lib/search/search.c

   1 /*
   2    Search text engine.
   3    Interface functions
   4
   5    Copyright (C) 2009 The Free Software Foundation, Inc.
   6
   7    Written by:
   8    Slava Zanko <slavazanko@gmail.com>, 2009.
   9
  10    This file is part of the Midnight Commander.
  11
  12    The Midnight Commander is free software; you can redistribute it
  13    and/or modify it under the terms of the GNU General Public License as
  14    published by the Free Software Foundation; either version 2 of the
  15    License, or (at your option) any later version.
  16
  17    The Midnight Commander is distributed in the hope that it will be
  18    useful, but WITHOUT ANY WARRANTY; without even the implied warranty
  19    of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  20    General Public License for more details.
  21
  22    You should have received a copy of the GNU General Public License
  23    along with this program; if not, write to the Free Software
  24    Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston,
  25    MA 02110-1301, USA.
  26  */
  27
  28 #include <config.h>
  29
  30 #include <stdlib.h>
  31 #include <sys/types.h>
  32
  33 #include "lib/global.h"
  34 #include "lib/strutil.h"
  35 #include "lib/search.h"
  36
  37 #include "src/charsets.h"
  38
  39 #include "internal.h"
  40
  41 /*** global variables ****************************************************************************/
  42
  43 /*** file scope macro definitions ****************************************************************/
  44
  45 /*** file scope type declarations ****************************************************************/
  46
  47 /*** file scope variables ************************************************************************/
  48
  49 static const mc_search_type_str_t mc_search__list_types[] = {
  50     {N_("Normal"), MC_SEARCH_T_NORMAL},
  51     {N_("&Regular expression"), MC_SEARCH_T_REGEX},
  52     {N_("Hexadecimal"), MC_SEARCH_T_HEX},
  53     {N_("Wildcard search"), MC_SEARCH_T_GLOB},
  54     {NULL, -1}
  55 };
  56
  57 /*** file scope functions ************************************************************************/
  58
  59 static mc_search_cond_t *
  60 mc_search__cond_struct_new (mc_search_t * lc_mc_search, const char *str,
  61                             gsize str_len, const char *charset)
  62 {
  63     mc_search_cond_t *mc_search_cond;
  64     mc_search_cond = g_malloc0 (sizeof (mc_search_cond_t));
  65
  66     mc_search_cond->str = g_string_new_len (str, str_len);
  67     mc_search_cond->charset = g_strdup (charset);
  68
  69     switch (lc_mc_search->search_type) {
  70     case MC_SEARCH_T_GLOB:
  71         mc_search__cond_struct_new_init_glob (charset, lc_mc_search, mc_search_cond);
  72         break;
  73     case MC_SEARCH_T_NORMAL:
  74         mc_search__cond_struct_new_init_normal (charset, lc_mc_search, mc_search_cond);
  75         break;
  76     case MC_SEARCH_T_REGEX:
  77         mc_search__cond_struct_new_init_regex (charset, lc_mc_search, mc_search_cond);
  78         break;
  79     case MC_SEARCH_T_HEX:
  80         mc_search__cond_struct_new_init_hex (charset, lc_mc_search, mc_search_cond);
  81         break;
  82     default:
  83         break;
  84     }
  85     return mc_search_cond;
  86 }
  87
  88 /* --------------------------------------------------------------------------------------------- */
  89
  90 static void
  91 mc_search__cond_struct_free (mc_search_cond_t * mc_search_cond)
  92 {
  93     if (mc_search_cond->upper)
  94         g_string_free (mc_search_cond->upper, TRUE);
  95
  96     if (mc_search_cond->lower)
  97         g_string_free (mc_search_cond->lower, TRUE);
  98
  99     g_string_free (mc_search_cond->str, TRUE);
 100     g_free (mc_search_cond->charset);
 101
 102 #ifdef SEARCH_TYPE_GLIB
 103     if (mc_search_cond->regex_handle)
 104         g_regex_unref (mc_search_cond->regex_handle);
 105 #else /* SEARCH_TYPE_GLIB */
 106     g_free (mc_search_cond->regex_handle);
 107 #endif /* SEARCH_TYPE_GLIB */
 108
 109     g_free (mc_search_cond);
 110 }
 111
 112 /* --------------------------------------------------------------------------------------------- */
 113
 114 static void
 115 mc_search__conditions_free (GPtrArray * array)
 116 {
 117     gsize loop1;
 118     mc_search_cond_t *lc_mc_search;
 119
 120     for (loop1 = 0; loop1 < array->len; loop1++) {
 121         lc_mc_search = (mc_search_cond_t *) g_ptr_array_index (array, loop1);
 122         mc_search__cond_struct_free (lc_mc_search);
 123     }
 124     g_ptr_array_free (array, TRUE);
 125 }
 126
 127 /* --------------------------------------------------------------------------------------------- */
 128
 129
 130
 131 /*** public functions ****************************************************************************/
 132
 133 mc_search_t *
 134 mc_search_new (const gchar * original, gsize str_len)
 135 {
 136     mc_search_t *lc_mc_search;
 137     if (!original)
 138         return NULL;
 139
 140     if ((gssize) str_len == -1) {
 141         str_len = strlen (original);
 142         if (str_len == 0)
 143             return NULL;
 144     }
 145
 146     lc_mc_search = g_malloc0 (sizeof (mc_search_t));
 147     lc_mc_search->original = g_strndup (original, str_len);
 148     lc_mc_search->original_len = str_len;
 149     return lc_mc_search;
 150 }
 151
 152 /* --------------------------------------------------------------------------------------------- */
 153
 154 void
 155 mc_search_free (mc_search_t * lc_mc_search)
 156 {
 157     if (lc_mc_search == NULL)
 158         return;
 159
 160     g_free (lc_mc_search->original);
 161     g_free (lc_mc_search->error_str);
 162
 163     if (lc_mc_search->conditions)
 164         mc_search__conditions_free (lc_mc_search->conditions);
 165
 166 #ifdef SEARCH_TYPE_GLIB
 167     if (lc_mc_search->regex_match_info)
 168         g_match_info_free (lc_mc_search->regex_match_info);
 169 #else /* SEARCH_TYPE_GLIB */
 170     g_free (lc_mc_search->regex_match_info);
 171 #endif /* SEARCH_TYPE_GLIB */
 172
 173     if (lc_mc_search->regex_buffer != NULL)
 174         g_string_free (lc_mc_search->regex_buffer, TRUE);
 175
 176     g_free (lc_mc_search);
 177 }
 178
 179 /* --------------------------------------------------------------------------------------------- */
 180
 181 gboolean
 182 mc_search_prepare (mc_search_t * lc_mc_search)
 183 {
 184     GPtrArray *ret;
 185     ret = g_ptr_array_new ();
 186 #ifdef HAVE_CHARSET
 187     if (lc_mc_search->is_all_charsets) {
 188         gsize loop1, recoded_str_len;
 189         gchar *buffer;
 190         for (loop1 = 0; loop1 < codepages->len; loop1++) {
 191             const char *id = ((codepage_desc *) g_ptr_array_index (codepages, loop1))->id;
 192             if (!g_ascii_strcasecmp (id, cp_source)) {
 193                 g_ptr_array_add (ret,
 194                                  mc_search__cond_struct_new (lc_mc_search, lc_mc_search->original,
 195                                                              lc_mc_search->original_len, cp_source));
 196                 continue;
 197             }
 198
 199             buffer =
 200                 mc_search__recode_str (lc_mc_search->original, lc_mc_search->original_len, cp_source,
 201                                        id, &recoded_str_len);
 202
 203             g_ptr_array_add (ret,
 204                              mc_search__cond_struct_new (lc_mc_search, buffer,
 205                                                          recoded_str_len, id));
 206             g_free (buffer);
 207         }
 208     } else {
 209         g_ptr_array_add (ret,
 210                          (gpointer) mc_search__cond_struct_new (lc_mc_search, lc_mc_search->original,
 211                                                                 lc_mc_search->original_len,
 212                                                                 cp_source));
 213     }
 214 #else
 215     g_ptr_array_add (ret,
 216                      (gpointer) mc_search__cond_struct_new (lc_mc_search, lc_mc_search->original,
 217                                                             lc_mc_search->original_len,
 218                                                             str_detect_termencoding ()));
 219 #endif
 220     lc_mc_search->conditions = ret;
 221
 222     return (lc_mc_search->error == MC_SEARCH_E_OK);
 223 }
 224
 225 /* --------------------------------------------------------------------------------------------- */
 226
 227 gboolean
 228 mc_search_run (mc_search_t * lc_mc_search, const void *user_data,
 229                gsize start_search, gsize end_search, gsize * found_len)
 230 {
 231     gboolean ret = FALSE;
 232
 233     if (!lc_mc_search)
 234         return FALSE;
 235     if (!mc_search_is_type_avail (lc_mc_search->search_type)) {
 236         lc_mc_search->error = MC_SEARCH_E_INPUT;
 237         lc_mc_search->error_str = g_strdup (_(STR_E_UNKNOWN_TYPE));
 238         return FALSE;
 239     }
 240 #ifdef SEARCH_TYPE_GLIB
 241     if (lc_mc_search->regex_match_info) {
 242         g_match_info_free (lc_mc_search->regex_match_info);
 243         lc_mc_search->regex_match_info = NULL;
 244     }
 245 #endif /* SEARCH_TYPE_GLIB */
 246
 247     lc_mc_search->error = MC_SEARCH_E_OK;
 248     g_free (lc_mc_search->error_str);
 249     lc_mc_search->error_str = NULL;
 250
 251     if ((lc_mc_search->conditions == NULL) && !mc_search_prepare (lc_mc_search))
 252         return FALSE;
 253
 254
 255     switch (lc_mc_search->search_type) {
 256     case MC_SEARCH_T_NORMAL:
 257         ret = mc_search__run_normal (lc_mc_search, user_data, start_search, end_search, found_len);
 258         break;
 259     case MC_SEARCH_T_REGEX:
 260         ret = mc_search__run_regex (lc_mc_search, user_data, start_search, end_search, found_len);
 261         break;
 262     case MC_SEARCH_T_GLOB:
 263         ret = mc_search__run_glob (lc_mc_search, user_data, start_search, end_search, found_len);
 264         break;
 265     case MC_SEARCH_T_HEX:
 266         ret = mc_search__run_hex (lc_mc_search, user_data, start_search, end_search, found_len);
 267         break;
 268     default:
 269         break;
 270     }
 271     return ret;
 272 }
 273
 274 /* --------------------------------------------------------------------------------------------- */
 275
 276 gboolean
 277 mc_search_is_type_avail (mc_search_type_t search_type)
 278 {
 279     switch (search_type) {
 280     case MC_SEARCH_T_GLOB:
 281     case MC_SEARCH_T_NORMAL:
 282     case MC_SEARCH_T_REGEX:
 283     case MC_SEARCH_T_HEX:
 284         return TRUE;
 285     default:
 286         break;
 287     }
 288     return FALSE;
 289 }
 290
 291 /* --------------------------------------------------------------------------------------------- */
 292
 293 const mc_search_type_str_t *
 294 mc_search_types_list_get (size_t *num)
 295 {
 296     /* don't count last NULL item */
 297     if (num != NULL)
 298         *num = sizeof (mc_search__list_types) / sizeof (mc_search__list_types[0]) - 1;
 299
 300     return mc_search__list_types;
 301 }
 302
 303 /* --------------------------------------------------------------------------------------------- */
 304
 305 GString *
 306 mc_search_prepare_replace_str (mc_search_t * lc_mc_search, GString * replace_str)
 307 {
 308     GString *ret;
 309
 310     if (lc_mc_search == NULL)
 311         return g_string_new_len (replace_str->str, replace_str->len);
 312
 313     if (replace_str == NULL || replace_str->str == NULL || replace_str->len == 0)
 314         return g_string_new ("");
 315
 316     switch (lc_mc_search->search_type) {
 317     case MC_SEARCH_T_REGEX:
 318         ret = mc_search_regex_prepare_replace_str (lc_mc_search, replace_str);
 319         break;
 320     case MC_SEARCH_T_GLOB:
 321         ret = mc_search_glob_prepare_replace_str (lc_mc_search, replace_str);
 322         break;
 323     case MC_SEARCH_T_NORMAL:
 324         ret = mc_search_normal_prepare_replace_str (lc_mc_search, replace_str);
 325         break;
 326     case MC_SEARCH_T_HEX:
 327         ret = mc_search_hex_prepare_replace_str (lc_mc_search, replace_str);
 328         break;
 329     default:
 330         ret = g_string_new_len (replace_str->str, replace_str->len);
 331         break;
 332     }
 333     return ret;
 334 }
 335
 336 /* --------------------------------------------------------------------------------------------- */
 337
 338 char *
 339 mc_search_prepare_replace_str2 (mc_search_t * lc_mc_search, char *replace_str)
 340 {
 341     GString *ret;
 342     GString *replace_str2;
 343
 344     replace_str2 = g_string_new (replace_str);
 345     ret = mc_search_prepare_replace_str (lc_mc_search, replace_str2);
 346     g_string_free (replace_str2, TRUE);
 347     return (ret != NULL) ? g_string_free (ret, FALSE) : NULL;
 348 }
 349
 350 /* --------------------------------------------------------------------------------------------- */
 351
 352 gboolean
 353 mc_search_is_fixed_search_str (mc_search_t * lc_mc_search)
 354 {
 355     if (lc_mc_search == NULL)
 356         return FALSE;
 357     switch (lc_mc_search->search_type) {
 358     case MC_SEARCH_T_REGEX:
 359     case MC_SEARCH_T_GLOB:
 360         return FALSE;
 361     default:
 362         return TRUE;
 363     }
 364 }
 365
 366 /* --------------------------------------------------------------------------------------------- */
 367
 368 gboolean
 369 mc_search (const gchar * pattern, const gchar * str, mc_search_type_t type)
 370 {
 371     gboolean ret;
 372     mc_search_t *search = mc_search_new (pattern, -1);
 373     if (search == NULL)
 374         return FALSE;
 375     search->search_type = type;
 376     search->is_case_sensitive = TRUE;
 377
 378     if (type == MC_SEARCH_T_GLOB)
 379         search->is_entire_line = TRUE;
 380
 381     ret = mc_search_run (search, str, 0, strlen (str), NULL);
 382     mc_search_free (search);
 383     return ret;
 384 }
 385
 386 /* --------------------------------------------------------------------------------------------- */
 387
 388 int
 389 mc_search_getstart_result_by_num (mc_search_t * lc_mc_search, int lc_index)
 390 {
 391     if (!lc_mc_search)
 392         return 0;
 393     if (lc_mc_search->search_type == MC_SEARCH_T_NORMAL)
 394         return 0;
 395 #ifdef SEARCH_TYPE_GLIB
 396     {
 397         gint start_pos;
 398         gint end_pos;
 399         g_match_info_fetch_pos (lc_mc_search->regex_match_info, lc_index, &start_pos, &end_pos);
 400         return (int) start_pos;
 401     }
 402 #else /* SEARCH_TYPE_GLIB */
 403     return lc_mc_search->iovector[lc_index * 2];
 404 #endif /* SEARCH_TYPE_GLIB */
 405
 406 }
 407
 408 /* --------------------------------------------------------------------------------------------- */
 409
 410 int
 411 mc_search_getend_result_by_num (mc_search_t * lc_mc_search, int lc_index)
 412 {
 413     if (!lc_mc_search)
 414         return 0;
 415     if (lc_mc_search->search_type == MC_SEARCH_T_NORMAL)
 416         return 0;
 417 #ifdef SEARCH_TYPE_GLIB
 418     {
 419         gint start_pos;
 420         gint end_pos;
 421         g_match_info_fetch_pos (lc_mc_search->regex_match_info, lc_index, &start_pos, &end_pos);
 422         return (int) end_pos;
 423     }
 424 #else /* SEARCH_TYPE_GLIB */
 425     return lc_mc_search->iovector[lc_index * 2 + 1];
 426 #endif /* SEARCH_TYPE_GLIB */
 427
 428 }
 429
 430 /* --------------------------------------------------------------------------------------------- */