libguile/strings.c

   1 /* Copyright (C) 1995,1996,1998,2000,2001 Free Software Foundation, Inc.
   2  *
   3  * This program is free software; you can redistribute it and/or modify
   4  * it under the terms of the GNU General Public License as published by
   5  * the Free Software Foundation; either version 2, or (at your option)
   6  * any later version.
   7  *
   8  * This program is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  11  * GNU General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public License
  14  * along with this software; see the file COPYING.  If not, write to
  15  * the Free Software Foundation, Inc., 59 Temple Place, Suite 330,
  16  * Boston, MA 02111-1307 USA
  17  *
  18  * As a special exception, the Free Software Foundation gives permission
  19  * for additional uses of the text contained in its release of GUILE.
  20  *
  21  * The exception is that, if you link the GUILE library with other files
  22  * to produce an executable, this does not by itself cause the
  23  * resulting executable to be covered by the GNU General Public License.
  24  * Your use of that executable is in no way restricted on account of
  25  * linking the GUILE library code into it.
  26  *
  27  * This exception does not however invalidate any other reasons why
  28  * the executable file might be covered by the GNU General Public License.
  29  *
  30  * This exception applies only to the code released by the
  31  * Free Software Foundation under the name GUILE.  If you copy
  32  * code from other Free Software Foundation releases into a copy of
  33  * GUILE, as the General Public License permits, the exception does
  34  * not apply to the code that you add in this way.  To avoid misleading
  35  * anyone as to the status of such modified files, you must delete
  36  * this exception notice from them.
  37  *
  38  * If you write modifications of your own for GUILE, it is your choice
  39  * whether to permit this exception to apply to your modifications.
  40  * If you do not wish that, delete this exception notice.  */
  41
  42 /* Software engineering face-lift by Greg J. Badros, 11-Dec-1999,
  43    gjb@cs.washington.edu, http://www.cs.washington.edu/homes/gjb */
  44
  45 \f
  46
  47 #include <string.h>
  48
  49 #include "libguile/_scm.h"
  50 #include "libguile/chars.h"
  51 #include "libguile/root.h"
  52 #include "libguile/strings.h"
  53 #include "libguile/validate.h"
  54 \f
  55
  56 /* {Strings}
  57  */
  58
  59 SCM_DEFINE (scm_string_p, "string?", 1, 0, 0,
  60             (SCM obj),
  61             "Return @code{#t} iff @var{obj} is a string, else returns\n"
  62             "@code{#f}.")
  63 #define FUNC_NAME s_scm_string_p
  64 {
  65   return SCM_BOOL (SCM_STRINGP (obj));
  66 }
  67 #undef FUNC_NAME
  68
  69 #if SCM_DEBUG_DEPRECATED == 0
  70
  71 /* The concept of read-only strings will disappear in next release
  72  * of Guile.
  73  */
  74
  75 SCM_DEFINE (scm_read_only_string_p, "read-only-string?", 1, 0, 0,
  76             (SCM obj),
  77             "Return @code{#t} if @var{obj} is either a string or a symbol,\n"
  78             "otherwise return @code{#f}.")
  79 #define FUNC_NAME s_scm_read_only_string_p
  80 {
  81   return SCM_BOOL(SCM_ROSTRINGP (obj));
  82 }
  83 #undef FUNC_NAME
  84
  85 #endif /* DEPRECATED */
  86
  87 SCM_REGISTER_PROC (s_scm_list_to_string, "list->string", 1, 0, 0, scm_string);
  88
  89 SCM_DEFINE (scm_string, "string", 0, 0, 1,
  90             (SCM chrs),
  91             "@deffnx primitive list->string chrs\n"
  92             "Return a newly allocated string composed of the arguments,\n"
  93             "@var{chrs}.")
  94 #define FUNC_NAME s_scm_string
  95 {
  96   SCM result;
  97
  98   {
  99     long i = scm_ilength (chrs);
 100
 101     SCM_ASSERT (i >= 0, chrs, SCM_ARGn, FUNC_NAME);
 102     result = scm_allocate_string (i);
 103   }
 104
 105   {
 106     unsigned char *data = SCM_STRING_UCHARS (result);
 107
 108     while (SCM_NNULLP (chrs))
 109       {
 110         SCM elt = SCM_CAR (chrs);
 111
 112         SCM_VALIDATE_CHAR (SCM_ARGn, elt);
 113         *data++ = SCM_CHAR (elt);
 114         chrs = SCM_CDR (chrs);
 115       }
 116   }
 117   return result;
 118 }
 119 #undef FUNC_NAME
 120
 121 #if (SCM_DEBUG_DEPRECATED == 0)
 122
 123 SCM
 124 scm_makstr (size_t len, int dummy)
 125 #define FUNC_NAME "scm_makstr"
 126 {
 127   SCM s;
 128   char *mem;
 129
 130   SCM_ASSERT_RANGE (1, scm_long2num (len), len <= SCM_STRING_MAX_LENGTH);
 131
 132   mem = (char *) scm_must_malloc (len + 1, FUNC_NAME);
 133   mem[len] = 0;
 134
 135   SCM_NEWCELL (s);
 136   SCM_SET_STRING_CHARS (s, mem);
 137   SCM_SET_STRING_LENGTH (s, len);
 138
 139   return s;
 140 }
 141 #undef FUNC_NAME
 142
 143 #endif  /* SCM_DEBUG_DEPRECATED == 0 */
 144
 145 /* converts C scm_array of strings to SCM scm_list of strings. */
 146 /* If argc < 0, a null terminated scm_array is assumed. */
 147
 148 SCM
 149 scm_makfromstrs (int argc, char **argv)
 150 {
 151   int i = argc;
 152   SCM lst = SCM_EOL;
 153   if (0 > i)
 154     for (i = 0; argv[i]; i++);
 155   while (i--)
 156     lst = scm_cons (scm_makfromstr (argv[i], (size_t) strlen (argv[i]), 0), lst);
 157   return lst;
 158 }
 159
 160
 161 /* This function must only be applied to memory obtained via malloc,
 162    since the GC is going to apply `free' to it when the string is
 163    dropped.
 164
 165    Also, s[len] must be `\0', since we promise that strings are
 166    null-terminated.  Perhaps we could handle non-null-terminated
 167    strings by claiming they're shared substrings of a string we just
 168    made up.  */
 169 SCM
 170 scm_take_str (char *s, size_t len)
 171 #define FUNC_NAME "scm_take_str"
 172 {
 173   SCM answer;
 174
 175   SCM_ASSERT_RANGE (2, scm_ulong2num (len), len <= SCM_STRING_MAX_LENGTH);
 176
 177   SCM_NEWCELL (answer);
 178   SCM_SET_STRING_CHARS (answer, s);
 179   SCM_SET_STRING_LENGTH (answer, len);
 180   scm_done_malloc (len + 1);
 181
 182   return answer;
 183 }
 184 #undef FUNC_NAME
 185
 186
 187 /* `s' must be a malloc'd string.  See scm_take_str.  */
 188 SCM
 189 scm_take0str (char *s)
 190 {
 191   return scm_take_str (s, strlen (s));
 192 }
 193
 194 SCM
 195 scm_makfromstr (const char *src, size_t len, int dummy)
 196 {
 197   SCM s = scm_allocate_string (len);
 198   char *dst = SCM_STRING_CHARS (s);
 199
 200   while (len--)
 201     *dst++ = *src++;
 202   return s;
 203 }
 204
 205 SCM
 206 scm_makfrom0str (const char *src)
 207 {
 208   if (!src) return SCM_BOOL_F;
 209   return scm_makfromstr (src, (size_t) strlen (src), 0);
 210 }
 211
 212
 213 SCM
 214 scm_makfrom0str_opt (const char *src)
 215 {
 216   return scm_makfrom0str (src);
 217 }
 218
 219
 220 SCM
 221 scm_allocate_string (size_t len)
 222 #define FUNC_NAME "scm_allocate_string"
 223 {
 224   char *mem;
 225   SCM s;
 226
 227   SCM_ASSERT_RANGE (1, scm_long2num (len), len <= SCM_STRING_MAX_LENGTH);
 228
 229   mem = (char *) scm_must_malloc (len + 1, FUNC_NAME);
 230   mem[len] = 0;
 231
 232   SCM_NEWCELL (s);
 233   SCM_SET_STRING_CHARS (s, mem);
 234   SCM_SET_STRING_LENGTH (s, len);
 235
 236   return s;
 237 }
 238 #undef FUNC_NAME
 239
 240
 241 SCM_DEFINE (scm_make_string, "make-string", 1, 1, 0,
 242             (SCM k, SCM chr),
 243             "Return a newly allocated string of\n"
 244             "length @var{k}.  If @var{chr} is given, then all elements of\n"
 245             "the string are initialized to @var{chr}, otherwise the contents\n"
 246             "of the @var{string} are unspecified.\n")
 247 #define FUNC_NAME s_scm_make_string
 248 {
 249   if (SCM_INUMP (k))
 250     {
 251       long int i = SCM_INUM (k);
 252       SCM res;
 253
 254       SCM_ASSERT_RANGE (1, k, i >= 0);
 255
 256       res = scm_allocate_string (i);
 257       if (!SCM_UNBNDP (chr))
 258         {
 259           unsigned char *dst;
 260
 261           SCM_VALIDATE_CHAR (2, chr);
 262
 263           dst = SCM_STRING_UCHARS (res);
 264           memset (dst, SCM_CHAR (chr), i);
 265         }
 266
 267       return res;
 268     }
 269   else if (SCM_BIGP (k))
 270     SCM_OUT_OF_RANGE (1, k);
 271   else
 272     SCM_WRONG_TYPE_ARG (1, k);
 273 }
 274 #undef FUNC_NAME
 275
 276
 277 SCM_DEFINE (scm_string_length, "string-length", 1, 0, 0,
 278             (SCM string),
 279             "Return the number of characters in @var{string}.")
 280 #define FUNC_NAME s_scm_string_length
 281 {
 282   SCM_VALIDATE_STRING (1, string);
 283   return SCM_MAKINUM (SCM_STRING_LENGTH (string));
 284 }
 285 #undef FUNC_NAME
 286
 287 SCM_DEFINE (scm_string_ref, "string-ref", 2, 0, 0,
 288             (SCM str, SCM k),
 289             "Return character @var{k} of @var{str} using zero-origin\n"
 290             "indexing. @var{k} must be a valid index of @var{str}.")
 291 #define FUNC_NAME s_scm_string_ref
 292 {
 293   long idx;
 294
 295   SCM_VALIDATE_STRING (1, str);
 296   SCM_VALIDATE_INUM_COPY (2, k, idx);
 297   SCM_ASSERT_RANGE (2, k, idx >= 0 && idx < SCM_STRING_LENGTH (str));
 298   return SCM_MAKE_CHAR (SCM_STRING_UCHARS (str)[idx]);
 299 }
 300 #undef FUNC_NAME
 301
 302
 303 SCM_DEFINE (scm_string_set_x, "string-set!", 3, 0, 0,
 304             (SCM str, SCM k, SCM chr),
 305             "Store @var{chr} in element @var{k} of @var{str} and return\n"
 306             "an unspecified value. @var{k} must be a valid index of\n"
 307             "@var{str}.")
 308 #define FUNC_NAME s_scm_string_set_x
 309 {
 310 #if (SCM_DEBUG_DEPRECATED == 0)
 311   SCM_VALIDATE_RWSTRING (1, str);
 312 #else
 313   SCM_VALIDATE_STRING (1, str);
 314 #endif
 315   SCM_VALIDATE_INUM_RANGE (2,k,0,SCM_STRING_LENGTH(str));
 316   SCM_VALIDATE_CHAR (3,chr);
 317   SCM_STRING_UCHARS (str)[SCM_INUM (k)] = SCM_CHAR (chr);
 318   return SCM_UNSPECIFIED;
 319 }
 320 #undef FUNC_NAME
 321
 322
 323 SCM_DEFINE (scm_substring, "substring", 2, 1, 0,
 324             (SCM str, SCM start, SCM end),
 325             "Return a newly allocated string formed from the characters\n"
 326             "of @var{str} beginning with index @var{start} (inclusive) and\n"
 327             "ending with index @var{end} (exclusive).\n"
 328             "@var{str} must be a string, @var{start} and @var{end} must be\n"
 329             "exact integers satisfying:\n\n"
 330             "0 <= @var{start} <= @var{end} <= (string-length @var{str}).")
 331 #define FUNC_NAME s_scm_substring
 332 {
 333   long int from;
 334   long int to;
 335
 336   SCM_VALIDATE_STRING (1, str);
 337   SCM_VALIDATE_INUM (2, start);
 338   SCM_VALIDATE_INUM_DEF (3, end, SCM_STRING_LENGTH (str));
 339
 340   from = SCM_INUM (start);
 341   SCM_ASSERT_RANGE (2, start, 0 <= from && from <= SCM_STRING_LENGTH (str));
 342   to = SCM_INUM (end);
 343   SCM_ASSERT_RANGE (3, end, from <= to && to <= SCM_STRING_LENGTH (str));
 344
 345   return scm_makfromstr (&SCM_STRING_CHARS (str)[from], (size_t) (to - from), 0);
 346 }
 347 #undef FUNC_NAME
 348
 349
 350 SCM_DEFINE (scm_string_append, "string-append", 0, 0, 1,
 351             (SCM args),
 352             "Return a newly allocated string whose characters form the\n"
 353             "concatenation of the given strings, @var{args}.")
 354 #define FUNC_NAME s_scm_string_append
 355 {
 356   SCM res;
 357   size_t i = 0;
 358   register SCM l, s;
 359   register unsigned char *data;
 360
 361   SCM_VALIDATE_REST_ARGUMENT (args);
 362   for (l = args; !SCM_NULLP (l); l = SCM_CDR (l)) {
 363     s = SCM_CAR (l);
 364     SCM_VALIDATE_STRING (SCM_ARGn,s);
 365     i += SCM_STRING_LENGTH (s);
 366   }
 367   res = scm_allocate_string (i);
 368   data = SCM_STRING_UCHARS (res);
 369   for (l = args;SCM_NIMP (l);l = SCM_CDR (l)) {
 370     s = SCM_CAR (l);
 371     for (i = 0;i<SCM_STRING_LENGTH (s);i++) *data++ = SCM_STRING_UCHARS (s)[i];
 372   }
 373   return res;
 374 }
 375 #undef FUNC_NAME
 376
 377 #if SCM_DEBUG_DEPRECATED == 0
 378
 379 /* Explicit shared substrings will disappear from Guile.
 380  *
 381  * Instead, "normal" strings will be implemented using sharing
 382  * internally, combined with a copy-on-write strategy.
 383  */
 384
 385 SCM_DEFINE (scm_make_shared_substring, "make-shared-substring", 1, 2, 0,
 386             (SCM str, SCM start, SCM end),
 387             "Return a shared substring of @var{str}.  The arguments are the\n"
 388             "same as for the @code{substring} function: the shared substring\n"
 389             "returned includes all of the text from @var{str} between\n"
 390             "indexes @var{start} (inclusive) and @var{end} (exclusive).  If\n"
 391             "@var{end} is omitted, it defaults to the end of @var{str}.  The\n"
 392             "shared substring returned by @code{make-shared-substring}\n"
 393             "occupies the same storage space as @var{str}.")
 394 #define FUNC_NAME s_scm_make_shared_substring
 395 {
 396   long f;
 397   long t;
 398   SCM answer;
 399   SCM len_str;
 400
 401   SCM_VALIDATE_ROSTRING (1,str);
 402   SCM_VALIDATE_INUM_DEF_COPY (2,start,0,f);
 403   SCM_VALIDATE_INUM_DEF_COPY (3,end,SCM_ROLENGTH(str),t);
 404
 405   SCM_ASSERT_RANGE (2,start,(f >= 0));
 406   SCM_ASSERT_RANGE (3,end, (f <= t) && (t <= SCM_ROLENGTH (str)));
 407
 408   SCM_NEWCELL (answer);
 409   SCM_NEWCELL (len_str);
 410
 411   SCM_DEFER_INTS;
 412   if (SCM_SUBSTRP (str))
 413     {
 414       long offset;
 415       offset = SCM_INUM (SCM_SUBSTR_OFFSET (str));
 416       f += offset;
 417       t += offset;
 418       SCM_SETCAR (len_str, SCM_MAKINUM (f));
 419       SCM_SETCDR (len_str, SCM_SUBSTR_STR (str));
 420       SCM_SETCDR (answer, len_str);
 421       SCM_SETLENGTH (answer, t - f, scm_tc7_substring);
 422     }
 423   else
 424     {
 425       SCM_SETCAR (len_str, SCM_MAKINUM (f));
 426       SCM_SETCDR (len_str, str);
 427       SCM_SETCDR (answer, len_str);
 428       SCM_SETLENGTH (answer, t - f, scm_tc7_substring);
 429     }
 430   SCM_ALLOW_INTS;
 431   return answer;
 432 }
 433 #undef FUNC_NAME
 434
 435 #endif /* DEPRECATED */
 436
 437 void
 438 scm_init_strings ()
 439 {
 440   scm_nullstr = scm_allocate_string (0);
 441
 442 #ifndef SCM_MAGIC_SNARFER
 443 #include "libguile/strings.x"
 444 #endif
 445 }
 446
 447
 448 /*
 449   Local Variables:
 450   c-file-style: "gnu"
 451   End:
 452 */