1 /* Copyright (C) 1995,1996,1997,1998,2000,2001, 2003, 2004, 2006, 2009 Free Software Foundation, Inc.
3 * This library is free software; you can redistribute it and/or
4 * modify it under the terms of the GNU Lesser General Public License
5 * as published by the Free Software Foundation; either version 3 of
6 * the License, or (at your option) any later version.
8 * This library is distributed in the hope that it will be useful, but
9 * WITHOUT ANY WARRANTY; without even the implied warranty of
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
11 * Lesser General Public License for more details.
13 * You should have received a copy of the GNU Lesser General Public
14 * License along with this library; if not, write to the Free Software
15 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
25 #include "libguile/_scm.h"
26 #include "libguile/chars.h"
27 #include "libguile/eval.h"
28 #include "libguile/hash.h"
29 #include "libguile/smob.h"
30 #include "libguile/variable.h"
31 #include "libguile/alist.h"
32 #include "libguile/fluids.h"
33 #include "libguile/strings.h"
34 #include "libguile/vectors.h"
35 #include "libguile/hashtab.h"
36 #include "libguile/weaks.h"
37 #include "libguile/modules.h"
38 #include "libguile/read.h"
39 #include "libguile/srfi-13.h"
41 #include "libguile/validate.h"
42 #include "libguile/symbols.h"
44 #include "libguile/private-options.h"
56 SCM_DEFINE (scm_sys_symbols
, "%symbols", 0, 0, 0,
58 "Return the system symbol obarray.")
59 #define FUNC_NAME s_scm_sys_symbols
71 /* In order to optimize reading speed, this function breaks part of
72 * the hashtable abstraction. The optimizations are:
74 * 1. The argument string can be compared directly to symbol objects
75 * without first creating an SCM string object. (This would have
76 * been necessary if we had used the hashtable API in hashtab.h.)
78 * 2. We can use the raw hash value stored in scm_i_symbol_hash (sym)
81 * Both optimizations might be possible without breaking the
82 * abstraction if the API in hashtab.c is improved.
86 scm_i_hash_symbol (SCM obj
, unsigned long n
, void *closure
)
88 return scm_i_symbol_hash (obj
) % n
;
92 lookup_interned_symbol (const char *name
, size_t len
,
93 unsigned long raw_hash
)
95 /* Try to find the symbol in the symbols table */
96 SCM result
= SCM_BOOL_F
;
97 SCM bucket
, elt
, previous_elt
;
98 unsigned long hash
= raw_hash
% SCM_HASHTABLE_N_BUCKETS (symbols
);
100 bucket
= SCM_HASHTABLE_BUCKET (symbols
, hash
);
101 for (elt
= bucket
, previous_elt
= SCM_BOOL_F
;
103 previous_elt
= elt
, elt
= SCM_CDR (elt
))
107 pair
= SCM_CAR (elt
);
108 if (!scm_is_pair (pair
))
111 if (SCM_WEAK_PAIR_CAR_DELETED_P (pair
))
113 /* PAIR is a weak pair whose key got nullified: remove it from
115 /* FIXME: Since this is done lazily, i.e., only when a new symbol
116 is to be inserted in a bucket containing deleted symbols, the
117 number of items in the hash table may remain erroneous for some
118 time, thus precluding proper rehashing. */
119 if (previous_elt
!= SCM_BOOL_F
)
120 SCM_SETCDR (previous_elt
, SCM_CDR (elt
));
122 bucket
= SCM_CDR (elt
);
124 SCM_HASHTABLE_DECREMENT (symbols
);
128 sym
= SCM_CAR (pair
);
130 if (scm_i_symbol_hash (sym
) == raw_hash
131 && scm_i_symbol_length (sym
) == len
)
133 const char *chrs
= scm_i_symbol_chars (sym
);
139 if (name
[i
] != chrs
[i
])
151 if (SCM_HASHTABLE_N_ITEMS (symbols
) < SCM_HASHTABLE_LOWER (symbols
))
152 /* We removed many symbols in this pass so trigger a rehashing. */
153 scm_i_rehash (symbols
, scm_i_hash_symbol
, 0, "lookup_interned_symbol");
158 /* Intern SYMBOL, an uninterned symbol. */
160 intern_symbol (SCM symbol
)
165 hash
= scm_i_symbol_hash (symbol
) % SCM_HASHTABLE_N_BUCKETS (symbols
);
166 slot
= SCM_HASHTABLE_BUCKET (symbols
, hash
);
167 cell
= scm_cons (symbol
, SCM_UNDEFINED
);
169 SCM_SET_HASHTABLE_BUCKET (symbols
, hash
, scm_cons (cell
, slot
));
170 SCM_HASHTABLE_INCREMENT (symbols
);
172 if (SCM_HASHTABLE_N_ITEMS (symbols
) > SCM_HASHTABLE_UPPER (symbols
))
173 scm_i_rehash (symbols
, scm_i_hash_symbol
, 0, "intern_symbol");
177 scm_i_c_mem2symbol (const char *name
, size_t len
)
180 size_t raw_hash
= scm_string_hash ((const unsigned char *) name
, len
);
182 symbol
= lookup_interned_symbol (name
, len
, raw_hash
);
183 if (scm_is_false (symbol
))
185 /* The symbol was not found, create it. */
186 symbol
= scm_i_c_make_symbol (name
, len
, 0, raw_hash
,
187 scm_cons (SCM_BOOL_F
, SCM_EOL
));
188 intern_symbol (symbol
);
195 scm_i_mem2symbol (SCM str
)
198 const char *name
= scm_i_string_chars (str
);
199 size_t len
= scm_i_string_length (str
);
200 size_t raw_hash
= scm_string_hash ((const unsigned char *) name
, len
);
202 symbol
= lookup_interned_symbol (name
, len
, raw_hash
);
203 if (scm_is_false (symbol
))
205 /* The symbol was not found, create it. */
206 symbol
= scm_i_make_symbol (str
, 0, raw_hash
,
207 scm_cons (SCM_BOOL_F
, SCM_EOL
));
208 intern_symbol (symbol
);
216 scm_i_mem2uninterned_symbol (SCM str
)
218 const char *name
= scm_i_string_chars (str
);
219 size_t len
= scm_i_string_length (str
);
220 size_t raw_hash
= scm_string_hash ((const unsigned char *) name
, len
);
222 return scm_i_make_symbol (str
, SCM_I_F_SYMBOL_UNINTERNED
,
223 raw_hash
, scm_cons (SCM_BOOL_F
, SCM_EOL
));
226 SCM_DEFINE (scm_symbol_p
, "symbol?", 1, 0, 0,
228 "Return @code{#t} if @var{obj} is a symbol, otherwise return\n"
230 #define FUNC_NAME s_scm_symbol_p
232 return scm_from_bool (scm_is_symbol (obj
));
236 SCM_DEFINE (scm_symbol_interned_p
, "symbol-interned?", 1, 0, 0,
238 "Return @code{#t} if @var{symbol} is interned, otherwise return\n"
240 #define FUNC_NAME s_scm_symbol_interned_p
242 SCM_VALIDATE_SYMBOL (1, symbol
);
243 return scm_from_bool (scm_i_symbol_is_interned (symbol
));
247 SCM_DEFINE (scm_make_symbol
, "make-symbol", 1, 0, 0,
249 "Return a new uninterned symbol with the name @var{name}. "
250 "The returned symbol is guaranteed to be unique and future "
251 "calls to @code{string->symbol} will not return it.")
252 #define FUNC_NAME s_scm_make_symbol
254 SCM_VALIDATE_STRING (1, name
);
255 return scm_i_mem2uninterned_symbol (name
);
259 SCM_DEFINE (scm_symbol_to_string
, "symbol->string", 1, 0, 0,
261 "Return the name of @var{symbol} as a string. If the symbol was\n"
262 "part of an object returned as the value of a literal expression\n"
263 "(section @pxref{Literal expressions,,,r5rs, The Revised^5\n"
264 "Report on Scheme}) or by a call to the @code{read} procedure,\n"
265 "and its name contains alphabetic characters, then the string\n"
266 "returned will contain characters in the implementation's\n"
267 "preferred standard case---some implementations will prefer\n"
268 "upper case, others lower case. If the symbol was returned by\n"
269 "@code{string->symbol}, the case of characters in the string\n"
270 "returned will be the same as the case in the string that was\n"
271 "passed to @code{string->symbol}. It is an error to apply\n"
272 "mutation procedures like @code{string-set!} to strings returned\n"
273 "by this procedure.\n"
275 "The following examples assume that the implementation's\n"
276 "standard case is lower case:\n"
279 "(symbol->string 'flying-fish) @result{} \"flying-fish\"\n"
280 "(symbol->string 'Martin) @result{} \"martin\"\n"
282 " (string->symbol \"Malvina\")) @result{} \"Malvina\"\n"
284 #define FUNC_NAME s_scm_symbol_to_string
286 SCM_VALIDATE_SYMBOL (1, s
);
287 return scm_i_symbol_substring (s
, 0, scm_i_symbol_length (s
));
292 SCM_DEFINE (scm_string_to_symbol
, "string->symbol", 1, 0, 0,
294 "Return the symbol whose name is @var{string}. This procedure\n"
295 "can create symbols with names containing special characters or\n"
296 "letters in the non-standard case, but it is usually a bad idea\n"
297 "to create such symbols because in some implementations of\n"
298 "Scheme they cannot be read as themselves. See\n"
299 "@code{symbol->string}.\n"
301 "The following examples assume that the implementation's\n"
302 "standard case is lower case:\n"
305 "(eq? 'mISSISSIppi 'mississippi) @result{} #t\n"
306 "(string->symbol \"mISSISSIppi\") @result{} @r{the symbol with name \"mISSISSIppi\"}\n"
307 "(eq? 'bitBlt (string->symbol \"bitBlt\")) @result{} #f\n"
309 " (string->symbol (symbol->string 'JollyWog))) @result{} #t\n"
310 "(string=? \"K. Harper, M.D.\"\n"
312 " (string->symbol \"K. Harper, M.D.\"))) @result{}#t\n"
314 #define FUNC_NAME s_scm_string_to_symbol
316 SCM_VALIDATE_STRING (1, string
);
317 return scm_i_mem2symbol (string
);
321 SCM_DEFINE (scm_string_ci_to_symbol
, "string-ci->symbol", 1, 0, 0,
323 "Return the symbol whose name is @var{str}. @var{str} is\n"
324 "converted to lowercase before the conversion is done, if Guile\n"
325 "is currently reading symbols case-insensitively.")
326 #define FUNC_NAME s_scm_string_ci_to_symbol
328 return scm_string_to_symbol (SCM_CASE_INSENSITIVE_P
329 ? scm_string_downcase(str
)
334 #define MAX_PREFIX_LENGTH 30
336 SCM_DEFINE (scm_gensym
, "gensym", 0, 1, 0,
338 "Create a new symbol with a name constructed from a prefix and\n"
339 "a counter value. The string @var{prefix} can be specified as\n"
340 "an optional argument. Default prefix is @code{ g}. The counter\n"
341 "is increased by 1 at each call. There is no provision for\n"
342 "resetting the counter.")
343 #define FUNC_NAME s_scm_gensym
345 static int gensym_counter
= 0;
349 char buf
[SCM_INTBUFLEN
];
351 if (SCM_UNBNDP (prefix
))
352 prefix
= scm_from_locale_string (" g");
354 /* mutex in case another thread looks and incs at the exact same moment */
355 scm_i_scm_pthread_mutex_lock (&scm_i_misc_mutex
);
356 n
= gensym_counter
++;
357 scm_i_pthread_mutex_unlock (&scm_i_misc_mutex
);
359 n_digits
= scm_iint2str (n
, 10, buf
);
360 suffix
= scm_from_locale_stringn (buf
, n_digits
);
361 name
= scm_string_append (scm_list_2 (prefix
, suffix
));
362 return scm_string_to_symbol (name
);
366 SCM_DEFINE (scm_symbol_hash
, "symbol-hash", 1, 0, 0,
368 "Return a hash value for @var{symbol}.")
369 #define FUNC_NAME s_scm_symbol_hash
371 SCM_VALIDATE_SYMBOL (1, symbol
);
372 return scm_from_ulong (scm_i_symbol_hash (symbol
));
376 SCM_DEFINE (scm_symbol_fref
, "symbol-fref", 1, 0, 0,
378 "Return the contents of @var{symbol}'s @dfn{function slot}.")
379 #define FUNC_NAME s_scm_symbol_fref
381 SCM_VALIDATE_SYMBOL (1, s
);
382 return SCM_CAR (SCM_CELL_OBJECT_3 (s
));
387 SCM_DEFINE (scm_symbol_pref
, "symbol-pref", 1, 0, 0,
389 "Return the @dfn{property list} currently associated with @var{symbol}.")
390 #define FUNC_NAME s_scm_symbol_pref
392 SCM_VALIDATE_SYMBOL (1, s
);
393 return SCM_CDR (SCM_CELL_OBJECT_3 (s
));
398 SCM_DEFINE (scm_symbol_fset_x
, "symbol-fset!", 2, 0, 0,
400 "Change the binding of @var{symbol}'s function slot.")
401 #define FUNC_NAME s_scm_symbol_fset_x
403 SCM_VALIDATE_SYMBOL (1, s
);
404 SCM_SETCAR (SCM_CELL_OBJECT_3 (s
), val
);
405 return SCM_UNSPECIFIED
;
410 SCM_DEFINE (scm_symbol_pset_x
, "symbol-pset!", 2, 0, 0,
412 "Change the binding of @var{symbol}'s property slot.")
413 #define FUNC_NAME s_scm_symbol_pset_x
415 SCM_VALIDATE_SYMBOL (1, s
);
416 SCM_SETCDR (SCM_CELL_OBJECT_3 (s
), val
);
417 return SCM_UNSPECIFIED
;
422 scm_from_locale_symbol (const char *sym
)
424 return scm_i_c_mem2symbol (sym
, strlen (sym
));
428 scm_from_locale_symboln (const char *sym
, size_t len
)
430 return scm_i_c_mem2symbol (sym
, len
);
434 scm_take_locale_symboln (char *sym
, size_t len
)
437 unsigned long raw_hash
;
439 if (len
== (size_t)-1)
443 /* Ensure STR is null terminated. A realloc for 1 extra byte should
444 often be satisfied from the alignment padding after the block, with
445 no actual data movement. */
446 sym
= scm_realloc (sym
, len
+1);
450 raw_hash
= scm_string_hash ((unsigned char *)sym
, len
);
451 res
= lookup_interned_symbol (sym
, len
, raw_hash
);
452 if (scm_is_false (res
))
454 res
= scm_i_c_take_symbol (sym
, len
, 0, raw_hash
,
455 scm_cons (SCM_BOOL_F
, SCM_EOL
));
465 scm_take_locale_symbol (char *sym
)
467 return scm_take_locale_symboln (sym
, (size_t)-1);
471 scm_symbols_prehistory ()
473 symbols
= scm_make_weak_key_hash_table (scm_from_int (2139));
480 #include "libguile/symbols.x"