for
596 @var{k} less than the length of @var{lst}. If @var{x} does not
597 occur in @var{lst}, then @code{#f} (not the empty list) is
602 @deffn {Scheme Procedure} member x lst
603 @deffnx {C Function} scm_member (x, lst)
604 Return the first sublist of @var{lst} whose car is
605 @code{equal?} to @var{x} where the sublists of @var{lst} are
606 the non-empty lists returned by @code{(list-tail @var{lst}
607 @var{k})} for @var{k} less than the length of @var{lst}. If
608 @var{x} does not occur in @var{lst}, then @code{#f} (not the
609 empty list) is returned.
611 See also SRFI-1 which has an extended @code{member} function
612 (@ref{SRFI-1 Searching}).
617 @subsubsection List Mapping
619 List processing is very convenient in Scheme because the process of
620 iterating over the elements of a list can be highly abstracted. The
621 procedures in this section are the most basic iterating procedures for
622 lists. They take a procedure and one or more lists as arguments, and
623 apply the procedure to each element of the list. They differ in their
627 @c begin (texi-doc-string "guile" "map")
628 @deffn {Scheme Procedure} map proc arg1 arg2 @dots{}
629 @deffnx {Scheme Procedure} map-in-order proc arg1 arg2 @dots{}
630 @deffnx {C Function} scm_map (proc, arg1, args)
631 Apply @var{proc} to each element of the list @var{arg1} (if only two
632 arguments are given), or to the corresponding elements of the argument
633 lists (if more than two arguments are given). The result(s) of the
634 procedure applications are saved and returned in a list. For
635 @code{map}, the order of procedure applications is not specified,
636 @code{map-in-order} applies the procedure from left to right to the list
641 @c begin (texi-doc-string "guile" "for-each")
642 @deffn {Scheme Procedure} for-each proc arg1 arg2 @dots{}
643 Like @code{map}, but the procedure is always applied from left to right,
644 and the result(s) of the procedure applications are thrown away. The
645 return value is not specified.
648 See also SRFI-1 which extends these functions to take lists of unequal
649 lengths (@ref{SRFI-1 Fold and Map}).
655 Vectors are sequences of Scheme objects. Unlike lists, the length of a
656 vector, once the vector is created, cannot be changed. The advantage of
657 vectors over lists is that the time required to access one element of a vector
658 given its @dfn{position} (synonymous with @dfn{index}), a zero-origin number,
659 is constant, whereas lists have an access time linear to the position of the
660 accessed element in the list.
662 Vectors can contain any kind of Scheme object; it is even possible to
663 have different types of objects in the same vector. For vectors
664 containing vectors, you may wish to use arrays, instead. Note, too,
665 that vectors are the special case of one dimensional non-uniform arrays
666 and that most array procedures operate happily on vectors
670 * Vector Syntax:: Read syntax for vectors.
671 * Vector Creation:: Dynamic vector creation and validation.
672 * Vector Accessors:: Accessing and modifying vector contents.
673 * Vector Accessing from C:: Ways to work with vectors from C.
674 * Uniform Numeric Vectors:: Vectors of unboxed numeric values.
679 @subsubsection Read Syntax for Vectors
681 Vectors can literally be entered in source code, just like strings,
682 characters or some of the other data types. The read syntax for vectors
683 is as follows: A sharp sign (@code{#}), followed by an opening
684 parentheses, all elements of the vector in their respective read syntax,
685 and finally a closing parentheses. The following are examples of the
686 read syntax for vectors; where the first vector only contains numbers
687 and the second three different object types: a string, a symbol and a
688 number in hexadecimal notation.
692 #("Hello" foo #xdeadbeef)
695 Like lists, vectors have to be quoted:
698 '#(a b c) @result{} #(a b c)
701 @node Vector Creation
702 @subsubsection Dynamic Vector Creation and Validation
704 Instead of creating a vector implicitly by using the read syntax just
705 described, you can create a vector dynamically by calling one of the
706 @code{vector} and @code{list->vector} primitives with the list of Scheme
707 values that you want to place into a vector. The size of the vector
708 thus created is determined implicitly by the number of arguments given.
711 @rnindex list->vector
712 @deffn {Scheme Procedure} vector . l
713 @deffnx {Scheme Procedure} list->vector l
714 @deffnx {C Function} scm_vector (l)
715 Return a newly allocated vector composed of the
716 given arguments. Analogous to @code{list}.
719 (vector 'a 'b 'c) @result{} #(a b c)
723 The inverse operation is @code{vector->list}:
725 @rnindex vector->list
726 @deffn {Scheme Procedure} vector->list v
727 @deffnx {C Function} scm_vector_to_list (v)
728 Return a newly allocated list composed of the elements of @var{v}.
731 (vector->list '#(dah dah didah)) @result{} (dah dah didah)
732 (list->vector '(dididit dah)) @result{} #(dididit dah)
736 To allocate a vector with an explicitly specified size, use
737 @code{make-vector}. With this primitive you can also specify an initial
738 value for the vector elements (the same value for all elements, that
742 @deffn {Scheme Procedure} make-vector len [fill]
743 @deffnx {C Function} scm_make_vector (len, fill)
744 Return a newly allocated vector of @var{len} elements. If a
745 second argument is given, then each position is initialized to
746 @var{fill}. Otherwise the initial contents of each position is
750 @deftypefn {C Function} SCM scm_c_make_vector (size_t k, SCM fill)
751 Like @code{scm_make_vector}, but the length is given as a @code{size_t}.
754 To check whether an arbitrary Scheme value @emph{is} a vector, use the
755 @code{vector?} primitive:
758 @deffn {Scheme Procedure} vector? obj
759 @deffnx {C Function} scm_vector_p (obj)
760 Return @code{#t} if @var{obj} is a vector, otherwise return
764 @deftypefn {C Function} int scm_is_vector (SCM obj)
765 Return non-zero when @var{obj} is a vector, otherwise return
769 @node Vector Accessors
770 @subsubsection Accessing and Modifying Vector Contents
772 @code{vector-length} and @code{vector-ref} return information about a
773 given vector, respectively its size and the elements that are contained
776 @rnindex vector-length
777 @deffn {Scheme Procedure} vector-length vector
778 @deffnx {C Function} scm_vector_length vector
779 Return the number of elements in @var{vector} as an exact integer.
782 @deftypefn {C Function} size_t scm_c_vector_length (SCM v)
783 Return the number of elements in @var{vector} as a @code{size_t}.
787 @deffn {Scheme Procedure} vector-ref vector k
788 @deffnx {C Function} scm_vector_ref vector k
789 Return the contents of position @var{k} of @var{vector}.
790 @var{k} must be a valid index of @var{vector}.
792 (vector-ref '#(1 1 2 3 5 8 13 21) 5) @result{} 8
793 (vector-ref '#(1 1 2 3 5 8 13 21)
794 (let ((i (round (* 2 (acos -1)))))
801 @deftypefn {C Function} SCM scm_c_vector_ref (SCM v, size_t k)
802 Return the contents of position @var{k} (a @code{size_t}) of
806 A vector created by one of the dynamic vector constructor procedures
807 (@pxref{Vector Creation}) can be modified using the following
810 @emph{NOTE:} According to R5RS, it is an error to use any of these
811 procedures on a literally read vector, because such vectors should be
812 considered as constants. Currently, however, Guile does not detect this
816 @deffn {Scheme Procedure} vector-set! vector k obj
817 @deffnx {C Function} scm_vector_set_x vector k obj
818 Store @var{obj} in position @var{k} of @var{vector}.
819 @var{k} must be a valid index of @var{vector}.
820 The value returned by @samp{vector-set!} is unspecified.
822 (let ((vec (vector 0 '(2 2 2 2) "Anna")))
823 (vector-set! vec 1 '("Sue" "Sue"))
824 vec) @result{} #(0 ("Sue" "Sue") "Anna")
828 @deftypefn {C Function} void scm_c_vector_set_x (SCM v, size_t k, SCM obj)
829 Store @var{obj} in position @var{k} (a @code{size_t}) of @var{v}.
832 @rnindex vector-fill!
833 @deffn {Scheme Procedure} vector-fill! v fill
834 @deffnx {C Function} scm_vector_fill_x (v, fill)
835 Store @var{fill} in every position of @var{vector}. The value
836 returned by @code{vector-fill!} is unspecified.
839 @deffn {Scheme Procedure} vector-copy vec
840 @deffnx {C Function} scm_vector_copy (vec)
841 Return a copy of @var{vec}.
844 @deffn {Scheme Procedure} vector-move-left! vec1 start1 end1 vec2 start2
845 @deffnx {C Function} scm_vector_move_left_x (vec1, start1, end1, vec2, start2)
846 Copy elements from @var{vec1}, positions @var{start1} to @var{end1},
847 to @var{vec2} starting at position @var{start2}. @var{start1} and
848 @var{start2} are inclusive indices; @var{end1} is exclusive.
850 @code{vector-move-left!} copies elements in leftmost order.
851 Therefore, in the case where @var{vec1} and @var{vec2} refer to the
852 same vector, @code{vector-move-left!} is usually appropriate when
853 @var{start1} is greater than @var{start2}.
856 @deffn {Scheme Procedure} vector-move-right! vec1 start1 end1 vec2 start2
857 @deffnx {C Function} scm_vector_move_right_x (vec1, start1, end1, vec2, start2)
858 Copy elements from @var{vec1}, positions @var{start1} to @var{end1},
859 to @var{vec2} starting at position @var{start2}. @var{start1} and
860 @var{start2} are inclusive indices; @var{end1} is exclusive.
862 @code{vector-move-right!} copies elements in rightmost order.
863 Therefore, in the case where @var{vec1} and @var{vec2} refer to the
864 same vector, @code{vector-move-right!} is usually appropriate when
865 @var{start1} is less than @var{start2}.
868 @node Vector Accessing from C
869 @subsubsection Vector Accessing from C
871 A vector can be read and modified from C with the functions
872 @code{scm_c_vector_ref} and @code{scm_c_vector_set_x}, for example. In
873 addition to these functions, there are two more ways to access vectors
874 from C that might be more efficient in certain situations: you can
875 restrict yourself to @dfn{simple vectors} and then use the very fast
876 @emph{simple vector macros}; or you can use the very general framework
877 for accessing all kinds of arrays (@pxref{Accessing Arrays from C}),
878 which is more verbose, but can deal efficiently with all kinds of
879 vectors (and arrays). For vectors, you can use the
880 @code{scm_vector_elements} and @code{scm_vector_writable_elements}
881 functions as shortcuts.
883 @deftypefn {C Function} int scm_is_simple_vector (SCM obj)
884 Return non-zero if @var{obj} is a simple vector, else return zero. A
885 simple vector is a vector that can be used with the @code{SCM_SIMPLE_*}
888 The following functions are guaranteed to return simple vectors:
889 @code{scm_make_vector}, @code{scm_c_make_vector}, @code{scm_vector},
890 @code{scm_list_to_vector}.
893 @deftypefn {C Macro} size_t SCM_SIMPLE_VECTOR_LENGTH (SCM vec)
894 Evaluates to the length of the simple vector @var{vec}. No type
898 @deftypefn {C Macro} SCM SCM_SIMPLE_VECTOR_REF (SCM vec, size_t idx)
899 Evaluates to the element at position @var{idx} in the simple vector
900 @var{vec}. No type or range checking is done.
903 @deftypefn {C Macro} void SCM_SIMPLE_VECTOR_SET (SCM vec, size_t idx, SCM val)
904 Sets the element at position @var{idx} in the simple vector
905 @var{vec} to @var{val}. No type or range checking is done.
908 @deftypefn {C Function} {const SCM *} scm_vector_elements (SCM vec, scm_t_array_handle *handle, size_t *lenp, ssize_t *incp)
909 Acquire a handle for the vector @var{vec} and return a pointer to the
910 elements of it. This pointer can only be used to read the elements of
911 @var{vec}. When @var{vec} is not a vector, an error is signaled. The
912 handle mustr eventually be released with
913 @code{scm_array_handle_release}.
915 The variables pointed to by @var{lenp} and @var{incp} are filled with
916 the number of elements of the vector and the increment (number of
917 elements) between successive elements, respectively. Successive
918 elements of @var{vec} need not be contiguous in their underlying
919 ``root vector'' returned here; hence the increment is not necessarily
920 equal to 1 and may well be negative too (@pxref{Shared Arrays}).
922 The following example shows the typical way to use this function. It
923 creates a list of all elements of @var{vec} (in reverse order).
926 scm_t_array_handle handle;
932 elt = scm_vector_elements (vec, &handle, &len, &inc);
934 for (i = 0; i < len; i++, elt += inc)
935 list = scm_cons (*elt, list);
936 scm_array_handle_release (&handle);
941 @deftypefn {C Function} {SCM *} scm_vector_writable_elements (SCM vec, scm_t_array_handle *handle, size_t *lenp, ssize_t *incp)
942 Like @code{scm_vector_elements} but the pointer can be used to modify
945 The following example shows the typical way to use this function. It
946 fills a vector with @code{#t}.
949 scm_t_array_handle handle;
954 elt = scm_vector_writable_elements (vec, &handle, &len, &inc);
955 for (i = 0; i < len; i++, elt += inc)
957 scm_array_handle_release (&handle);
962 @node Uniform Numeric Vectors
963 @subsubsection Uniform Numeric Vectors
965 A uniform numeric vector is a vector whose elements are all of a single
966 numeric type. Guile offers uniform numeric vectors for signed and
967 unsigned 8-bit, 16-bit, 32-bit, and 64-bit integers, two sizes of
968 floating point values, and complex floating-point numbers of these two
969 sizes. @xref{SRFI-4}, for more information.
971 For many purposes, bytevectors work just as well as uniform vectors, and have
972 the advantage that they integrate well with binary input and output.
973 @xref{Bytevectors}, for more information on bytevectors.
976 @subsection Bit Vectors
979 Bit vectors are zero-origin, one-dimensional arrays of booleans. They
980 are displayed as a sequence of @code{0}s and @code{1}s prefixed by
984 (make-bitvector 8 #f) @result{}
988 Bit vectors are also generalized vectors, @xref{Generalized
989 Vectors}, and can thus be used with the array procedures, @xref{Arrays}.
990 Bit vectors are the special case of one dimensional bit arrays.
992 @deffn {Scheme Procedure} bitvector? obj
993 @deffnx {C Function} scm_bitvector_p (obj)
994 Return @code{#t} when @var{obj} is a bitvector, else
998 @deftypefn {C Function} int scm_is_bitvector (SCM obj)
999 Return @code{1} when @var{obj} is a bitvector, else return @code{0}.
1002 @deffn {Scheme Procedure} make-bitvector len [fill]
1003 @deffnx {C Function} scm_make_bitvector (len, fill)
1004 Create a new bitvector of length @var{len} and
1005 optionally initialize all elements to @var{fill}.
1008 @deftypefn {C Function} SCM scm_c_make_bitvector (size_t len, SCM fill)
1009 Like @code{scm_make_bitvector}, but the length is given as a
1013 @deffn {Scheme Procedure} bitvector . bits
1014 @deffnx {C Function} scm_bitvector (bits)
1015 Create a new bitvector with the arguments as elements.
1018 @deffn {Scheme Procedure} bitvector-length vec
1019 @deffnx {C Function} scm_bitvector_length (vec)
1020 Return the length of the bitvector @var{vec}.
1023 @deftypefn {C Function} size_t scm_c_bitvector_length (SCM vec)
1024 Like @code{scm_bitvector_length}, but the length is returned as a
1028 @deffn {Scheme Procedure} bitvector-ref vec idx
1029 @deffnx {C Function} scm_bitvector_ref (vec, idx)
1030 Return the element at index @var{idx} of the bitvector
1034 @deftypefn {C Function} SCM scm_c_bitvector_ref (SCM obj, size_t idx)
1035 Return the element at index @var{idx} of the bitvector
1039 @deffn {Scheme Procedure} bitvector-set! vec idx val
1040 @deffnx {C Function} scm_bitvector_set_x (vec, idx, val)
1041 Set the element at index @var{idx} of the bitvector
1042 @var{vec} when @var{val} is true, else clear it.
1045 @deftypefn {C Function} SCM scm_c_bitvector_set_x (SCM obj, size_t idx, SCM val)
1046 Set the element at index @var{idx} of the bitvector
1047 @var{vec} when @var{val} is true, else clear it.
1050 @deffn {Scheme Procedure} bitvector-fill! vec val
1051 @deffnx {C Function} scm_bitvector_fill_x (vec, val)
1052 Set all elements of the bitvector
1053 @var{vec} when @var{val} is true, else clear them.
1056 @deffn {Scheme Procedure} list->bitvector list
1057 @deffnx {C Function} scm_list_to_bitvector (list)
1058 Return a new bitvector initialized with the elements
1062 @deffn {Scheme Procedure} bitvector->list vec
1063 @deffnx {C Function} scm_bitvector_to_list (vec)
1064 Return a new list initialized with the elements
1065 of the bitvector @var{vec}.
1068 @deffn {Scheme Procedure} bit-count bool bitvector
1069 @deffnx {C Function} scm_bit_count (bool, bitvector)
1070 Return a count of how many entries in @var{bitvector} are equal to
1071 @var{bool}. For example,
1074 (bit-count #f #*000111000) @result{} 6
1078 @deffn {Scheme Procedure} bit-position bool bitvector start
1079 @deffnx {C Function} scm_bit_position (bool, bitvector, start)
1080 Return the index of the first occurrence of @var{bool} in
1081 @var{bitvector}, starting from @var{start}. If there is no @var{bool}
1082 entry between @var{start} and the end of @var{bitvector}, then return
1083 @code{#f}. For example,
1086 (bit-position #t #*000101 0) @result{} 3
1087 (bit-position #f #*0001111 3) @result{} #f
1091 @deffn {Scheme Procedure} bit-invert! bitvector
1092 @deffnx {C Function} scm_bit_invert_x (bitvector)
1093 Modify @var{bitvector} by replacing each element with its negation.
1096 @deffn {Scheme Procedure} bit-set*! bitvector uvec bool
1097 @deffnx {C Function} scm_bit_set_star_x (bitvector, uvec, bool)
1098 Set entries of @var{bitvector} to @var{bool}, with @var{uvec}
1099 selecting the entries to change. The return value is unspecified.
1101 If @var{uvec} is a bit vector, then those entries where it has
1102 @code{#t} are the ones in @var{bitvector} which are set to @var{bool}.
1103 @var{uvec} and @var{bitvector} must be the same length. When
1104 @var{bool} is @code{#t} it's like @var{uvec} is OR'ed into
1105 @var{bitvector}. Or when @var{bool} is @code{#f} it can be seen as an
1109 (define bv #*01000010)
1110 (bit-set*! bv #*10010001 #t)
1112 @result{} #*11010011
1115 If @var{uvec} is a uniform vector of unsigned long integers, then
1116 they're indexes into @var{bitvector} which are set to @var{bool}.
1119 (define bv #*01000010)
1120 (bit-set*! bv #u(5 2 7) #t)
1122 @result{} #*01100111
1126 @deffn {Scheme Procedure} bit-count* bitvector uvec bool
1127 @deffnx {C Function} scm_bit_count_star (bitvector, uvec, bool)
1128 Return a count of how many entries in @var{bitvector} are equal to
1129 @var{bool}, with @var{uvec} selecting the entries to consider.
1131 @var{uvec} is interpreted in the same way as for @code{bit-set*!}
1132 above. Namely, if @var{uvec} is a bit vector then entries which have
1133 @code{#t} there are considered in @var{bitvector}. Or if @var{uvec}
1134 is a uniform vector of unsigned long integers then it's the indexes in
1135 @var{bitvector} to consider.
1140 (bit-count* #*01110111 #*11001101 #t) @result{} 3
1141 (bit-count* #*01110111 #u(7 0 4) #f) @result{} 2
1145 @deftypefn {C Function} {const scm_t_uint32 *} scm_bitvector_elements (SCM vec, scm_t_array_handle *handle, size_t *offp, size_t *lenp, ssize_t *incp)
1146 Like @code{scm_vector_elements} (@pxref{Vector Accessing from C}), but
1147 for bitvectors. The variable pointed to by @var{offp} is set to the
1148 value returned by @code{scm_array_handle_bit_elements_offset}. See
1149 @code{scm_array_handle_bit_elements} for how to use the returned
1150 pointer and the offset.
1153 @deftypefn {C Function} {scm_t_uint32 *} scm_bitvector_writable_elements (SCM vec, scm_t_array_handle *handle, size_t *offp, size_t *lenp, ssize_t *incp)
1154 Like @code{scm_bitvector_elements}, but the pointer is good for reading
1158 @node Generalized Vectors
1159 @subsection Generalized Vectors
1161 Guile has a number of data types that are generally vector-like:
1162 strings, uniform numeric vectors, bytevectors, bitvectors, and of course
1163 ordinary vectors of arbitrary Scheme values. These types are disjoint:
1164 a Scheme value belongs to at most one of the five types listed above.
1166 If you want to gloss over this distinction and want to treat all four
1167 types with common code, you can use the procedures in this section.
1168 They work with the @emph{generalized vector} type, which is the union
1169 of the five vector-like types.
1171 @deffn {Scheme Procedure} generalized-vector? obj
1172 @deffnx {C Function} scm_generalized_vector_p (obj)
1173 Return @code{#t} if @var{obj} is a vector, bytevector, string,
1174 bitvector, or uniform numeric vector.
1177 @deffn {Scheme Procedure} generalized-vector-length v
1178 @deffnx {C Function} scm_generalized_vector_length (v)
1179 Return the length of the generalized vector @var{v}.
1182 @deffn {Scheme Procedure} generalized-vector-ref v idx
1183 @deffnx {C Function} scm_generalized_vector_ref (v, idx)
1184 Return the element at index @var{idx} of the
1185 generalized vector @var{v}.
1188 @deffn {Scheme Procedure} generalized-vector-set! v idx val
1189 @deffnx {C Function} scm_generalized_vector_set_x (v, idx, val)
1190 Set the element at index @var{idx} of the
1191 generalized vector @var{v} to @var{val}.
1194 @deffn {Scheme Procedure} generalized-vector->list v
1195 @deffnx {C Function} scm_generalized_vector_to_list (v)
1196 Return a new list whose elements are the elements of the
1197 generalized vector @var{v}.
1200 @deftypefn {C Function} int scm_is_generalized_vector (SCM obj)
1201 Return @code{1} if @var{obj} is a vector, string,
1202 bitvector, or uniform numeric vector; else return @code{0}.
1205 @deftypefn {C Function} size_t scm_c_generalized_vector_length (SCM v)
1206 Return the length of the generalized vector @var{v}.
1209 @deftypefn {C Function} SCM scm_c_generalized_vector_ref (SCM v, size_t idx)
1210 Return the element at index @var{idx} of the generalized vector @var{v}.
1213 @deftypefn {C Function} void scm_c_generalized_vector_set_x (SCM v, size_t idx, SCM val)
1214 Set the element at index @var{idx} of the generalized vector @var{v}
1218 @deftypefn {C Function} void scm_generalized_vector_get_handle (SCM v, scm_t_array_handle *handle)
1219 Like @code{scm_array_get_handle} but an error is signalled when @var{v}
1220 is not of rank one. You can use @code{scm_array_handle_ref} and
1221 @code{scm_array_handle_set} to read and write the elements of @var{v},
1222 or you can use functions like @code{scm_array_handle_<foo>_elements} to
1223 deal with specific types of vectors.
1230 @dfn{Arrays} are a collection of cells organized into an arbitrary
1231 number of dimensions. Each cell can be accessed in constant time by
1232 supplying an index for each dimension.
1234 In the current implementation, an array uses a generalized vector for
1235 the actual storage of its elements. Any kind of generalized vector
1236 will do, so you can have arrays of uniform numeric values, arrays of
1237 characters, arrays of bits, and of course, arrays of arbitrary Scheme
1238 values. For example, arrays with an underlying @code{c64vector} might
1239 be nice for digital signal processing, while arrays made from a
1240 @code{u8vector} might be used to hold gray-scale images.
1242 The number of dimensions of an array is called its @dfn{rank}. Thus,
1243 a matrix is an array of rank 2, while a vector has rank 1. When
1244 accessing an array element, you have to specify one exact integer for
1245 each dimension. These integers are called the @dfn{indices} of the
1246 element. An array specifies the allowed range of indices for each
1247 dimension via an inclusive lower and upper bound. These bounds can
1248 well be negative, but the upper bound must be greater than or equal to
1249 the lower bound minus one. When all lower bounds of an array are
1250 zero, it is called a @dfn{zero-origin} array.
1252 Arrays can be of rank 0, which could be interpreted as a scalar.
1253 Thus, a zero-rank array can store exactly one object and the list of
1254 indices of this element is the empty list.
1256 Arrays contain zero elements when one of their dimensions has a zero
1257 length. These empty arrays maintain information about their shape: a
1258 matrix with zero columns and 3 rows is different from a matrix with 3
1259 columns and zero rows, which again is different from a vector of
1262 Generalized vectors, such as strings, uniform numeric vectors,
1263 bytevectors, bit vectors and ordinary vectors, are the special case of
1264 one dimensional arrays.
1268 * Array Procedures::
1270 * Accessing Arrays from C::
1274 @subsubsection Array Syntax
1276 An array is displayed as @code{#} followed by its rank, followed by a
1277 tag that describes the underlying vector, optionally followed by
1278 information about its shape, and finally followed by the cells,
1279 organized into dimensions using parentheses.
1281 In more words, the array tag is of the form
1284 #<rank><vectag><@@lower><:len><@@lower><:len>...
1287 where @code{<rank>} is a positive integer in decimal giving the rank of
1288 the array. It is omitted when the rank is 1 and the array is non-shared
1289 and has zero-origin (see below). For shared arrays and for a non-zero
1290 origin, the rank is always printed even when it is 1 to distinguish
1291 them from ordinary vectors.
1293 The @code{<vectag>} part is the tag for a uniform numeric vector, like
1294 @code{u8}, @code{s16}, etc, @code{b} for bitvectors, or @code{a} for
1295 strings. It is empty for ordinary vectors.
1297 The @code{<@@lower>} part is a @samp{@@} character followed by a signed
1298 integer in decimal giving the lower bound of a dimension. There is one
1299 @code{<@@lower>} for each dimension. When all lower bounds are zero,
1300 all @code{<@@lower>} parts are omitted.
1302 The @code{<:len>} part is a @samp{:} character followed by an unsigned
1303 integer in decimal giving the length of a dimension. Like for the lower
1304 bounds, there is one @code{<:len>} for each dimension, and the
1305 @code{<:len>} part always follows the @code{<@@lower>} part for a
1306 dimension. Lengths are only then printed when they can't be deduced
1307 from the nested lists of elements of the array literal, which can happen
1308 when at least one length is zero.
1310 As a special case, an array of rank 0 is printed as
1311 @code{#0<vectag>(<scalar>)}, where @code{<scalar>} is the result of
1312 printing the single element of the array.
1318 is an ordinary array of rank 1 with lower bound 0 in dimension 0.
1319 (I.e., a regular vector.)
1322 is an ordinary array of rank 1 with lower bound 2 in dimension 0.
1324 @item #2((1 2 3) (4 5 6))
1325 is a non-uniform array of rank 2; a 3@cross{}3 matrix with index ranges 0..2
1329 is a uniform u8 array of rank 1.
1331 @item #2u32@@2@@3((1 2) (2 3))
1332 is a uniform u8 array of rank 2 with index ranges 2..3 and 3..4.
1335 is a two-dimensional array with index ranges 0..-1 and 0..-1, i.e. both
1336 dimensions have length zero.
1339 is a two-dimensional array with index ranges 0..-1 and 0..1, i.e. the
1340 first dimension has length zero, but the second has length 2.
1343 is a rank-zero array with contents 12.
1347 In addition, bytevectors are also arrays, but use a different syntax
1348 (@pxref{Bytevectors}):
1353 is a 3-byte long bytevector, with contents 1, 2, 3.
1357 @node Array Procedures
1358 @subsubsection Array Procedures
1360 When an array is created, the range of each dimension must be
1361 specified, e.g., to create a 2@cross{}3 array with a zero-based index:
1364 (make-array 'ho 2 3) @result{} #2((ho ho ho) (ho ho ho))
1367 The range of each dimension can also be given explicitly, e.g., another
1368 way to create the same array:
1371 (make-array 'ho '(0 1) '(0 2)) @result{} #2((ho ho ho) (ho ho ho))
1374 The following procedures can be used with arrays (or vectors). An
1375 argument shown as @var{idx}@dots{} means one parameter for each
1376 dimension in the array. A @var{idxlist} argument means a list of such
1377 values, one for each dimension.
1380 @deffn {Scheme Procedure} array? obj
1381 @deffnx {C Function} scm_array_p (obj, unused)
1382 Return @code{#t} if the @var{obj} is an array, and @code{#f} if
1385 The second argument to scm_array_p is there for historical reasons,
1386 but it is not used. You should always pass @code{SCM_UNDEFINED} as
1390 @deffn {Scheme Procedure} typed-array? obj type
1391 @deffnx {C Function} scm_typed_array_p (obj, type)
1392 Return @code{#t} if the @var{obj} is an array of type @var{type}, and
1396 @deftypefn {C Function} int scm_is_array (SCM obj)
1397 Return @code{1} if the @var{obj} is an array and @code{0} if not.
1400 @deftypefn {C Function} int scm_is_typed_array (SCM obj, SCM type)
1401 Return @code{0} if the @var{obj} is an array of type @var{type}, and
1405 @deffn {Scheme Procedure} make-array fill bound @dots{}
1406 @deffnx {C Function} scm_make_array (fill, bounds)
1407 Equivalent to @code{(make-typed-array #t @var{fill} @var{bound} ...)}.
1410 @deffn {Scheme Procedure} make-typed-array type fill bound @dots{}
1411 @deffnx {C Function} scm_make_typed_array (type, fill, bounds)
1412 Create and return an array that has as many dimensions as there are
1413 @var{bound}s and (maybe) fill it with @var{fill}.
1415 The underlying storage vector is created according to @var{type},
1416 which must be a symbol whose name is the `vectag' of the array as
1417 explained above, or @code{#t} for ordinary, non-specialized arrays.
1419 For example, using the symbol @code{f64} for @var{type} will create an
1420 array that uses a @code{f64vector} for storing its elements, and
1421 @code{a} will use a string.
1423 When @var{fill} is not the special @emph{unspecified} value, the new
1424 array is filled with @var{fill}. Otherwise, the initial contents of
1425 the array is unspecified. The special @emph{unspecified} value is
1426 stored in the variable @code{*unspecified*} so that for example
1427 @code{(make-typed-array 'u32 *unspecified* 4)} creates a uninitialized
1428 @code{u32} vector of length 4.
1430 Each @var{bound} may be a positive non-zero integer @var{N}, in which
1431 case the index for that dimension can range from 0 through @var{N-1}; or
1432 an explicit index range specifier in the form @code{(LOWER UPPER)},
1433 where both @var{lower} and @var{upper} are integers, possibly less than
1434 zero, and possibly the same number (however, @var{lower} cannot be
1435 greater than @var{upper}).
1438 @deffn {Scheme Procedure} list->array dimspec list
1439 Equivalent to @code{(list->typed-array #t @var{dimspec}
1443 @deffn {Scheme Procedure} list->typed-array type dimspec list
1444 @deffnx {C Function} scm_list_to_typed_array (type, dimspec, list)
1445 Return an array of the type indicated by @var{type} with elements the
1446 same as those of @var{list}.
1448 The argument @var{dimspec} determines the number of dimensions of the
1449 array and their lower bounds. When @var{dimspec} is an exact integer,
1450 it gives the number of dimensions directly and all lower bounds are
1451 zero. When it is a list of exact integers, then each element is the
1452 lower index bound of a dimension, and there will be as many dimensions
1453 as elements in the list.
1456 @deffn {Scheme Procedure} array-type array
1457 Return the type of @var{array}. This is the `vectag' used for
1458 printing @var{array} (or @code{#t} for ordinary arrays) and can be
1459 used with @code{make-typed-array} to create an array of the same kind
1463 @deffn {Scheme Procedure} array-ref array idx @dots{}
1464 Return the element at @code{(idx @dots{})} in @var{array}.
1467 (define a (make-array 999 '(1 2) '(3 4)))
1468 (array-ref a 2 4) @result{} 999
1472 @deffn {Scheme Procedure} array-in-bounds? array idx @dots{}
1473 @deffnx {C Function} scm_array_in_bounds_p (array, idxlist)
1474 Return @code{#t} if the given index would be acceptable to
1478 (define a (make-array #f '(1 2) '(3 4)))
1479 (array-in-bounds? a 2 3) @result{} #t
1480 (array-in-bounds? a 0 0) @result{} #f
1484 @deffn {Scheme Procedure} array-set! array obj idx @dots{}
1485 @deffnx {C Function} scm_array_set_x (array, obj, idxlist)
1486 Set the element at @code{(idx @dots{})} in @var{array} to @var{obj}.
1487 The return value is unspecified.
1490 (define a (make-array #f '(0 1) '(0 1)))
1491 (array-set! a #t 1 1)
1492 a @result{} #2((#f #f) (#f #t))
1496 @deffn {Scheme Procedure} array-shape array
1497 @deffnx {Scheme Procedure} array-dimensions array
1498 @deffnx {C Function} scm_array_dimensions (array)
1499 Return a list of the bounds for each dimension of @var{array}.
1501 @code{array-shape} gives @code{(@var{lower} @var{upper})} for each
1502 dimension. @code{array-dimensions} instead returns just
1503 @math{@var{upper}+1} for dimensions with a 0 lower bound. Both are
1504 suitable as input to @code{make-array}.
1509 (define a (make-array 'foo '(-1 3) 5))
1510 (array-shape a) @result{} ((-1 3) (0 4))
1511 (array-dimensions a) @result{} ((-1 3) 5)
1515 @deffn {Scheme Procedure} array-rank obj
1516 @deffnx {C Function} scm_array_rank (obj)
1517 Return the rank of @var{array}.
1520 @deftypefn {C Function} size_t scm_c_array_rank (SCM array)
1521 Return the rank of @var{array} as a @code{size_t}.
1524 @deffn {Scheme Procedure} array->list array
1525 @deffnx {C Function} scm_array_to_list (array)
1526 Return a list consisting of all the elements, in order, of
1530 @c FIXME: Describe how the order affects the copying (it matters for
1531 @c shared arrays with the same underlying root vector, presumably).
1533 @deffn {Scheme Procedure} array-copy! src dst
1534 @deffnx {Scheme Procedure} array-copy-in-order! src dst
1535 @deffnx {C Function} scm_array_copy_x (src, dst)
1536 Copy every element from vector or array @var{src} to the corresponding
1537 element of @var{dst}. @var{dst} must have the same rank as @var{src},
1538 and be at least as large in each dimension. The return value is
1542 @deffn {Scheme Procedure} array-fill! array fill
1543 @deffnx {C Function} scm_array_fill_x (array, fill)
1544 Store @var{fill} in every element of @var{array}. The value returned
1548 @c begin (texi-doc-string "guile" "array-equal?")
1549 @deffn {Scheme Procedure} array-equal? array1 array2 @dots{}
1550 Return @code{#t} if all arguments are arrays with the same shape, the
1551 same type, and have corresponding elements which are either
1552 @code{equal?} or @code{array-equal?}. This function differs from
1553 @code{equal?} (@pxref{Equality}) in that all arguments must be arrays.
1556 @c FIXME: array-map! accepts no source arrays at all, and in that
1557 @c case makes calls "(proc)". Is that meant to be a documented
1560 @c FIXME: array-for-each doesn't say what happens if the sources have
1561 @c different index ranges. The code currently iterates over the
1562 @c indices of the first and expects the others to cover those. That
1563 @c at least vaguely matches array-map!, but is is meant to be a
1564 @c documented feature?
1566 @deffn {Scheme Procedure} array-map! dst proc src1 @dots{} srcN
1567 @deffnx {Scheme Procedure} array-map-in-order! dst proc src1 @dots{} srcN
1568 @deffnx {C Function} scm_array_map_x (dst, proc, srclist)
1569 Set each element of the @var{dst} array to values obtained from calls
1570 to @var{proc}. The value returned is unspecified.
1572 Each call is @code{(@var{proc} @var{elem1} @dots{} @var{elemN})},
1573 where each @var{elem} is from the corresponding @var{src} array, at
1574 the @var{dst} index. @code{array-map-in-order!} makes the calls in
1575 row-major order, @code{array-map!} makes them in an unspecified order.
1577 The @var{src} arrays must have the same number of dimensions as
1578 @var{dst}, and must have a range for each dimension which covers the
1579 range in @var{dst}. This ensures all @var{dst} indices are valid in
1583 @deffn {Scheme Procedure} array-for-each proc src1 @dots{} srcN
1584 @deffnx {C Function} scm_array_for_each (proc, src1, srclist)
1585 Apply @var{proc} to each tuple of elements of @var{src1} @dots{}
1586 @var{srcN}, in row-major order. The value returned is unspecified.
1589 @deffn {Scheme Procedure} array-index-map! dst proc
1590 @deffnx {C Function} scm_array_index_map_x (dst, proc)
1591 Set each element of the @var{dst} array to values returned by calls to
1592 @var{proc}. The value returned is unspecified.
1594 Each call is @code{(@var{proc} @var{i1} @dots{} @var{iN})}, where
1595 @var{i1}@dots{}@var{iN} is the destination index, one parameter for
1596 each dimension. The order in which the calls are made is unspecified.
1598 For example, to create a @m{4\times4, 4x4} matrix representing a
1602 \advance\leftskip by 2\lispnarrowing {
1620 (define a (make-array #f 4 4))
1621 (array-index-map! a (lambda (i j)
1622 (modulo (+ i j) 4)))
1626 @deffn {Scheme Procedure} uniform-array-read! ra [port_or_fd [start [end]]]
1627 @deffnx {C Function} scm_uniform_array_read_x (ra, port_or_fd, start, end)
1628 Attempt to read all elements of @var{ura}, in lexicographic order, as
1629 binary objects from @var{port-or-fdes}.
1630 If an end of file is encountered,
1631 the objects up to that point are put into @var{ura}
1632 (starting at the beginning) and the remainder of the array is
1635 The optional arguments @var{start} and @var{end} allow
1636 a specified region of a vector (or linearized array) to be read,
1637 leaving the remainder of the vector unchanged.
1639 @code{uniform-array-read!} returns the number of objects read.
1640 @var{port-or-fdes} may be omitted, in which case it defaults to the value
1641 returned by @code{(current-input-port)}.
1644 @deffn {Scheme Procedure} uniform-array-write v [port_or_fd [start [end]]]
1645 @deffnx {C Function} scm_uniform_array_write (v, port_or_fd, start, end)
1646 Writes all elements of @var{ura} as binary objects to
1649 The optional arguments @var{start}
1651 a specified region of a vector (or linearized array) to be written.
1653 The number of objects actually written is returned.
1654 @var{port-or-fdes} may be
1655 omitted, in which case it defaults to the value returned by
1656 @code{(current-output-port)}.
1660 @subsubsection Shared Arrays
1662 @deffn {Scheme Procedure} make-shared-array oldarray mapfunc bound @dots{}
1663 @deffnx {C Function} scm_make_shared_array (oldarray, mapfunc, boundlist)
1664 Return a new array which shares the storage of @var{oldarray}.
1665 Changes made through either affect the same underlying storage. The
1666 @var{bound@dots{}} arguments are the shape of the new array, the same
1667 as @code{make-array} (@pxref{Array Procedures}).
1669 @var{mapfunc} translates coordinates from the new array to the
1670 @var{oldarray}. It's called as @code{(@var{mapfunc} newidx1 @dots{})}
1671 with one parameter for each dimension of the new array, and should
1672 return a list of indices for @var{oldarray}, one for each dimension of
1675 @var{mapfunc} must be affine linear, meaning that each @var{oldarray}
1676 index must be formed by adding integer multiples (possibly negative)
1677 of some or all of @var{newidx1} etc, plus a possible integer offset.
1678 The multiples and offset must be the same in each call.
1681 One good use for a shared array is to restrict the range of some
1682 dimensions, so as to apply say @code{array-for-each} or
1683 @code{array-fill!} to only part of an array. The plain @code{list}
1684 function can be used for @var{mapfunc} in this case, making no changes
1685 to the index values. For example,
1688 (make-shared-array #2((a b c) (d e f) (g h i)) list 3 2)
1689 @result{} #2((a b) (d e) (g h))
1692 The new array can have fewer dimensions than @var{oldarray}, for
1693 example to take a column from an array.
1696 (make-shared-array #2((a b c) (d e f) (g h i))
1697 (lambda (i) (list i 2))
1702 A diagonal can be taken by using the single new array index for both
1703 row and column in the old array. For example,
1706 (make-shared-array #2((a b c) (d e f) (g h i))
1707 (lambda (i) (list i i))
1712 Dimensions can be increased by for instance considering portions of a
1713 one dimensional array as rows in a two dimensional array.
1714 (@code{array-contents} below can do the opposite, flattening an
1718 (make-shared-array #1(a b c d e f g h i j k l)
1719 (lambda (i j) (list (+ (* i 3) j)))
1721 @result{} #2((a b c) (d e f) (g h i) (j k l))
1724 By negating an index the order that elements appear can be reversed.
1725 The following just reverses the column order,
1728 (make-shared-array #2((a b c) (d e f) (g h i))
1729 (lambda (i j) (list i (- 2 j)))
1731 @result{} #2((c b a) (f e d) (i h g))
1734 A fixed offset on indexes allows for instance a change from a 0 based
1738 (define x #2((a b c) (d e f) (g h i)))
1739 (define y (make-shared-array x
1740 (lambda (i j) (list (1- i) (1- j)))
1742 (array-ref x 0 0) @result{} a
1743 (array-ref y 1 1) @result{} a
1746 A multiple on an index allows every Nth element of an array to be
1747 taken. The following is every third element,
1750 (make-shared-array #1(a b c d e f g h i j k l)
1751 (lambda (i) (list (* i 3)))
1753 @result{} #1(a d g j)
1756 The above examples can be combined to make weird and wonderful
1757 selections from an array, but it's important to note that because
1758 @var{mapfunc} must be affine linear, arbitrary permutations are not
1761 In the current implementation, @var{mapfunc} is not called for every
1762 access to the new array but only on some sample points to establish a
1763 base and stride for new array indices in @var{oldarray} data. A few
1764 sample points are enough because @var{mapfunc} is linear.
1767 @deffn {Scheme Procedure} shared-array-increments array
1768 @deffnx {C Function} scm_shared_array_increments (array)
1769 For each dimension, return the distance between elements in the root vector.
1772 @deffn {Scheme Procedure} shared-array-offset array
1773 @deffnx {C Function} scm_shared_array_offset (array)
1774 Return the root vector index of the first element in the array.
1777 @deffn {Scheme Procedure} shared-array-root array
1778 @deffnx {C Function} scm_shared_array_root (array)
1779 Return the root vector of a shared array.
1782 @deffn {Scheme Procedure} array-contents array [strict]
1783 @deffnx {C Function} scm_array_contents (array, strict)
1784 If @var{array} may be @dfn{unrolled} into a one dimensional shared array
1785 without changing their order (last subscript changing fastest), then
1786 @code{array-contents} returns that shared array, otherwise it returns
1787 @code{#f}. All arrays made by @code{make-array} and
1788 @code{make-typed-array} may be unrolled, some arrays made by
1789 @code{make-shared-array} may not be.
1791 If the optional argument @var{strict} is provided, a shared array will
1792 be returned only if its elements are stored internally contiguous in
1796 @deffn {Scheme Procedure} transpose-array array dim1 @dots{}
1797 @deffnx {C Function} scm_transpose_array (array, dimlist)
1798 Return an array sharing contents with @var{array}, but with
1799 dimensions arranged in a different order. There must be one
1800 @var{dim} argument for each dimension of @var{array}.
1801 @var{dim1}, @var{dim2}, @dots{} should be integers between 0
1802 and the rank of the array to be returned. Each integer in that
1803 range must appear at least once in the argument list.
1805 The values of @var{dim1}, @var{dim2}, @dots{} correspond to
1806 dimensions in the array to be returned, and their positions in the
1807 argument list to dimensions of @var{array}. Several @var{dim}s
1808 may have the same value, in which case the returned array will
1809 have smaller rank than @var{array}.
1812 (transpose-array '#2((a b) (c d)) 1 0) @result{} #2((a c) (b d))
1813 (transpose-array '#2((a b) (c d)) 0 0) @result{} #1(a d)
1814 (transpose-array '#3(((a b c) (d e f)) ((1 2 3) (4 5 6))) 1 1 0) @result{}
1815 #2((a 4) (b 5) (c 6))
1819 @node Accessing Arrays from C
1820 @subsubsection Accessing Arrays from C
1822 Arrays, especially uniform numeric arrays, are useful to efficiently
1823 represent large amounts of rectangularily organized information, such as
1824 matrices, images, or generally blobs of binary data. It is desirable to
1825 access these blobs in a C like manner so that they can be handed to
1826 external C code such as linear algebra libraries or image processing
1829 While pointers to the elements of an array are in use, the array itself
1830 must be protected so that the pointer remains valid. Such a protected
1831 array is said to be @dfn{reserved}. A reserved array can be read but
1832 modifications to it that would cause the pointer to its elements to
1833 become invalid are prevented. When you attempt such a modification, an
1836 (This is similar to locking the array while it is in use, but without
1837 the danger of a deadlock. In a multi-threaded program, you will need
1838 additional synchronization to avoid modifying reserved arrays.)
1840 You must take care to always unreserve an array after reserving it,
1841 even in the presence of non-local exits. If a non-local exit can
1842 happen between these two calls, you should install a dynwind context
1843 that releases the array when it is left (@pxref{Dynamic Wind}).
1845 In addition, array reserving and unreserving must be properly
1846 paired. For instance, when reserving two or more arrays in a certain
1847 order, you need to unreserve them in the opposite order.
1849 Once you have reserved an array and have retrieved the pointer to its
1850 elements, you must figure out the layout of the elements in memory.
1851 Guile allows slices to be taken out of arrays without actually making a
1852 copy, such as making an alias for the diagonal of a matrix that can be
1853 treated as a vector. Arrays that result from such an operation are not
1854 stored contiguously in memory and when working with their elements
1855 directly, you need to take this into account.
1857 The layout of array elements in memory can be defined via a
1858 @emph{mapping function} that computes a scalar position from a vector of
1859 indices. The scalar position then is the offset of the element with the
1860 given indices from the start of the storage block of the array.
1862 In Guile, this mapping function is restricted to be @dfn{affine}: all
1863 mapping functions of Guile arrays can be written as @code{p = b +
1864 c[0]*i[0] + c[1]*i[1] + ... + c[n-1]*i[n-1]} where @code{i[k]} is the
1865 @nicode{k}th index and @code{n} is the rank of the array. For
1866 example, a matrix of size 3x3 would have @code{b == 0}, @code{c[0] ==
1867 3} and @code{c[1] == 1}. When you transpose this matrix (with
1868 @code{transpose-array}, say), you will get an array whose mapping
1869 function has @code{b == 0}, @code{c[0] == 1} and @code{c[1] == 3}.
1871 The function @code{scm_array_handle_dims} gives you (indirect) access to
1872 the coefficients @code{c[k]}.
1875 Note that there are no functions for accessing the elements of a
1876 character array yet. Once the string implementation of Guile has been
1877 changed to use Unicode, we will provide them.
1879 @deftp {C Type} scm_t_array_handle
1880 This is a structure type that holds all information necessary to manage
1881 the reservation of arrays as explained above. Structures of this type
1882 must be allocated on the stack and must only be accessed by the
1883 functions listed below.
1886 @deftypefn {C Function} void scm_array_get_handle (SCM array, scm_t_array_handle *handle)
1887 Reserve @var{array}, which must be an array, and prepare @var{handle} to
1888 be used with the functions below. You must eventually call
1889 @code{scm_array_handle_release} on @var{handle}, and do this in a
1890 properly nested fashion, as explained above. The structure pointed to
1891 by @var{handle} does not need to be initialized before calling this
1895 @deftypefn {C Function} void scm_array_handle_release (scm_t_array_handle *handle)
1896 End the array reservation represented by @var{handle}. After a call to
1897 this function, @var{handle} might be used for another reservation.
1900 @deftypefn {C Function} size_t scm_array_handle_rank (scm_t_array_handle *handle)
1901 Return the rank of the array represented by @var{handle}.
1904 @deftp {C Type} scm_t_array_dim
1905 This structure type holds information about the layout of one dimension
1906 of an array. It includes the following fields:
1911 The lower and upper bounds (both inclusive) of the permissible index
1912 range for the given dimension. Both values can be negative, but
1913 @var{lbnd} is always less than or equal to @var{ubnd}.
1916 The distance from one element of this dimension to the next. Note, too,
1917 that this can be negative.
1921 @deftypefn {C Function} {const scm_t_array_dim *} scm_array_handle_dims (scm_t_array_handle *handle)
1922 Return a pointer to a C vector of information about the dimensions of
1923 the array represented by @var{handle}. This pointer is valid as long as
1924 the array remains reserved. As explained above, the
1925 @code{scm_t_array_dim} structures returned by this function can be used
1926 calculate the position of an element in the storage block of the array
1929 This position can then be used as an index into the C array pointer
1930 returned by the various @code{scm_array_handle_<foo>_elements}
1931 functions, or with @code{scm_array_handle_ref} and
1932 @code{scm_array_handle_set}.
1934 Here is how one can compute the position @var{pos} of an element given
1935 its indices in the vector @var{indices}:
1938 ssize_t indices[RANK];
1939 scm_t_array_dim *dims;
1944 for (i = 0; i < RANK; i++)
1946 if (indices[i] < dims[i].lbnd || indices[i] > dims[i].ubnd)
1948 pos += (indices[i] - dims[i].lbnd) * dims[i].inc;
1953 @deftypefn {C Function} ssize_t scm_array_handle_pos (scm_t_array_handle *handle, SCM indices)
1954 Compute the position corresponding to @var{indices}, a list of
1955 indices. The position is computed as described above for
1956 @code{scm_array_handle_dims}. The number of the indices and their
1957 range is checked and an appropriate error is signalled for invalid
1961 @deftypefn {C Function} SCM scm_array_handle_ref (scm_t_array_handle *handle, ssize_t pos)
1962 Return the element at position @var{pos} in the storage block of the
1963 array represented by @var{handle}. Any kind of array is acceptable. No
1964 range checking is done on @var{pos}.
1967 @deftypefn {C Function} void scm_array_handle_set (scm_t_array_handle *handle, ssize_t pos, SCM val)
1968 Set the element at position @var{pos} in the storage block of the array
1969 represented by @var{handle} to @var{val}. Any kind of array is
1970 acceptable. No range checking is done on @var{pos}. An error is
1971 signalled when the array can not store @var{val}.
1974 @deftypefn {C Function} {const SCM *} scm_array_handle_elements (scm_t_array_handle *handle)
1975 Return a pointer to the elements of a ordinary array of general Scheme
1976 values (i.e., a non-uniform array) for reading. This pointer is valid
1977 as long as the array remains reserved.
1980 @deftypefn {C Function} {SCM *} scm_array_handle_writable_elements (scm_t_array_handle *handle)
1981 Like @code{scm_array_handle_elements}, but the pointer is good for
1982 reading and writing.
1985 @deftypefn {C Function} {const void *} scm_array_handle_uniform_elements (scm_t_array_handle *handle)
1986 Return a pointer to the elements of a uniform numeric array for reading.
1987 This pointer is valid as long as the array remains reserved. The size
1988 of each element is given by @code{scm_array_handle_uniform_element_size}.
1991 @deftypefn {C Function} {void *} scm_array_handle_uniform_writable_elements (scm_t_array_handle *handle)
1992 Like @code{scm_array_handle_uniform_elements}, but the pointer is good
1993 reading and writing.
1996 @deftypefn {C Function} size_t scm_array_handle_uniform_element_size (scm_t_array_handle *handle)
1997 Return the size of one element of the uniform numeric array represented
2001 @deftypefn {C Function} {const scm_t_uint8 *} scm_array_handle_u8_elements (scm_t_array_handle *handle)
2002 @deftypefnx {C Function} {const scm_t_int8 *} scm_array_handle_s8_elements (scm_t_array_handle *handle)
2003 @deftypefnx {C Function} {const scm_t_uint16 *} scm_array_handle_u16_elements (scm_t_array_handle *handle)
2004 @deftypefnx {C Function} {const scm_t_int16 *} scm_array_handle_s16_elements (scm_t_array_handle *handle)
2005 @deftypefnx {C Function} {const scm_t_uint32 *} scm_array_handle_u32_elements (scm_t_array_handle *handle)
2006 @deftypefnx {C Function} {const scm_t_int32 *} scm_array_handle_s32_elements (scm_t_array_handle *handle)
2007 @deftypefnx {C Function} {const scm_t_uint64 *} scm_array_handle_u64_elements (scm_t_array_handle *handle)
2008 @deftypefnx {C Function} {const scm_t_int64 *} scm_array_handle_s64_elements (scm_t_array_handle *handle)
2009 @deftypefnx {C Function} {const float *} scm_array_handle_f32_elements (scm_t_array_handle *handle)
2010 @deftypefnx {C Function} {const double *} scm_array_handle_f64_elements (scm_t_array_handle *handle)
2011 @deftypefnx {C Function} {const float *} scm_array_handle_c32_elements (scm_t_array_handle *handle)
2012 @deftypefnx {C Function} {const double *} scm_array_handle_c64_elements (scm_t_array_handle *handle)
2013 Return a pointer to the elements of a uniform numeric array of the
2014 indicated kind for reading. This pointer is valid as long as the array
2017 The pointers for @code{c32} and @code{c64} uniform numeric arrays point
2018 to pairs of floating point numbers. The even index holds the real part,
2019 the odd index the imaginary part of the complex number.
2022 @deftypefn {C Function} {scm_t_uint8 *} scm_array_handle_u8_writable_elements (scm_t_array_handle *handle)
2023 @deftypefnx {C Function} {scm_t_int8 *} scm_array_handle_s8_writable_elements (scm_t_array_handle *handle)
2024 @deftypefnx {C Function} {scm_t_uint16 *} scm_array_handle_u16_writable_elements (scm_t_array_handle *handle)
2025 @deftypefnx {C Function} {scm_t_int16 *} scm_array_handle_s16_writable_elements (scm_t_array_handle *handle)
2026 @deftypefnx {C Function} {scm_t_uint32 *} scm_array_handle_u32_writable_elements (scm_t_array_handle *handle)
2027 @deftypefnx {C Function} {scm_t_int32 *} scm_array_handle_s32_writable_elements (scm_t_array_handle *handle)
2028 @deftypefnx {C Function} {scm_t_uint64 *} scm_array_handle_u64_writable_elements (scm_t_array_handle *handle)
2029 @deftypefnx {C Function} {scm_t_int64 *} scm_array_handle_s64_writable_elements (scm_t_array_handle *handle)
2030 @deftypefnx {C Function} {float *} scm_array_handle_f32_writable_elements (scm_t_array_handle *handle)
2031 @deftypefnx {C Function} {double *} scm_array_handle_f64_writable_elements (scm_t_array_handle *handle)
2032 @deftypefnx {C Function} {float *} scm_array_handle_c32_writable_elements (scm_t_array_handle *handle)
2033 @deftypefnx {C Function} {double *} scm_array_handle_c64_writable_elements (scm_t_array_handle *handle)
2034 Like @code{scm_array_handle_<kind>_elements}, but the pointer is good
2035 for reading and writing.
2038 @deftypefn {C Function} {const scm_t_uint32 *} scm_array_handle_bit_elements (scm_t_array_handle *handle)
2039 Return a pointer to the words that store the bits of the represented
2040 array, which must be a bit array.
2042 Unlike other arrays, bit arrays have an additional offset that must be
2043 figured into index calculations. That offset is returned by
2044 @code{scm_array_handle_bit_elements_offset}.
2046 To find a certain bit you first need to calculate its position as
2047 explained above for @code{scm_array_handle_dims} and then add the
2048 offset. This gives the absolute position of the bit, which is always a
2049 non-negative integer.
2051 Each word of the bit array storage block contains exactly 32 bits, with
2052 the least significant bit in that word having the lowest absolute
2053 position number. The next word contains the next 32 bits.
2055 Thus, the following code can be used to access a bit whose position
2056 according to @code{scm_array_handle_dims} is given in @var{pos}:
2060 scm_t_array_handle handle;
2064 size_t word_pos, mask;
2066 scm_array_get_handle (&bit_array, &handle);
2067 bits = scm_array_handle_bit_elements (&handle);
2070 abs_pos = pos + scm_array_handle_bit_elements_offset (&handle);
2071 word_pos = abs_pos / 32;
2072 mask = 1L << (abs_pos % 32);
2074 if (bits[word_pos] & mask)
2077 scm_array_handle_release (&handle);
2082 @deftypefn {C Function} {scm_t_uint32 *} scm_array_handle_bit_writable_elements (scm_t_array_handle *handle)
2083 Like @code{scm_array_handle_bit_elements} but the pointer is good for
2084 reading and writing. You must take care not to modify bits outside of
2085 the allowed index range of the array, even for contiguous arrays.
2093 The @code{(ice-9 vlist)} module provides an implementation of the @dfn{VList}
2094 data structure designed by Phil Bagwell in 2002. VLists are immutable lists,
2095 which can contain any Scheme object. They improve on standard Scheme linked
2096 lists in several areas:
2100 Random access has typically constant-time complexity.
2103 Computing the length of a VList has time complexity logarithmic in the number of
2107 VLists use less storage space than standard lists.
2110 VList elements are stored in contiguous regions, which improves memory locality
2111 and leads to more efficient use of hardware caches.
2114 The idea behind VLists is to store vlist elements in increasingly large
2115 contiguous blocks (implemented as vectors here). These blocks are linked to one
2116 another using a pointer to the next block and an offset within that block. The
2117 size of these blocks form a geometric series with ratio
2118 @code{block-growth-factor} (2 by default).
2120 The VList structure also serves as the basis for the @dfn{VList-based hash
2121 lists} or ``vhashes'', an immutable dictionary type (@pxref{VHashes}).
2123 However, the current implementation in @code{(ice-9 vlist)} has several
2124 noteworthy shortcomings:
2129 It is @emph{not} thread-safe. Although operations on vlists are all
2130 @dfn{referentially transparent} (i.e., purely functional), adding elements to a
2131 vlist with @code{vlist-cons} mutates part of its internal structure, which makes
2132 it non-thread-safe. This could be fixed, but it would slow down
2136 @code{vlist-cons} always allocates at least as much memory as @code{cons}.
2137 Again, Phil Bagwell describes how to fix it, but that would require tuning the
2138 garbage collector in a way that may not be generally beneficial.
2141 @code{vlist-cons} is a Scheme procedure compiled to bytecode, and it does not
2142 compete with the straightforward C implementation of @code{cons}, and with the
2143 fact that the VM has a special @code{cons} instruction.
2147 We hope to address these in the future.
2149 The programming interface exported by @code{(ice-9 vlist)} is defined below.
2150 Most of it is the same as SRFI-1 with an added @code{vlist-} prefix to function
2153 @deffn {Scheme Procedure} vlist? obj
2154 Return true if @var{obj} is a VList.
2157 @defvr {Scheme Variable} vlist-null
2158 The empty VList. Note that it's possible to create an empty VList not
2159 @code{eq?} to @code{vlist-null}; thus, callers should always use
2160 @code{vlist-null?} when testing whether a VList is empty.
2163 @deffn {Scheme Procedure} vlist-null? vlist
2164 Return true if @var{vlist} is empty.
2167 @deffn {Scheme Procedure} vlist-cons item vlist
2168 Return a new vlist with @var{item} as its head and @var{vlist} as its tail.
2171 @deffn {Scheme Procedure} vlist-head vlist
2172 Return the head of @var{vlist}.
2175 @deffn {Scheme Procedure} vlist-tail vlist
2176 Return the tail of @var{vlist}.
2179 @defvr {Scheme Variable} block-growth-factor
2180 A fluid that defines the growth factor of VList blocks, 2 by default.
2183 The functions below provide the usual set of higher-level list operations.
2185 @deffn {Scheme Procedure} vlist-fold proc init vlist
2186 @deffnx {Scheme Procedure} vlist-fold-right proc init vlist
2187 Fold over @var{vlist}, calling @var{proc} for each element, as for SRFI-1
2188 @code{fold} and @code{fold-right} (@pxref{SRFI-1, @code{fold}}).
2191 @deffn {Scheme Procedure} vlist-ref vlist index
2192 Return the element at index @var{index} in @var{vlist}. This is typically a
2193 constant-time operation.
2196 @deffn {Scheme Procedure} vlist-length vlist
2197 Return the length of @var{vlist}. This is typically logarithmic in the number
2198 of elements in @var{vlist}.
2201 @deffn {Scheme Procedure} vlist-reverse vlist
2202 Return a new @var{vlist} whose content are those of @var{vlist} in reverse
2206 @deffn {Scheme Procedure} vlist-map proc vlist
2207 Map @var{proc} over the elements of @var{vlist} and return a new vlist.
2210 @deffn {Scheme Procedure} vlist-for-each proc vlist
2211 Call @var{proc} on each element of @var{vlist}. The result is unspecified.
2214 @deffn {Scheme Procedure} vlist-drop vlist count
2215 Return a new vlist that does not contain the @var{count} first elements of
2216 @var{vlist}. This is typically a constant-time operation.
2219 @deffn {Scheme Procedure} vlist-take vlist count
2220 Return a new vlist that contains only the @var{count} first elements of
2224 @deffn {Scheme Procedure} vlist-filter pred vlist
2225 Return a new vlist containing all the elements from @var{vlist} that satisfy
2229 @deffn {Scheme Procedure} vlist-delete x vlist [equal?]
2230 Return a new vlist corresponding to @var{vlist} without the elements
2231 @var{equal?} to @var{x}.
2234 @deffn {Scheme Procedure} vlist-unfold p f g seed [tail-gen]
2235 @deffnx {Scheme Procedure} vlist-unfold-right p f g seed [tail]
2236 Return a new vlist, as for SRFI-1 @code{unfold} and @code{unfold-right}
2237 (@pxref{SRFI-1, @code{unfold}}).
2240 @deffn {Scheme Procedure} vlist-append vlists ...
2241 Append the given vlists and return the resulting vlist.
2244 @deffn {Scheme Procedure} list->vlist lst
2245 Return a new vlist whose contents correspond to @var{lst}.
2248 @deffn {Scheme Procedure} vlist->list vlist
2249 Return a new list whose contents match those of @var{vlist}.
2257 A @dfn{record type} is a first class object representing a user-defined
2258 data type. A @dfn{record} is an instance of a record type.
2260 @deffn {Scheme Procedure} record? obj
2261 Return @code{#t} if @var{obj} is a record of any type and @code{#f}
2264 Note that @code{record?} may be true of any Scheme value; there is no
2265 promise that records are disjoint with other Scheme types.
2268 @deffn {Scheme Procedure} make-record-type type-name field-names [print]
2269 Create and return a new @dfn{record-type descriptor}.
2271 @var{type-name} is a string naming the type. Currently it's only used
2272 in the printed representation of records, and in diagnostics.
2273 @var{field-names} is a list of symbols naming the fields of a record
2274 of the type. Duplicates are not allowed among these symbols.
2277 (make-record-type "employee" '(name age salary))
2280 The optional @var{print} argument is a function used by
2281 @code{display}, @code{write}, etc, for printing a record of the new
2282 type. It's called as @code{(@var{print} record port)} and should look
2283 at @var{record} and write to @var{port}.
2286 @deffn {Scheme Procedure} record-constructor rtd [field-names]
2287 Return a procedure for constructing new members of the type represented
2288 by @var{rtd}. The returned procedure accepts exactly as many arguments
2289 as there are symbols in the given list, @var{field-names}; these are
2290 used, in order, as the initial values of those fields in a new record,
2291 which is returned by the constructor procedure. The values of any
2292 fields not named in that list are unspecified. The @var{field-names}
2293 argument defaults to the list of field names in the call to
2294 @code{make-record-type} that created the type represented by @var{rtd};
2295 if the @var{field-names} argument is provided, it is an error if it
2296 contains any duplicates or any symbols not in the default list.
2299 @deffn {Scheme Procedure} record-predicate rtd
2300 Return a procedure for testing membership in the type represented by
2301 @var{rtd}. The returned procedure accepts exactly one argument and
2302 returns a true value if the argument is a member of the indicated record
2303 type; it returns a false value otherwise.
2306 @deffn {Scheme Procedure} record-accessor rtd field-name
2307 Return a procedure for reading the value of a particular field of a
2308 member of the type represented by @var{rtd}. The returned procedure
2309 accepts exactly one argument which must be a record of the appropriate
2310 type; it returns the current value of the field named by the symbol
2311 @var{field-name} in that record. The symbol @var{field-name} must be a
2312 member of the list of field-names in the call to @code{make-record-type}
2313 that created the type represented by @var{rtd}.
2316 @deffn {Scheme Procedure} record-modifier rtd field-name
2317 Return a procedure for writing the value of a particular field of a
2318 member of the type represented by @var{rtd}. The returned procedure
2319 accepts exactly two arguments: first, a record of the appropriate type,
2320 and second, an arbitrary Scheme value; it modifies the field named by
2321 the symbol @var{field-name} in that record to contain the given value.
2322 The returned value of the modifier procedure is unspecified. The symbol
2323 @var{field-name} must be a member of the list of field-names in the call
2324 to @code{make-record-type} that created the type represented by
2328 @deffn {Scheme Procedure} record-type-descriptor record
2329 Return a record-type descriptor representing the type of the given
2330 record. That is, for example, if the returned descriptor were passed to
2331 @code{record-predicate}, the resulting predicate would return a true
2332 value when passed the given record. Note that it is not necessarily the
2333 case that the returned descriptor is the one that was passed to
2334 @code{record-constructor} in the call that created the constructor
2335 procedure that created the given record.
2338 @deffn {Scheme Procedure} record-type-name rtd
2339 Return the type-name associated with the type represented by rtd. The
2340 returned value is @code{eqv?} to the @var{type-name} argument given in
2341 the call to @code{make-record-type} that created the type represented by
2345 @deffn {Scheme Procedure} record-type-fields rtd
2346 Return a list of the symbols naming the fields in members of the type
2347 represented by @var{rtd}. The returned value is @code{equal?} to the
2348 field-names argument given in the call to @code{make-record-type} that
2349 created the type represented by @var{rtd}.
2354 @subsection Structures
2357 A @dfn{structure} is a first class data type which holds Scheme values
2358 or C words in fields numbered 0 upwards. A @dfn{vtable} represents a
2359 structure type, giving field types and permissions, and an optional
2360 print function for @code{write} etc.
2362 Structures are lower level than records (@pxref{Records}) but have
2363 some extra features. The vtable system allows sets of types be
2364 constructed, with class data. The uninterpreted words can
2365 inter-operate with C code, allowing arbitrary pointers or other values
2366 to be stored along side usual Scheme @code{SCM} values.
2370 * Structure Basics::
2375 @node Vtables, Structure Basics, Structures, Structures
2376 @subsubsection Vtables
2378 A vtable is a structure type, specifying its layout, and other
2379 information. A vtable is actually itself a structure, but there's no
2380 need to worray about that initially (@pxref{Vtable Contents}.)
2382 @deffn {Scheme Procedure} make-vtable fields [print]
2383 Create a new vtable.
2385 @var{fields} is a string describing the fields in the structures to be
2386 created. Each field is represented by two characters, a type letter
2387 and a permissions letter, for example @code{"pw"}. The types are as
2392 @code{p} -- a Scheme value. ``p'' stands for ``protected'' meaning
2393 it's protected against garbage collection.
2396 @code{u} -- an arbitrary word of data (an @code{scm_t_bits}). At the
2397 Scheme level it's read and written as an unsigned integer. ``u''
2398 stands for ``uninterpreted'' (it's not treated as a Scheme value), or
2399 ``unprotected'' (it's not marked during GC), or ``unsigned long'' (its
2400 size), or all of these things.
2403 @code{s} -- a self-reference. Such a field holds the @code{SCM} value
2404 of the structure itself (a circular reference). This can be useful in
2405 C code where you might have a pointer to the data array, and want to
2406 get the Scheme @code{SCM} handle for the structure. In Scheme code it
2410 The second letter for each field is a permission code,
2414 @code{w} -- writable, the field can be read and written.
2416 @code{r} -- read-only, the field can be read but not written.
2418 @code{o} -- opaque, the field can be neither read nor written at the
2419 Scheme level. This can be used for fields which should only be used
2422 @code{W},@code{R},@code{O} -- a tail array, with permissions for the
2423 array fields as per @code{w},@code{r},@code{o}.
2426 A tail array is further fields at the end of a structure. The last
2427 field in the layout string might be for instance @samp{pW} to have a
2428 tail of writable Scheme-valued fields. The @samp{pW} field itself
2429 holds the tail size, and the tail fields come after it.
2431 Here are some examples.
2434 (make-vtable "pw") ;; one writable field
2435 (make-vtable "prpw") ;; one read-only and one writable
2436 (make-vtable "pwuwuw") ;; one scheme and two uninterpreted
2438 (make-vtable "prpW") ;; one fixed then a tail array
2441 The optional @var{print} argument is a function called by
2442 @code{display} and @code{write} (etc) to give a printed representation
2443 of a structure created from this vtable. It's called
2444 @code{(@var{print} struct port)} and should look at @var{struct} and
2445 write to @var{port}. The default print merely gives a form like
2446 @samp{#<struct ADDR:ADDR>} with a pair of machine addresses.
2448 The following print function for example shows the two fields of its
2453 (lambda (struct port)
2455 (display (struct-ref struct 0) port)
2456 (display " and " port)
2457 (display (struct-ref struct 1) port)
2458 (display ">" port)))
2463 @node Structure Basics, Vtable Contents, Vtables, Structures
2464 @subsubsection Structure Basics
2466 This section describes the basic procedures for working with
2467 structures. @code{make-struct} creates a structure, and
2468 @code{struct-ref} and @code{struct-set!} access write fields.
2470 @deffn {Scheme Procedure} make-struct vtable tail-size [init...]
2471 @deffnx {C Function} scm_make_struct (vtable, tail_size, init_list)
2472 Create a new structure, with layout per the given @var{vtable}
2475 @var{tail-size} is the size of the tail array if @var{vtable}
2476 specifies a tail array. @var{tail-size} should be 0 when @var{vtable}
2477 doesn't specify a tail array.
2479 The optional @var{init}@dots{} arguments are initial values for the
2480 fields of the structure (and the tail array). This is the only way to
2481 put values in read-only fields. If there are fewer @var{init}
2482 arguments than fields then the defaults are @code{#f} for a Scheme
2483 field (type @code{p}) or 0 for an uninterpreted field (type @code{u}).
2485 Type @code{s} self-reference fields, permission @code{o} opaque
2486 fields, and the count field of a tail array are all ignored for the
2487 @var{init} arguments, ie.@: an argument is not consumed by such a
2488 field. An @code{s} is always set to the structure itself, an @code{o}
2489 is always set to @code{#f} or 0 (with the intention that C code will
2490 do something to it later), and the tail count is always the given
2496 (define v (make-vtable "prpwpw"))
2497 (define s (make-struct v 0 123 "abc" 456))
2498 (struct-ref s 0) @result{} 123
2499 (struct-ref s 1) @result{} "abc"
2503 (define v (make-vtable "prpW"))
2504 (define s (make-struct v 6 "fixed field" 'x 'y))
2505 (struct-ref s 0) @result{} "fixed field"
2506 (struct-ref s 1) @result{} 2 ;; tail size
2507 (struct-ref s 2) @result{} x ;; tail array ...
2508 (struct-ref s 3) @result{} y
2509 (struct-ref s 4) @result{} #f
2513 @deffn {Scheme Procedure} struct? obj
2514 @deffnx {C Function} scm_struct_p (obj)
2515 Return @code{#t} if @var{obj} is a structure, or @code{#f} if not.
2518 @deffn {Scheme Procedure} struct-ref struct n
2519 @deffnx {C Function} scm_struct_ref (struct, n)
2520 Return the contents of field number @var{n} in @var{struct}. The
2521 first field is number 0.
2523 An error is thrown if @var{n} is out of range, or if the field cannot
2524 be read because it's @code{o} opaque.
2527 @deffn {Scheme Procedure} struct-set! struct n value
2528 @deffnx {C Function} scm_struct_set_x (struct, n, value)
2529 Set field number @var{n} in @var{struct} to @var{value}. The first
2532 An error is thrown if @var{n} is out of range, or if the field cannot
2533 be written because it's @code{r} read-only or @code{o} opaque.
2536 @deffn {Scheme Procedure} struct-vtable struct
2537 @deffnx {C Function} scm_struct_vtable (struct)
2538 Return the vtable used by @var{struct}.
2540 This can be used to examine the layout of an unknown structure, see
2541 @ref{Vtable Contents}.
2545 @node Vtable Contents, Vtable Vtables, Structure Basics, Structures
2546 @subsubsection Vtable Contents
2548 A vtable is itself a structure, with particular fields that hold
2549 information about the structures to be created. These include the
2550 fields of those structures, and the print function for them. The
2551 variables below allow access to those fields.
2553 @deffn {Scheme Procedure} struct-vtable? obj
2554 @deffnx {C Function} scm_struct_vtable_p (obj)
2555 Return @code{#t} if @var{obj} is a vtable structure.
2557 Note that because vtables are simply structures with a particular
2558 layout, @code{struct-vtable?} can potentially return true on an
2559 application structure which merely happens to look like a vtable.
2562 @defvr {Scheme Variable} vtable-index-layout
2563 @defvrx {C Macro} scm_vtable_index_layout
2564 The field number of the layout specification in a vtable. The layout
2565 specification is a symbol like @code{pwpw} formed from the fields
2566 string passed to @code{make-vtable}, or created by
2567 @code{make-struct-layout} (@pxref{Vtable Vtables}).
2570 (define v (make-vtable "pwpw" 0))
2571 (struct-ref v vtable-index-layout) @result{} pwpw
2574 This field is read-only, since the layout of structures using a vtable
2578 @defvr {Scheme Variable} vtable-index-vtable
2579 @defvrx {C Macro} scm_vtable_index_vtable
2580 A self-reference to the vtable, ie.@: a type @code{s} field. This is
2581 used by C code within Guile and has no use at the Scheme level.
2584 @defvr {Scheme Variable} vtable-index-printer
2585 @defvrx {C Macro} scm_vtable_index_printer
2586 The field number of the printer function. This field contains @code{#f}
2587 if the default print function should be used.
2590 (define (my-print-func struct port)
2592 (define v (make-vtable "pwpw" my-print-func))
2593 (struct-ref v vtable-index-printer) @result{} my-print-func
2596 This field is writable, allowing the print function to be changed
2600 @deffn {Scheme Procedure} struct-vtable-name vtable
2601 @deffnx {Scheme Procedure} set-struct-vtable-name! vtable name
2602 @deffnx {C Function} scm_struct_vtable_name (vtable)
2603 @deffnx {C Function} scm_set_struct_vtable_name_x (vtable, name)
2604 Get or set the name of @var{vtable}. @var{name} is a symbol and is
2605 used in the default print function when printing structures created
2609 (define v (make-vtable "pw"))
2610 (set-struct-vtable-name! v 'my-name)
2612 (define s (make-struct v 0))
2613 (display s) @print{} #<my-name b7ab3ae0:b7ab3730>
2617 @deffn {Scheme Procedure} struct-vtable-tag vtable
2618 @deffnx {C Function} scm_struct_vtable_tag (vtable)
2619 Return the tag of the given @var{vtable}.
2621 @c FIXME: what can be said about what this means?
2626 @node Vtable Vtables, , Vtable Contents, Structures
2627 @subsubsection Vtable Vtables
2629 As noted above, a vtable is a structure and that structure is itself
2630 described by a vtable. Such a ``vtable of a vtable'' can be created
2631 with @code{make-vtable-vtable} below. This can be used to build sets
2632 of related vtables, possibly with extra application fields.
2634 This second level of vtable can be a little confusing. The ball
2635 example below is a typical use, adding a ``class data'' field to the
2636 vtables, from which instance structures are created. The current
2637 implementation of Guile's own records (@pxref{Records}) does something
2638 similar, a record type descriptor is a vtable with room to hold the
2639 field names of the records to be created from it.
2641 @deffn {Scheme Procedure} make-vtable-vtable user-fields tail-size [print]
2642 @deffnx {C Function} scm_make_vtable_vtable (user_fields, tail_size, print_and_init_list)
2643 Create a ``vtable-vtable'' which can be used to create vtables. This
2644 vtable-vtable is also a vtable, and is self-describing, meaning its
2645 vtable is itself. The following is a simple usage.
2648 (define vt-vt (make-vtable-vtable "" 0))
2649 (define vt (make-struct vt-vt 0
2650 (make-struct-layout "pwpw"))
2651 (define s (make-struct vt 0 123 456))
2653 (struct-ref s 0) @result{} 123
2656 @code{make-struct} is used to create a vtable from the vtable-vtable.
2657 The first initializer is a layout object (field
2658 @code{vtable-index-layout}), usually obtained from
2659 @code{make-struct-layout} (below). An optional second initializer is
2660 a printer function (field @code{vtable-index-printer}), used as
2661 described under @code{make-vtable} (@pxref{Vtables}).
2664 @var{user-fields} is a layout string giving extra fields to have in
2665 the vtables. A vtable starts with some base fields as per @ref{Vtable
2666 Contents}, and @var{user-fields} is appended. The @var{user-fields}
2667 start at field number @code{vtable-offset-user} (below), and exist in
2668 both the vtable-vtable and in the vtables created from it. Such
2669 fields provide space for ``class data''. For example,
2672 (define vt-of-vt (make-vtable-vtable "pw" 0))
2673 (define vt (make-struct vt-of-vt 0))
2674 (struct-set! vt vtable-offset-user "my class data")
2677 @var{tail-size} is the size of the tail array in the vtable-vtable
2678 itself, if @var{user-fields} specifies a tail array. This should be 0
2679 if nothing extra is required or the format has no tail array. The
2680 tail array field such as @samp{pW} holds the tail array size, as
2681 usual, and is followed by the extra space.
2684 (define vt-vt (make-vtable-vtable "pW" 20))
2685 (define my-vt-tail-start (1+ vtable-offset-user))
2686 (struct-set! vt-vt (+ 3 my-vt-tail-start) "data in tail")
2689 The optional @var{print} argument is used by @code{display} and
2690 @code{write} (etc) to print the vtable-vtable and any vtables created
2691 from it. It's called as @code{(@var{print} vtable port)} and should
2692 look at @var{vtable} and write to @var{port}. The default is the
2693 usual structure print function, which just gives machine addresses.
2696 @deffn {Scheme Procedure} make-struct-layout fields
2697 @deffnx {C Function} scm_make_struct_layout (fields)
2698 Return a structure layout symbol, from a @var{fields} string.
2699 @var{fields} is as described under @code{make-vtable}
2700 (@pxref{Vtables}). An invalid @var{fields} string is an error.
2703 (make-struct-layout "prpW") @result{} prpW
2704 (make-struct-layout "blah") @result{} ERROR
2708 @defvr {Scheme Variable} vtable-offset-user
2709 @defvrx {C Macro} scm_vtable_offset_user
2710 The first field in a vtable which is available for application use.
2711 Such fields only exist when specified by @var{user-fields} in
2712 @code{make-vtable-vtable} above.
2716 Here's an extended vtable-vtable example, creating classes of
2717 ``balls''. Each class has a ``colour'', which is fixed. Instances of
2718 those classes are created, and such each such ball has an ``owner'',
2719 which can be changed.
2722 (define ball-root (make-vtable-vtable "pr" 0))
2724 (define (make-ball-type ball-color)
2725 (make-struct ball-root 0
2726 (make-struct-layout "pw")
2728 (format port "#<a ~A ball owned by ~A>"
2732 (define (color ball)
2733 (struct-ref (struct-vtable ball) vtable-offset-user))
2734 (define (owner ball)
2735 (struct-ref ball 0))
2737 (define red (make-ball-type 'red))
2738 (define green (make-ball-type 'green))
2740 (define (make-ball type owner) (make-struct type 0 owner))
2742 (define ball (make-ball green 'Nisse))
2743 ball @result{} #<a green ball owned by Nisse>
2747 @node Dictionary Types
2748 @subsection Dictionary Types
2750 A @dfn{dictionary} object is a data structure used to index
2751 information in a user-defined way. In standard Scheme, the main
2752 aggregate data types are lists and vectors. Lists are not really
2753 indexed at all, and vectors are indexed only by number
2754 (e.g. @code{(vector-ref foo 5)}). Often you will find it useful
2755 to index your data on some other type; for example, in a library
2756 catalog you might want to look up a book by the name of its
2757 author. Dictionaries are used to help you organize information in
2760 An @dfn{association list} (or @dfn{alist} for short) is a list of
2761 key-value pairs. Each pair represents a single quantity or
2762 object; the @code{car} of the pair is a key which is used to
2763 identify the object, and the @code{cdr} is the object's value.
2765 A @dfn{hash table} also permits you to index objects with
2766 arbitrary keys, but in a way that makes looking up any one object
2767 extremely fast. A well-designed hash system makes hash table
2768 lookups almost as fast as conventional array or vector references.
2770 Alists are popular among Lisp programmers because they use only
2771 the language's primitive operations (lists, @dfn{car}, @dfn{cdr}
2772 and the equality primitives). No changes to the language core are
2773 necessary. Therefore, with Scheme's built-in list manipulation
2774 facilities, it is very convenient to handle data stored in an
2775 association list. Also, alists are highly portable and can be
2776 easily implemented on even the most minimal Lisp systems.
2778 However, alists are inefficient, especially for storing large
2779 quantities of data. Because we want Guile to be useful for large
2780 software systems as well as small ones, Guile provides a rich set
2781 of tools for using either association lists or hash tables.
2783 @node Association Lists
2784 @subsection Association Lists
2785 @tpindex Association Lists
2787 @cindex association List
2791 An association list is a conventional data structure that is often used
2792 to implement simple key-value databases. It consists of a list of
2793 entries in which each entry is a pair. The @dfn{key} of each entry is
2794 the @code{car} of the pair and the @dfn{value} of each entry is the
2798 ASSOCIATION LIST ::= '( (KEY1 . VALUE1)
2806 Association lists are also known, for short, as @dfn{alists}.
2808 The structure of an association list is just one example of the infinite
2809 number of possible structures that can be built using pairs and lists.
2810 As such, the keys and values in an association list can be manipulated
2811 using the general list structure procedures @code{cons}, @code{car},
2812 @code{cdr}, @code{set-car!}, @code{set-cdr!} and so on. However,
2813 because association lists are so useful, Guile also provides specific
2814 procedures for manipulating them.
2817 * Alist Key Equality::
2818 * Adding or Setting Alist Entries::
2819 * Retrieving Alist Entries::
2820 * Removing Alist Entries::
2821 * Sloppy Alist Functions::
2825 @node Alist Key Equality
2826 @subsubsection Alist Key Equality
2828 All of Guile's dedicated association list procedures, apart from
2829 @code{acons}, come in three flavours, depending on the level of equality
2830 that is required to decide whether an existing key in the association
2831 list is the same as the key that the procedure call uses to identify the
2836 Procedures with @dfn{assq} in their name use @code{eq?} to determine key
2840 Procedures with @dfn{assv} in their name use @code{eqv?} to determine
2844 Procedures with @dfn{assoc} in their name use @code{equal?} to
2845 determine key equality.
2848 @code{acons} is an exception because it is used to build association
2849 lists which do not require their entries' keys to be unique.
2851 @node Adding or Setting Alist Entries
2852 @subsubsection Adding or Setting Alist Entries
2854 @code{acons} adds a new entry to an association list and returns the
2855 combined association list. The combined alist is formed by consing the
2856 new entry onto the head of the alist specified in the @code{acons}
2857 procedure call. So the specified alist is not modified, but its
2858 contents become shared with the tail of the combined alist that
2859 @code{acons} returns.
2861 In the most common usage of @code{acons}, a variable holding the
2862 original association list is updated with the combined alist:
2865 (set! address-list (acons name address address-list))
2868 In such cases, it doesn't matter that the old and new values of
2869 @code{address-list} share some of their contents, since the old value is
2870 usually no longer independently accessible.
2872 Note that @code{acons} adds the specified new entry regardless of
2873 whether the alist may already contain entries with keys that are, in
2874 some sense, the same as that of the new entry. Thus @code{acons} is
2875 ideal for building alists where there is no concept of key uniqueness.
2878 (set! task-list (acons 3 "pay gas bill" '()))
2881 ((3 . "pay gas bill"))
2883 (set! task-list (acons 3 "tidy bedroom" task-list))
2886 ((3 . "tidy bedroom") (3 . "pay gas bill"))
2889 @code{assq-set!}, @code{assv-set!} and @code{assoc-set!} are used to add
2890 or replace an entry in an association list where there @emph{is} a
2891 concept of key uniqueness. If the specified association list already
2892 contains an entry whose key is the same as that specified in the
2893 procedure call, the existing entry is replaced by the new one.
2894 Otherwise, the new entry is consed onto the head of the old association
2895 list to create the combined alist. In all cases, these procedures
2896 return the combined alist.
2898 @code{assq-set!} and friends @emph{may} destructively modify the
2899 structure of the old association list in such a way that an existing
2900 variable is correctly updated without having to @code{set!} it to the
2906 (("mary" . "34 Elm Road") ("james" . "16 Bow Street"))
2908 (assoc-set! address-list "james" "1a London Road")
2910 (("mary" . "34 Elm Road") ("james" . "1a London Road"))
2914 (("mary" . "34 Elm Road") ("james" . "1a London Road"))
2920 (assoc-set! address-list "bob" "11 Newington Avenue")
2922 (("bob" . "11 Newington Avenue") ("mary" . "34 Elm Road")
2923 ("james" . "1a London Road"))
2927 (("mary" . "34 Elm Road") ("james" . "1a London Road"))
2930 The only safe way to update an association list variable when adding or
2931 replacing an entry like this is to @code{set!} the variable to the
2936 (assoc-set! address-list "bob" "11 Newington Avenue"))
2939 (("bob" . "11 Newington Avenue") ("mary" . "34 Elm Road")
2940 ("james" . "1a London Road"))
2943 Because of this slight inconvenience, you may find it more convenient to
2944 use hash tables to store dictionary data. If your application will not
2945 be modifying the contents of an alist very often, this may not make much
2948 If you need to keep the old value of an association list in a form
2949 independent from the list that results from modification by
2950 @code{acons}, @code{assq-set!}, @code{assv-set!} or @code{assoc-set!},
2951 use @code{list-copy} to copy the old association list before modifying
2954 @deffn {Scheme Procedure} acons key value alist
2955 @deffnx {C Function} scm_acons (key, value, alist)
2956 Add a new key-value pair to @var{alist}. A new pair is
2957 created whose car is @var{key} and whose cdr is @var{value}, and the
2958 pair is consed onto @var{alist}, and the new list is returned. This
2959 function is @emph{not} destructive; @var{alist} is not modified.
2962 @deffn {Scheme Procedure} assq-set! alist key val
2963 @deffnx {Scheme Procedure} assv-set! alist key value
2964 @deffnx {Scheme Procedure} assoc-set! alist key value
2965 @deffnx {C Function} scm_assq_set_x (alist, key, val)
2966 @deffnx {C Function} scm_assv_set_x (alist, key, val)
2967 @deffnx {C Function} scm_assoc_set_x (alist, key, val)
2968 Reassociate @var{key} in @var{alist} with @var{value}: find any existing
2969 @var{alist} entry for @var{key} and associate it with the new
2970 @var{value}. If @var{alist} does not contain an entry for @var{key},
2971 add a new one. Return the (possibly new) alist.
2973 These functions do not attempt to verify the structure of @var{alist},
2974 and so may cause unusual results if passed an object that is not an
2978 @node Retrieving Alist Entries
2979 @subsubsection Retrieving Alist Entries
2984 @code{assq}, @code{assv} and @code{assoc} find the entry in an alist
2985 for a given key, and return the @code{(@var{key} . @var{value})} pair.
2986 @code{assq-ref}, @code{assv-ref} and @code{assoc-ref} do a similar
2987 lookup, but return just the @var{value}.
2989 @deffn {Scheme Procedure} assq key alist
2990 @deffnx {Scheme Procedure} assv key alist
2991 @deffnx {Scheme Procedure} assoc key alist
2992 @deffnx {C Function} scm_assq (key, alist)
2993 @deffnx {C Function} scm_assv (key, alist)
2994 @deffnx {C Function} scm_assoc (key, alist)
2995 Return the first entry in @var{alist} with the given @var{key}. The
2996 return is the pair @code{(KEY . VALUE)} from @var{alist}. If there's
2997 no matching entry the return is @code{#f}.
2999 @code{assq} compares keys with @code{eq?}, @code{assv} uses
3000 @code{eqv?} and @code{assoc} uses @code{equal?}. See also SRFI-1
3001 which has an extended @code{assoc} (@ref{SRFI-1 Association Lists}).
3004 @deffn {Scheme Procedure} assq-ref alist key
3005 @deffnx {Scheme Procedure} assv-ref alist key
3006 @deffnx {Scheme Procedure} assoc-ref alist key
3007 @deffnx {C Function} scm_assq_ref (alist, key)
3008 @deffnx {C Function} scm_assv_ref (alist, key)
3009 @deffnx {C Function} scm_assoc_ref (alist, key)
3010 Return the value from the first entry in @var{alist} with the given
3011 @var{key}, or @code{#f} if there's no such entry.
3013 @code{assq-ref} compares keys with @code{eq?}, @code{assv-ref} uses
3014 @code{eqv?} and @code{assoc-ref} uses @code{equal?}.
3016 Notice these functions have the @var{key} argument last, like other
3017 @code{-ref} functions, but this is opposite to what @code{assq}
3020 When the return is @code{#f} it can be either @var{key} not found, or
3021 an entry which happens to have value @code{#f} in the @code{cdr}. Use
3022 @code{assq} etc above if you need to differentiate these cases.
3026 @node Removing Alist Entries
3027 @subsubsection Removing Alist Entries
3029 To remove the element from an association list whose key matches a
3030 specified key, use @code{assq-remove!}, @code{assv-remove!} or
3031 @code{assoc-remove!} (depending, as usual, on the level of equality
3032 required between the key that you specify and the keys in the
3035 As with @code{assq-set!} and friends, the specified alist may or may not
3036 be modified destructively, and the only safe way to update a variable
3037 containing the alist is to @code{set!} it to the value that
3038 @code{assq-remove!} and friends return.
3043 (("bob" . "11 Newington Avenue") ("mary" . "34 Elm Road")
3044 ("james" . "1a London Road"))
3046 (set! address-list (assoc-remove! address-list "mary"))
3049 (("bob" . "11 Newington Avenue") ("james" . "1a London Road"))
3052 Note that, when @code{assq/v/oc-remove!} is used to modify an
3053 association list that has been constructed only using the corresponding
3054 @code{assq/v/oc-set!}, there can be at most one matching entry in the
3055 alist, so the question of multiple entries being removed in one go does
3056 not arise. If @code{assq/v/oc-remove!} is applied to an association
3057 list that has been constructed using @code{acons}, or an
3058 @code{assq/v/oc-set!} with a different level of equality, or any mixture
3059 of these, it removes only the first matching entry from the alist, even
3060 if the alist might contain further matching entries. For example:
3063 (define address-list '())
3064 (set! address-list (assq-set! address-list "mary" "11 Elm Street"))
3065 (set! address-list (assq-set! address-list "mary" "57 Pine Drive"))
3068 (("mary" . "57 Pine Drive") ("mary" . "11 Elm Street"))
3070 (set! address-list (assoc-remove! address-list "mary"))
3073 (("mary" . "11 Elm Street"))
3076 In this example, the two instances of the string "mary" are not the same
3077 when compared using @code{eq?}, so the two @code{assq-set!} calls add
3078 two distinct entries to @code{address-list}. When compared using
3079 @code{equal?}, both "mary"s in @code{address-list} are the same as the
3080 "mary" in the @code{assoc-remove!} call, but @code{assoc-remove!} stops
3081 after removing the first matching entry that it finds, and so one of the
3082 "mary" entries is left in place.
3084 @deffn {Scheme Procedure} assq-remove! alist key
3085 @deffnx {Scheme Procedure} assv-remove! alist key
3086 @deffnx {Scheme Procedure} assoc-remove! alist key
3087 @deffnx {C Function} scm_assq_remove_x (alist, key)
3088 @deffnx {C Function} scm_assv_remove_x (alist, key)
3089 @deffnx {C Function} scm_assoc_remove_x (alist, key)
3090 Delete the first entry in @var{alist} associated with @var{key}, and return
3091 the resulting alist.
3094 @node Sloppy Alist Functions
3095 @subsubsection Sloppy Alist Functions
3097 @code{sloppy-assq}, @code{sloppy-assv} and @code{sloppy-assoc} behave
3098 like the corresponding non-@code{sloppy-} procedures, except that they
3099 return @code{#f} when the specified association list is not well-formed,
3100 where the non-@code{sloppy-} versions would signal an error.
3102 Specifically, there are two conditions for which the non-@code{sloppy-}
3103 procedures signal an error, which the @code{sloppy-} procedures handle
3104 instead by returning @code{#f}. Firstly, if the specified alist as a
3105 whole is not a proper list:
3108 (assoc "mary" '((1 . 2) ("key" . "door") . "open sesame"))
3110 ERROR: In procedure assoc in expression (assoc "mary" (quote #)):
3111 ERROR: Wrong type argument in position 2 (expecting
3112 association list): ((1 . 2) ("key" . "door") . "open sesame")
3114 (sloppy-assoc "mary" '((1 . 2) ("key" . "door") . "open sesame"))
3120 Secondly, if one of the entries in the specified alist is not a pair:
3123 (assoc 2 '((1 . 1) 2 (3 . 9)))
3125 ERROR: In procedure assoc in expression (assoc 2 (quote #)):
3126 ERROR: Wrong type argument in position 2 (expecting
3127 association list): ((1 . 1) 2 (3 . 9))
3129 (sloppy-assoc 2 '((1 . 1) 2 (3 . 9)))
3134 Unless you are explicitly working with badly formed association lists,
3135 it is much safer to use the non-@code{sloppy-} procedures, because they
3136 help to highlight coding and data errors that the @code{sloppy-}
3137 versions would silently cover up.
3139 @deffn {Scheme Procedure} sloppy-assq key alist
3140 @deffnx {C Function} scm_sloppy_assq (key, alist)
3141 Behaves like @code{assq} but does not do any error checking.
3142 Recommended only for use in Guile internals.
3145 @deffn {Scheme Procedure} sloppy-assv key alist
3146 @deffnx {C Function} scm_sloppy_assv (key, alist)
3147 Behaves like @code{assv} but does not do any error checking.
3148 Recommended only for use in Guile internals.
3151 @deffn {Scheme Procedure} sloppy-assoc key alist
3152 @deffnx {C Function} scm_sloppy_assoc (key, alist)
3153 Behaves like @code{assoc} but does not do any error checking.
3154 Recommended only for use in Guile internals.
3158 @subsubsection Alist Example
3160 Here is a longer example of how alists may be used in practice.
3163 (define capitals '(("New York" . "Albany")
3164 ("Oregon" . "Salem")
3165 ("Florida" . "Miami")))
3167 ;; What's the capital of Oregon?
3168 (assoc "Oregon" capitals) @result{} ("Oregon" . "Salem")
3169 (assoc-ref capitals "Oregon") @result{} "Salem"
3171 ;; We left out South Dakota.
3173 (assoc-set! capitals "South Dakota" "Pierre"))
3175 @result{} (("South Dakota" . "Pierre")
3176 ("New York" . "Albany")
3177 ("Oregon" . "Salem")
3178 ("Florida" . "Miami"))
3180 ;; And we got Florida wrong.
3182 (assoc-set! capitals "Florida" "Tallahassee"))
3184 @result{} (("South Dakota" . "Pierre")
3185 ("New York" . "Albany")
3186 ("Oregon" . "Salem")
3187 ("Florida" . "Tallahassee"))
3189 ;; After Oregon secedes, we can remove it.
3191 (assoc-remove! capitals "Oregon"))
3193 @result{} (("South Dakota" . "Pierre")
3194 ("New York" . "Albany")
3195 ("Florida" . "Tallahassee"))
3199 @subsection VList-Based Hash Lists or ``VHashes''
3201 @cindex VList-based hash lists
3204 The @code{(ice-9 vlist)} module provides an implementation of @dfn{VList-based
3205 hash lists} (@pxref{VLists}). VList-based hash lists, or @dfn{vhashes}, are an
3206 immutable dictionary type similar to association lists that maps @dfn{keys} to
3207 @dfn{values}. However, unlike association lists, accessing a value given its
3208 key is typically a constant-time operation.
3210 The VHash programming interface of @code{(ice-9 vlist)} is mostly the same as
3211 that of association lists found in SRFI-1, with procedure names prefixed by
3212 @code{vhash-} instead of @code{vlist-} (@pxref{SRFI-1 Association Lists}).
3214 In addition, vhashes can be manipulated using VList operations:
3217 (vlist-head (vhash-consq 'a 1 vlist-null))
3220 (define vh1 (vhash-consq 'b 2 (vhash-consq 'a 1 vlist-null)))
3221 (define vh2 (vhash-consq 'c 3 (vlist-tail vh1)))
3230 @result{} ((c . 3) (a . 1))
3233 However, keep in mind that procedures that construct new VLists
3234 (@code{vlist-map}, @code{vlist-filter}, etc.) return raw VLists, not vhashes:
3237 (define vh (alist->vhash '((a . 1) (b . 2) (c . 3)) hashq))
3242 ;; This will create a raw vlist.
3243 (vlist-filter (lambda (key+value) (odd? (cdr key+value))) vh))
3245 @result{} ERROR: Wrong type argument in position 2
3248 @result{} ((a . 1) (c . 3))
3251 @deffn {Scheme Procedure} vhash? obj
3252 Return true if @var{obj} is a vhash.
3255 @deffn {Scheme Procedure} vhash-cons key value vhash [hash-proc]
3256 @deffnx {Scheme Procedure} vhash-consq key value vhash
3257 @deffnx {Scheme Procedure} vhash-consv key value vhash
3258 Return a new hash list based on @var{vhash} where @var{key} is associated with
3259 @var{value}, using @var{hash-proc} to compute the hash of @var{key}.
3260 @var{vhash} must be either @code{vlist-null} or a vhash returned by a previous
3261 call to @code{vhash-cons}. @var{hash-proc} defaults to @code{hash} (@pxref{Hash
3262 Table Reference, @code{hash} procedure}). With @code{vhash-consq}, the
3263 @code{hashq} hash function is used; with @code{vhash-consv} the @code{hashv}
3264 hash function is used.
3266 All @code{vhash-cons} calls made to construct a vhash should use the same
3267 @var{hash-proc}. Failing to do that, the result is undefined.
3270 @deffn {Scheme Procedure} vhash-assoc key vhash [equal? [hash-proc]]
3271 @deffnx {Scheme Procedure} vhash-assq key vhash
3272 @deffnx {Scheme Procedure} vhash-assv key vhash
3273 Return the first key/value pair from @var{vhash} whose key is equal to @var{key}
3274 according to the @var{equal?} equality predicate (which defaults to
3275 @code{equal?}), and using @var{hash-proc} (which defaults to @code{hash}) to
3276 compute the hash of @var{key}. The second form uses @code{eq?} as the equality
3277 predicate and @code{hashq} as the hash function; the last form uses @code{eqv?}
3280 Note that it is important to consistently use the same hash function for
3281 @var{hash-proc} as was passed to @code{vhash-cons}. Failing to do that, the
3282 result is unpredictable.
3285 @deffn {Scheme Procedure} vhash-delete key vhash [equal? [hash-proc]]
3286 @deffnx {Scheme Procedure} vhash-delq key vhash
3287 @deffnx {Scheme Procedure} vhash-delv key vhash
3288 Remove all associations from @var{vhash} with @var{key}, comparing keys with
3289 @var{equal?} (which defaults to @code{equal?}), and computing the hash of
3290 @var{key} using @var{hash-proc} (which defaults to @code{hash}). The second
3291 form uses @code{eq?} as the equality predicate and @code{hashq} as the hash
3292 function; the last one uses @code{eqv?} and @code{hashv}.
3294 Again the choice of @var{hash-proc} must be consistent with previous calls to
3298 @deffn {Scheme Procedure} vhash-fold proc vhash
3299 Fold over the key/pair elements of @var{vhash}. For each pair call @var{proc}
3300 as @code{(@var{proc} key value result)}.
3303 @deffn {Scheme Procedure} alist->vhash alist [hash-proc]
3304 Return the vhash corresponding to @var{alist}, an association list, using
3305 @var{hash-proc} to compute key hashes. When omitted, @var{hash-proc} defaults
3311 @subsection Hash Tables
3312 @tpindex Hash Tables
3314 Hash tables are dictionaries which offer similar functionality as
3315 association lists: They provide a mapping from keys to values. The
3316 difference is that association lists need time linear in the size of
3317 elements when searching for entries, whereas hash tables can normally
3318 search in constant time. The drawback is that hash tables require a
3319 little bit more memory, and that you can not use the normal list
3320 procedures (@pxref{Lists}) for working with them.
3322 Guile provides two types of hashtables. One is an abstract data type
3323 that can only be manipulated with the functions in this section. The
3324 other type is concrete: it uses a normal vector with alists as
3325 elements. The advantage of the abstract hash tables is that they will
3326 be automatically resized when they become too full or too empty.
3329 * Hash Table Examples:: Demonstration of hash table usage.
3330 * Hash Table Reference:: Hash table procedure descriptions.
3334 @node Hash Table Examples
3335 @subsubsection Hash Table Examples
3337 For demonstration purposes, this section gives a few usage examples of
3338 some hash table procedures, together with some explanation what they do.
3340 First we start by creating a new hash table with 31 slots, and
3341 populate it with two key/value pairs.
3344 (define h (make-hash-table 31))
3346 ;; This is an opaque object
3351 ;; We can also use a vector of alists.
3352 (define h (make-vector 7 '()))
3356 #(() () () () () () ())
3358 ;; Inserting into a hash table can be done with hashq-set!
3359 (hashq-set! h 'foo "bar")
3363 (hashq-set! h 'braz "zonk")
3367 ;; Or with hash-create-handle!
3368 (hashq-create-handle! h 'frob #f)
3372 ;; The vector now contains three elements in the alists and the frob
3373 ;; entry is at index (hashq 'frob).
3376 #(((braz . "zonk")) ((foo . "bar")) () () () () ((frob . #f)))
3384 You can get the value for a given key with the procedure
3385 @code{hashq-ref}, but the problem with this procedure is that you
3386 cannot reliably determine whether a key does exists in the table. The
3387 reason is that the procedure returns @code{#f} if the key is not in
3388 the table, but it will return the same value if the key is in the
3389 table and just happens to have the value @code{#f}, as you can see in
3390 the following examples.
3401 (hashq-ref h 'not-there)
3406 Better is to use the procedure @code{hashq-get-handle}, which makes a
3407 distinction between the two cases. Just like @code{assq}, this
3408 procedure returns a key/value-pair on success, and @code{#f} if the
3412 (hashq-get-handle h 'foo)
3416 (hashq-get-handle h 'not-there)
3421 There is no procedure for calculating the number of key/value-pairs in
3422 a hash table, but @code{hash-fold} can be used for doing exactly that.
3425 (hash-fold (lambda (key value seed) (+ 1 seed)) 0 h)
3430 @node Hash Table Reference
3431 @subsubsection Hash Table Reference
3433 @c FIXME: Describe in broad terms what happens for resizing, and what
3434 @c the initial size means for this.
3436 Like the association list functions, the hash table functions come in
3437 several varieties, according to the equality test used for the keys.
3438 Plain @code{hash-} functions use @code{equal?}, @code{hashq-}
3439 functions use @code{eq?}, @code{hashv-} functions use @code{eqv?}, and
3440 the @code{hashx-} functions use an application supplied test.
3442 A single @code{make-hash-table} creates a hash table suitable for use
3443 with any set of functions, but it's imperative that just one set is
3444 then used consistently, or results will be unpredictable.
3446 Hash tables are implemented as a vector indexed by a hash value formed
3447 from the key, with an association list of key/value pairs for each
3448 bucket in case distinct keys hash together. Direct access to the
3449 pairs in those lists is provided by the @code{-handle-} functions.
3450 The abstract kind of hash tables hide the vector in an opaque object
3451 that represents the hash table, while for the concrete kind the vector
3452 @emph{is} the hashtable.
3454 When the number of table entries in an abstract hash table goes above
3455 a threshold, the vector is made larger and the entries are rehashed,
3456 to prevent the bucket lists from becoming too long and slowing down
3457 accesses. When the number of entries goes below a threshold, the
3458 vector is shrunk to save space.
3460 A abstract hash table is created with @code{make-hash-table}. To
3461 create a vector that is suitable as a hash table, use
3462 @code{(make-vector @var{size} '())}, for example.
3464 For the @code{hashx-} ``extended'' routines, an application supplies a
3465 @var{hash} function producing an integer index like @code{hashq} etc
3466 below, and an @var{assoc} alist search function like @code{assq} etc
3467 (@pxref{Retrieving Alist Entries}). Here's an example of such
3468 functions implementing case-insensitive hashing of string keys,
3471 (use-modules (srfi srfi-1)
3474 (define (my-hash str size)
3475 (remainder (string-hash-ci str) size))
3476 (define (my-assoc str alist)
3477 (find (lambda (pair) (string-ci=? str (car pair))) alist))
3479 (define my-table (make-hash-table))
3480 (hashx-set! my-hash my-assoc my-table "foo" 123)
3482 (hashx-ref my-hash my-assoc my-table "FOO")
3486 In a @code{hashx-} @var{hash} function the aim is to spread keys
3487 across the vector, so bucket lists don't become long. But the actual
3488 values are arbitrary as long as they're in the range 0 to
3489 @math{@var{size}-1}. Helpful functions for forming a hash value, in
3490 addition to @code{hashq} etc below, include @code{symbol-hash}
3491 (@pxref{Symbol Keys}), @code{string-hash} and @code{string-hash-ci}
3492 (@pxref{String Comparison}), and @code{char-set-hash}
3493 (@pxref{Character Set Predicates/Comparison}).
3496 @deffn {Scheme Procedure} make-hash-table [size]
3497 Create a new abstract hash table object, with an optional minimum
3500 When @var{size} is given, the table vector will still grow and shrink
3501 automatically, as described above, but with @var{size} as a minimum.
3502 If an application knows roughly how many entries the table will hold
3503 then it can use @var{size} to avoid rehashing when initial entries are
3507 @deffn {Scheme Procedure} hash-table? obj
3508 @deffnx {C Function} scm_hash_table_p (obj)
3509 Return @code{#t} if @var{obj} is a abstract hash table object.
3512 @deffn {Scheme Procedure} hash-clear! table
3513 @deffnx {C Function} scm_hash_clear_x (table)
3514 Remove all items from @var{table} (without triggering a resize).
3517 @deffn {Scheme Procedure} hash-ref table key [dflt]
3518 @deffnx {Scheme Procedure} hashq-ref table key [dflt]
3519 @deffnx {Scheme Procedure} hashv-ref table key [dflt]
3520 @deffnx {Scheme Procedure} hashx-ref hash assoc table key [dflt]
3521 @deffnx {C Function} scm_hash_ref (table, key, dflt)
3522 @deffnx {C Function} scm_hashq_ref (table, key, dflt)
3523 @deffnx {C Function} scm_hashv_ref (table, key, dflt)
3524 @deffnx {C Function} scm_hashx_ref (hash, assoc, table, key, dflt)
3525 Lookup @var{key} in the given hash @var{table}, and return the
3526 associated value. If @var{key} is not found, return @var{dflt}, or
3527 @code{#f} if @var{dflt} is not given.
3530 @deffn {Scheme Procedure} hash-set! table key val
3531 @deffnx {Scheme Procedure} hashq-set! table key val
3532 @deffnx {Scheme Procedure} hashv-set! table key val
3533 @deffnx {Scheme Procedure} hashx-set! hash assoc table key val
3534 @deffnx {C Function} scm_hash_set_x (table, key, val)
3535 @deffnx {C Function} scm_hashq_set_x (table, key, val)
3536 @deffnx {C Function} scm_hashv_set_x (table, key, val)
3537 @deffnx {C Function} scm_hashx_set_x (hash, assoc, table, key, val)
3538 Associate @var{val} with @var{key} in the given hash @var{table}. If
3539 @var{key} is already present then it's associated value is changed.
3540 If it's not present then a new entry is created.
3543 @deffn {Scheme Procedure} hash-remove! table key
3544 @deffnx {Scheme Procedure} hashq-remove! table key
3545 @deffnx {Scheme Procedure} hashv-remove! table key
3546 @deffnx {Scheme Procedure} hashx-remove! hash assoc table key
3547 @deffnx {C Function} scm_hash_remove_x (table, key)
3548 @deffnx {C Function} scm_hashq_remove_x (table, key)
3549 @deffnx {C Function} scm_hashv_remove_x (table, key)
3550 @deffnx {C Function} scm_hashx_remove_x (hash, assoc, table, key)
3551 Remove any association for @var{key} in the given hash @var{table}.
3552 If @var{key} is not in @var{table} then nothing is done.
3555 @deffn {Scheme Procedure} hash key size
3556 @deffnx {Scheme Procedure} hashq key size
3557 @deffnx {Scheme Procedure} hashv key size
3558 @deffnx {C Function} scm_hash (key, size)
3559 @deffnx {C Function} scm_hashq (key, size)
3560 @deffnx {C Function} scm_hashv (key, size)
3561 Return a hash value for @var{key}. This is a number in the range
3562 @math{0} to @math{@var{size}-1}, which is suitable for use in a hash
3563 table of the given @var{size}.
3565 Note that @code{hashq} and @code{hashv} may use internal addresses of
3566 objects, so if an object is garbage collected and re-created it can
3567 have a different hash value, even when the two are notionally
3568 @code{eq?}. For instance with symbols,
3571 (hashq 'something 123) @result{} 19
3573 (hashq 'something 123) @result{} 62
3576 In normal use this is not a problem, since an object entered into a
3577 hash table won't be garbage collected until removed. It's only if
3578 hashing calculations are somehow separated from normal references that
3579 its lifetime needs to be considered.
3582 @deffn {Scheme Procedure} hash-get-handle table key
3583 @deffnx {Scheme Procedure} hashq-get-handle table key
3584 @deffnx {Scheme Procedure} hashv-get-handle table key
3585 @deffnx {Scheme Procedure} hashx-get-handle hash assoc table key
3586 @deffnx {C Function} scm_hash_get_handle (table, key)
3587 @deffnx {C Function} scm_hashq_get_handle (table, key)
3588 @deffnx {C Function} scm_hashv_get_handle (table, key)
3589 @deffnx {C Function} scm_hashx_get_handle (hash, assoc, table, key)
3590 Return the @code{(@var{key} . @var{value})} pair for @var{key} in the
3591 given hash @var{table}, or @code{#f} if @var{key} is not in
3595 @deffn {Scheme Procedure} hash-create-handle! table key init
3596 @deffnx {Scheme Procedure} hashq-create-handle! table key init
3597 @deffnx {Scheme Procedure} hashv-create-handle! table key init
3598 @deffnx {Scheme Procedure} hashx-create-handle! hash assoc table key init
3599 @deffnx {C Function} scm_hash_create_handle_x (table, key, init)
3600 @deffnx {C Function} scm_hashq_create_handle_x (table, key, init)
3601 @deffnx {C Function} scm_hashv_create_handle_x (table, key, init)
3602 @deffnx {C Function} scm_hashx_create_handle_x (hash, assoc, table, key, init)
3603 Return the @code{(@var{key} . @var{value})} pair for @var{key} in the
3604 given hash @var{table}. If @var{key} is not in @var{table} then
3605 create an entry for it with @var{init} as the value, and return that
3609 @deffn {Scheme Procedure} hash-map->list proc table
3610 @deffnx {Scheme Procedure} hash-for-each proc table
3611 @deffnx {C Function} scm_hash_map_to_list (proc, table)
3612 @deffnx {C Function} scm_hash_for_each (proc, table)
3613 Apply @var{proc} to the entries in the given hash @var{table}. Each
3614 call is @code{(@var{proc} @var{key} @var{value})}. @code{hash-map->list}
3615 returns a list of the results from these calls, @code{hash-for-each}
3616 discards the results and returns an unspecified value.
3618 Calls are made over the table entries in an unspecified order, and for
3619 @code{hash-map->list} the order of the values in the returned list is
3620 unspecified. Results will be unpredictable if @var{table} is modified
3623 For example the following returns a new alist comprising all the
3624 entries from @code{mytable}, in no particular order.
3627 (hash-map->list cons mytable)
3631 @deffn {Scheme Procedure} hash-for-each-handle proc table
3632 @deffnx {C Function} scm_hash_for_each_handle (proc, table)
3633 Apply @var{proc} to the entries in the given hash @var{table}. Each
3634 call is @code{(@var{proc} @var{handle})}, where @var{handle} is a
3635 @code{(@var{key} . @var{value})} pair. Return an unspecified value.
3637 @code{hash-for-each-handle} differs from @code{hash-for-each} only in
3638 the argument list of @var{proc}.
3641 @deffn {Scheme Procedure} hash-fold proc init table
3642 @deffnx {C Function} scm_hash_fold (proc, init, table)
3643 Accumulate a result by applying @var{proc} to the elements of the
3644 given hash @var{table}. Each call is @code{(@var{proc} @var{key}
3645 @var{value} @var{prior-result})}, where @var{key} and @var{value} are
3646 from the @var{table} and @var{prior-result} is the return from the
3647 previous @var{proc} call. For the first call, @var{prior-result} is
3648 the given @var{init} value.
3650 Calls are made over the table entries in an unspecified order.
3651 Results will be unpredictable if @var{table} is modified while
3652 @code{hash-fold} is running.
3654 For example, the following returns a count of how many keys in
3655 @code{mytable} are strings.
3658 (hash-fold (lambda (key value prior)
3659 (if (string? key) (1+ prior) prior))
3666 @c TeX-master: "guile.texi"