1 /* Copyright (C) 1995, 1996, 1997, 1998, 2000, 2001, 2003, 2004,
2 * 2006, 2009, 2011, 2013 Free Software Foundation, Inc.
4 * This library is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU Lesser General Public License
6 * as published by the Free Software Foundation; either version 3 of
7 * the License, or (at your option) any later version.
9 * This library is distributed in the hope that it will be useful, but
10 * WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
12 * Lesser General Public License for more details.
14 * You should have received a copy of the GNU Lesser General Public
15 * License along with this library; if not, write to the Free Software
16 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
28 #include "libguile/_scm.h"
29 #include "libguile/chars.h"
30 #include "libguile/eval.h"
31 #include "libguile/hash.h"
32 #include "libguile/smob.h"
33 #include "libguile/variable.h"
34 #include "libguile/alist.h"
35 #include "libguile/fluids.h"
36 #include "libguile/strings.h"
37 #include "libguile/vectors.h"
38 #include "libguile/weak-set.h"
39 #include "libguile/modules.h"
40 #include "libguile/read.h"
41 #include "libguile/srfi-13.h"
43 #include "libguile/validate.h"
44 #include "libguile/symbols.h"
46 #include "libguile/private-options.h"
58 SCM_DEFINE (scm_sys_symbols
, "%symbols", 0, 0, 0,
60 "Return the system symbol obarray.")
61 #define FUNC_NAME s_scm_sys_symbols
74 scm_i_hash_symbol (SCM obj
, unsigned long n
, void *closure
)
76 return scm_i_symbol_hash (obj
) % n
;
79 struct string_lookup_data
82 unsigned long string_hash
;
86 string_lookup_predicate_fn (SCM sym
, void *closure
)
88 struct string_lookup_data
*data
= closure
;
90 if (scm_i_symbol_hash (sym
) == data
->string_hash
91 && scm_i_symbol_length (sym
) == scm_i_string_length (data
->string
))
93 size_t n
= scm_i_symbol_length (sym
);
95 if (scm_i_symbol_ref (sym
, n
) != scm_i_string_ref (data
->string
, n
))
104 lookup_interned_symbol (SCM name
, unsigned long raw_hash
, SCM obarray
)
106 struct string_lookup_data data
;
109 data
.string_hash
= raw_hash
;
111 return scm_c_weak_set_lookup (obarray
, raw_hash
,
112 string_lookup_predicate_fn
,
116 struct latin1_lookup_data
120 unsigned long string_hash
;
124 latin1_lookup_predicate_fn (SCM sym
, void *closure
)
126 struct latin1_lookup_data
*data
= closure
;
128 return scm_i_symbol_hash (sym
) == data
->string_hash
129 && scm_i_is_narrow_symbol (sym
)
130 && scm_i_symbol_length (sym
) == data
->len
131 && strncmp (scm_i_symbol_chars (sym
), data
->str
, data
->len
) == 0;
135 lookup_interned_latin1_symbol (const char *str
, size_t len
,
136 unsigned long raw_hash
,
139 struct latin1_lookup_data data
;
143 data
.string_hash
= raw_hash
;
145 return scm_c_weak_set_lookup (obarray
, raw_hash
,
146 latin1_lookup_predicate_fn
,
150 struct utf8_lookup_data
154 unsigned long string_hash
;
158 utf8_string_equals_wide_string (const scm_t_uint8
*narrow
, size_t nlen
,
159 const scm_t_wchar
*wide
, size_t wlen
)
161 size_t byte_idx
= 0, char_idx
= 0;
163 while (byte_idx
< nlen
&& char_idx
< wlen
)
168 nbytes
= u8_mbtouc (&c
, narrow
+ byte_idx
, nlen
- byte_idx
);
171 else if (c
== 0xfffd)
174 else if (c
!= wide
[char_idx
])
181 return byte_idx
== nlen
&& char_idx
== wlen
;
185 utf8_lookup_predicate_fn (SCM sym
, void *closure
)
187 struct utf8_lookup_data
*data
= closure
;
189 if (scm_i_symbol_hash (sym
) != data
->string_hash
)
192 if (scm_i_is_narrow_symbol (sym
))
193 return (scm_i_symbol_length (sym
) == data
->len
194 && strncmp (scm_i_symbol_chars (sym
), data
->str
, data
->len
) == 0);
196 return utf8_string_equals_wide_string ((const scm_t_uint8
*) data
->str
,
198 scm_i_symbol_wide_chars (sym
),
199 scm_i_symbol_length (sym
));
203 lookup_interned_utf8_symbol (const char *str
, size_t len
,
204 unsigned long raw_hash
,
207 struct utf8_lookup_data data
;
211 data
.string_hash
= raw_hash
;
213 return scm_c_weak_set_lookup (obarray
, raw_hash
,
214 utf8_lookup_predicate_fn
,
219 symbol_lookup_predicate_fn (SCM sym
, void *closure
)
221 SCM other
= SCM_PACK_POINTER (closure
);
223 if (scm_i_symbol_hash (sym
) == scm_i_symbol_hash (other
)
224 && scm_i_symbol_length (sym
) == scm_i_symbol_length (other
))
226 if (scm_i_is_narrow_symbol (sym
))
227 return scm_i_is_narrow_symbol (other
)
228 && (strncmp (scm_i_symbol_chars (sym
),
229 scm_i_symbol_chars (other
),
230 scm_i_symbol_length (other
)) == 0);
233 (scm_string_equal_p (scm_symbol_to_string (sym
),
234 scm_symbol_to_string (other
)));
240 scm_i_str2symbol (SCM str
, SCM obarray
)
243 size_t raw_hash
= scm_i_string_hash (str
);
245 symbol
= lookup_interned_symbol (str
, raw_hash
, obarray
);
246 if (scm_is_true (symbol
))
250 /* The symbol was not found, create it. */
251 symbol
= scm_i_make_symbol (str
, 0, raw_hash
,
252 scm_cons (SCM_BOOL_F
, SCM_EOL
));
254 /* Might return a different symbol, if another one was interned at
256 return scm_c_weak_set_add_x (obarray
, raw_hash
,
257 symbol_lookup_predicate_fn
,
258 SCM_UNPACK_POINTER (symbol
), symbol
);
264 scm_i_str2uninterned_symbol (SCM str
)
266 size_t raw_hash
= scm_i_string_hash (str
);
268 return scm_i_make_symbol (str
, SCM_I_F_SYMBOL_UNINTERNED
,
269 raw_hash
, scm_cons (SCM_BOOL_F
, SCM_EOL
));
272 SCM_DEFINE (scm_symbol_p
, "symbol?", 1, 0, 0,
274 "Return @code{#t} if @var{obj} is a symbol, otherwise return\n"
276 #define FUNC_NAME s_scm_symbol_p
278 return scm_from_bool (scm_is_symbol (obj
));
282 SCM_DEFINE (scm_symbol_interned_p
, "symbol-interned?", 1, 0, 0,
284 "Return @code{#t} if @var{symbol} is interned, otherwise return\n"
286 #define FUNC_NAME s_scm_symbol_interned_p
288 SCM_VALIDATE_SYMBOL (1, symbol
);
289 return scm_from_bool (scm_i_symbol_is_interned (symbol
));
293 SCM_DEFINE (scm_make_symbol
, "make-symbol", 1, 0, 0,
295 "Return a new uninterned symbol with the name @var{name}. "
296 "The returned symbol is guaranteed to be unique and future "
297 "calls to @code{string->symbol} will not return it.")
298 #define FUNC_NAME s_scm_make_symbol
300 SCM_VALIDATE_STRING (1, name
);
301 return scm_i_str2uninterned_symbol (name
);
305 SCM_DEFINE (scm_symbol_to_string
, "symbol->string", 1, 0, 0,
307 "Return the name of @var{symbol} as a string. If the symbol was\n"
308 "part of an object returned as the value of a literal expression\n"
309 "(section @pxref{Literal expressions,,,r5rs, The Revised^5\n"
310 "Report on Scheme}) or by a call to the @code{read} procedure,\n"
311 "and its name contains alphabetic characters, then the string\n"
312 "returned will contain characters in the implementation's\n"
313 "preferred standard case---some implementations will prefer\n"
314 "upper case, others lower case. If the symbol was returned by\n"
315 "@code{string->symbol}, the case of characters in the string\n"
316 "returned will be the same as the case in the string that was\n"
317 "passed to @code{string->symbol}. It is an error to apply\n"
318 "mutation procedures like @code{string-set!} to strings returned\n"
319 "by this procedure.\n"
321 "The following examples assume that the implementation's\n"
322 "standard case is lower case:\n"
325 "(symbol->string 'flying-fish) @result{} \"flying-fish\"\n"
326 "(symbol->string 'Martin) @result{} \"martin\"\n"
328 " (string->symbol \"Malvina\")) @result{} \"Malvina\"\n"
330 #define FUNC_NAME s_scm_symbol_to_string
332 SCM_VALIDATE_SYMBOL (1, s
);
333 return scm_i_symbol_substring (s
, 0, scm_i_symbol_length (s
));
338 SCM_DEFINE (scm_string_to_symbol
, "string->symbol", 1, 0, 0,
340 "Return the symbol whose name is @var{string}. This procedure\n"
341 "can create symbols with names containing special characters or\n"
342 "letters in the non-standard case, but it is usually a bad idea\n"
343 "to create such symbols because in some implementations of\n"
344 "Scheme they cannot be read as themselves. See\n"
345 "@code{symbol->string}.\n"
347 "The following examples assume that the implementation's\n"
348 "standard case is lower case:\n"
351 "(eq? 'mISSISSIppi 'mississippi) @result{} #t\n"
352 "(string->symbol \"mISSISSIppi\") @result{} @r{the symbol with name \"mISSISSIppi\"}\n"
353 "(eq? 'bitBlt (string->symbol \"bitBlt\")) @result{} #f\n"
355 " (string->symbol (symbol->string 'JollyWog))) @result{} #t\n"
356 "(string=? \"K. Harper, M.D.\"\n"
358 " (string->symbol \"K. Harper, M.D.\"))) @result{}#t\n"
360 #define FUNC_NAME s_scm_string_to_symbol
362 SCM_VALIDATE_STRING (1, string
);
363 return scm_i_str2symbol (string
, symbols
);
367 SCM_DEFINE (scm_string_ci_to_symbol
, "string-ci->symbol", 1, 0, 0,
369 "Return the symbol whose name is @var{str}. @var{str} is\n"
370 "converted to lowercase before the conversion is done, if Guile\n"
371 "is currently reading symbols case-insensitively.")
372 #define FUNC_NAME s_scm_string_ci_to_symbol
374 return scm_string_to_symbol (SCM_CASE_INSENSITIVE_P
375 ? scm_string_downcase(str
)
380 SCM_DEFINE (scm_make_obarray
, "make-obarray", 0, 0, 0,
382 "Return a fresh obarray.")
383 #define FUNC_NAME s_scm_make_obarray
385 return scm_c_make_weak_set (0);
389 SCM_DEFINE (scm_find_symbol
, "find-symbol", 1, 1, 0,
390 (SCM string
, SCM obarray
),
391 "Return the symbol named @var{string} if it is present in\n"
392 "@var{obarray}. Return false otherwise.")
393 #define FUNC_NAME s_scm_find_symbol
395 if (SCM_UNBNDP (obarray
))
398 return lookup_interned_symbol (string
,
399 scm_i_string_hash (string
),
404 SCM_DEFINE (scm_intern
, "intern", 1, 1, 0,
405 (SCM string
, SCM obarray
),
406 "Intern @var{string} in @var{obarray}.")
407 #define FUNC_NAME s_scm_intern
409 if (SCM_UNBNDP (obarray
))
412 SCM_VALIDATE_STRING (1, string
);
413 return scm_i_str2symbol (string
, obarray
);
417 SCM_DEFINE (scm_unintern
, "unintern", 1, 1, 0,
418 (SCM symbol
, SCM obarray
),
419 "Unintern @var{symbol} from @var{obarray}.")
420 #define FUNC_NAME s_scm_unintern
422 if (SCM_UNBNDP (obarray
))
425 scm_weak_set_remove_x (obarray
, symbol
);
426 return SCM_UNSPECIFIED
;
430 SCM_DEFINE (scm_obarray_for_each
, "obarray-for-each", 1, 1, 0,
431 (SCM proc
, SCM obarray
),
433 #define FUNC_NAME s_scm_obarray_for_each
435 if (SCM_UNBNDP (obarray
))
438 scm_weak_set_for_each (proc
, obarray
);
439 return SCM_UNSPECIFIED
;
443 /* The default prefix for `gensym'd symbols. */
444 static SCM default_gensym_prefix
;
446 #define MAX_PREFIX_LENGTH 30
448 SCM_DEFINE (scm_gensym
, "gensym", 0, 1, 0,
450 "Create a new symbol with a name constructed from a prefix and\n"
451 "a counter value. The string @var{prefix} can be specified as\n"
452 "an optional argument. Default prefix is @code{ g}. The counter\n"
453 "is increased by 1 at each call. There is no provision for\n"
454 "resetting the counter.")
455 #define FUNC_NAME s_scm_gensym
457 static int gensym_counter
= 0;
461 char buf
[SCM_INTBUFLEN
];
463 if (SCM_UNBNDP (prefix
))
464 prefix
= default_gensym_prefix
;
466 /* mutex in case another thread looks and incs at the exact same moment */
467 scm_i_scm_pthread_mutex_lock (&scm_i_misc_mutex
);
468 n
= gensym_counter
++;
469 scm_i_pthread_mutex_unlock (&scm_i_misc_mutex
);
471 n_digits
= scm_iint2str (n
, 10, buf
);
472 suffix
= scm_from_latin1_stringn (buf
, n_digits
);
473 name
= scm_string_append (scm_list_2 (prefix
, suffix
));
474 return scm_string_to_symbol (name
);
478 SCM_DEFINE (scm_symbol_hash
, "symbol-hash", 1, 0, 0,
480 "Return a hash value for @var{symbol}.")
481 #define FUNC_NAME s_scm_symbol_hash
483 SCM_VALIDATE_SYMBOL (1, symbol
);
484 return scm_from_ulong (scm_i_symbol_hash (symbol
));
488 SCM_DEFINE (scm_symbol_fref
, "symbol-fref", 1, 0, 0,
490 "Return the contents of the symbol @var{s}'s @dfn{function slot}.")
491 #define FUNC_NAME s_scm_symbol_fref
493 SCM_VALIDATE_SYMBOL (1, s
);
494 return SCM_CAR (SCM_CELL_OBJECT_3 (s
));
499 SCM_DEFINE (scm_symbol_pref
, "symbol-pref", 1, 0, 0,
501 "Return the @dfn{property list} currently associated with the\n"
503 #define FUNC_NAME s_scm_symbol_pref
505 SCM_VALIDATE_SYMBOL (1, s
);
506 return SCM_CDR (SCM_CELL_OBJECT_3 (s
));
511 SCM_DEFINE (scm_symbol_fset_x
, "symbol-fset!", 2, 0, 0,
513 "Change the binding of the symbol @var{s}'s function slot.")
514 #define FUNC_NAME s_scm_symbol_fset_x
516 SCM_VALIDATE_SYMBOL (1, s
);
517 SCM_SETCAR (SCM_CELL_OBJECT_3 (s
), val
);
518 return SCM_UNSPECIFIED
;
523 SCM_DEFINE (scm_symbol_pset_x
, "symbol-pset!", 2, 0, 0,
525 "Change the binding of the symbol @var{s}'s property slot.")
526 #define FUNC_NAME s_scm_symbol_pset_x
528 SCM_VALIDATE_SYMBOL (1, s
);
529 SCM_SETCDR (SCM_CELL_OBJECT_3 (s
), val
);
530 return SCM_UNSPECIFIED
;
535 scm_from_locale_symbol (const char *sym
)
537 return scm_from_locale_symboln (sym
, -1);
541 scm_from_locale_symboln (const char *sym
, size_t len
)
543 SCM str
= scm_from_locale_stringn (sym
, len
);
544 return scm_i_str2symbol (str
, symbols
);
548 scm_take_locale_symboln (char *sym
, size_t len
)
552 str
= scm_take_locale_stringn (sym
, len
);
553 return scm_i_str2symbol (str
, symbols
);
557 scm_take_locale_symbol (char *sym
)
559 return scm_take_locale_symboln (sym
, (size_t)-1);
563 scm_from_latin1_symbol (const char *sym
)
565 return scm_from_latin1_symboln (sym
, -1);
569 scm_from_latin1_symboln (const char *sym
, size_t len
)
574 if (len
== (size_t) -1)
576 hash
= scm_i_latin1_string_hash (sym
, len
);
578 ret
= lookup_interned_latin1_symbol (sym
, len
, hash
, symbols
);
579 if (scm_is_false (ret
))
581 SCM str
= scm_from_latin1_stringn (sym
, len
);
582 ret
= scm_i_str2symbol (str
, symbols
);
589 scm_from_utf8_symbol (const char *sym
)
591 return scm_from_utf8_symboln (sym
, -1);
595 scm_from_utf8_symboln (const char *sym
, size_t len
)
600 if (len
== (size_t) -1)
602 hash
= scm_i_utf8_string_hash (sym
, len
);
604 ret
= lookup_interned_utf8_symbol (sym
, len
, hash
, symbols
);
605 if (scm_is_false (ret
))
607 SCM str
= scm_from_utf8_stringn (sym
, len
);
608 ret
= scm_i_str2symbol (str
, symbols
);
615 scm_symbols_prehistory ()
617 symbols
= scm_c_make_weak_set (5000);
624 #include "libguile/symbols.x"
626 default_gensym_prefix
= scm_from_latin1_string (" g");