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 "../src/global.h"
  34 #include "../src/search/search.h"
  35 #include "../src/search/internal.h"
  36 #include "../src/strutil.h"
  37 #include "../src/charsets.h"
  38
  39 /*** global variables ****************************************************************************/
  40
  41 /*** file scope macro definitions ****************************************************************/
  42
  43 /*** file scope type declarations ****************************************************************/
  44
  45 /*** file scope variables ************************************************************************/
  46
  47 static const mc_search_type_str_t mc_search__list_types[] = {
  48     {N_("Normal"), MC_SEARCH_T_NORMAL},
  49     {N_("&Regular expression"), MC_SEARCH_T_REGEX},
  50     {N_("Hexadecimal"), MC_SEARCH_T_HEX},
  51     {N_("Wildcard search"), MC_SEARCH_T_GLOB},
  52     {NULL, -1}
  53 };
  54
  55 /*** file scope functions ************************************************************************/
  56
  57 static mc_search_cond_t *
  58 mc_search__cond_struct_new (mc_search_t * lc_mc_search, const char *str,
  59                             gsize str_len, const char *charset)
  60 {
  61     mc_search_cond_t *mc_search_cond;
  62     mc_search_cond = g_malloc0 (sizeof (mc_search_cond_t));
  63
  64     mc_search_cond->str = g_string_new_len (str, str_len);
  65     mc_search_cond->len = str_len;
  66     mc_search_cond->charset = g_strdup (charset);
  67
  68     switch (lc_mc_search->search_type) {
  69     case MC_SEARCH_T_GLOB:
  70         mc_search__cond_struct_new_init_glob (charset, lc_mc_search, mc_search_cond);
  71         break;
  72     case MC_SEARCH_T_NORMAL:
  73         mc_search__cond_struct_new_init_normal (charset, lc_mc_search, mc_search_cond);
  74         break;
  75     case MC_SEARCH_T_REGEX:
  76         mc_search__cond_struct_new_init_regex (charset, lc_mc_search, mc_search_cond);
  77         break;
  78     case MC_SEARCH_T_HEX:
  79         mc_search__cond_struct_new_init_hex (charset, lc_mc_search, mc_search_cond);
  80         break;
  81     default:
  82         break;
  83     }
  84     return mc_search_cond;
  85 }
  86
  87 /* --------------------------------------------------------------------------------------------- */
  88
  89 static void
  90 mc_search__cond_struct_free (mc_search_cond_t * mc_search_cond)
  91 {
  92     if (mc_search_cond->upper)
  93         g_string_free (mc_search_cond->upper, TRUE);
  94
  95     if (mc_search_cond->lower)
  96         g_string_free (mc_search_cond->lower, TRUE);
  97
  98     g_string_free (mc_search_cond->str, TRUE);
  99     g_free (mc_search_cond->charset);
 100
 101 #ifdef SEARCH_TYPE_GLIB
 102     if (mc_search_cond->regex_handle)
 103         g_regex_unref (mc_search_cond->regex_handle);
 104 #else /* SEARCH_TYPE_GLIB */
 105     g_free (mc_search_cond->regex_handle);
 106 #endif /* SEARCH_TYPE_GLIB */
 107
 108     g_free (mc_search_cond);
 109 }
 110
 111 /* --------------------------------------------------------------------------------------------- */
 112
 113 static void
 114 mc_search__conditions_free (GPtrArray * array)
 115 {
 116     gsize loop1;
 117     mc_search_cond_t *lc_mc_search;
 118
 119     for (loop1 = 0; loop1 < array->len; loop1++) {
 120         lc_mc_search = (mc_search_cond_t *) g_ptr_array_index (array, loop1);
 121         mc_search__cond_struct_free (lc_mc_search);
 122     }
 123     g_ptr_array_free (array, TRUE);
 124 }
 125
 126 /* --------------------------------------------------------------------------------------------- */
 127
 128
 129
 130 /*** public functions ****************************************************************************/
 131
 132 mc_search_t *
 133 mc_search_new (const gchar * original, gsize str_len)
 134 {
 135     mc_search_t *lc_mc_search;
 136     if (!original)
 137         return NULL;
 138
 139     if ((gssize) str_len == -1) {
 140         str_len = strlen (original);
 141         if (str_len == 0)
 142             return NULL;
 143     }
 144
 145     lc_mc_search = g_malloc0 (sizeof (mc_search_t));
 146     lc_mc_search->original = g_strndup (original, str_len);
 147     lc_mc_search->original_len = str_len;
 148     return lc_mc_search;
 149 }
 150
 151 /* --------------------------------------------------------------------------------------------- */
 152
 153 void
 154 mc_search_free (mc_search_t * lc_mc_search)
 155 {
 156     if (!lc_mc_search)
 157         return;
 158
 159     g_free (lc_mc_search->original);
 160     g_free (lc_mc_search->error_str);
 161
 162     if (lc_mc_search->conditions)
 163         mc_search__conditions_free (lc_mc_search->conditions);
 164
 165 #ifdef SEARCH_TYPE_GLIB
 166     if (lc_mc_search->regex_match_info)
 167         g_match_info_free (lc_mc_search->regex_match_info);
 168 #else /* SEARCH_TYPE_GLIB */
 169     g_free (lc_mc_search->regex_match_info);
 170 #endif /* SEARCH_TYPE_GLIB */
 171
 172     if (lc_mc_search->regex_buffer != NULL)
 173         g_string_free (lc_mc_search->regex_buffer, TRUE);
 174
 175     g_free (lc_mc_search);
 176 }
 177
 178 /* --------------------------------------------------------------------------------------------- */
 179
 180 gboolean
 181 mc_search_prepare (mc_search_t * lc_mc_search)
 182 {
 183     GPtrArray *ret;
 184     ret = g_ptr_array_new ();
 185 #ifdef HAVE_CHARSET
 186     if (lc_mc_search->is_all_charsets) {
 187         gsize loop1, recoded_str_len;
 188         gchar *buffer;
 189         for (loop1 = 0; loop1 < (gsize) n_codepages; loop1++) {
 190             if (!g_ascii_strcasecmp (codepages[loop1].id, cp_source)) {
 191                 g_ptr_array_add (ret,
 192                                  mc_search__cond_struct_new (lc_mc_search, lc_mc_search->original,
 193                                                              lc_mc_search->original_len, cp_source));
 194                 continue;
 195             }
 196
 197             buffer =
 198                 mc_search__recode_str (lc_mc_search->original, lc_mc_search->original_len, cp_source,
 199                                        codepages[loop1].id, &recoded_str_len);
 200
 201             g_ptr_array_add (ret,
 202                              mc_search__cond_struct_new (lc_mc_search, buffer,
 203                                                          recoded_str_len, codepages[loop1].id));
 204             g_free (buffer);
 205         }
 206     } else {
 207         g_ptr_array_add (ret,
 208                          (gpointer) mc_search__cond_struct_new (lc_mc_search, lc_mc_search->original,
 209                                                                 lc_mc_search->original_len,
 210                                                                 cp_source));
 211     }
 212 #else
 213     g_ptr_array_add (ret,
 214                      (gpointer) mc_search__cond_struct_new (lc_mc_search, lc_mc_search->original,
 215                                                             lc_mc_search->original_len,
 216                                                             str_detect_termencoding ()));
 217 #endif
 218     lc_mc_search->conditions = ret;
 219
 220     return (lc_mc_search->error == MC_SEARCH_E_OK);
 221 }
 222
 223 /* --------------------------------------------------------------------------------------------- */
 224
 225 gboolean
 226 mc_search_run (mc_search_t * lc_mc_search, const void *user_data,
 227                gsize start_search, gsize end_search, gsize * found_len)
 228 {
 229     gboolean ret = FALSE;
 230
 231     if (!lc_mc_search)
 232         return FALSE;
 233     if (!mc_search_is_type_avail (lc_mc_search->search_type)) {
 234         lc_mc_search->error = MC_SEARCH_E_INPUT;
 235         lc_mc_search->error_str = g_strdup (_(STR_E_UNKNOWN_TYPE));
 236         return FALSE;
 237     }
 238 #ifdef SEARCH_TYPE_GLIB
 239     if (lc_mc_search->regex_match_info) {
 240         g_match_info_free (lc_mc_search->regex_match_info);
 241         lc_mc_search->regex_match_info = NULL;
 242     }
 243 #endif /* SEARCH_TYPE_GLIB */
 244
 245     lc_mc_search->error = MC_SEARCH_E_OK;
 246     g_free (lc_mc_search->error_str);
 247     lc_mc_search->error_str = NULL;
 248
 249     if ((lc_mc_search->conditions == NULL) && !mc_search_prepare (lc_mc_search))
 250         return FALSE;
 251
 252
 253     switch (lc_mc_search->search_type) {
 254     case MC_SEARCH_T_NORMAL:
 255         ret = mc_search__run_normal (lc_mc_search, user_data, start_search, end_search, found_len);
 256         break;
 257     case MC_SEARCH_T_REGEX:
 258         ret = mc_search__run_regex (lc_mc_search, user_data, start_search, end_search, found_len);
 259         break;
 260     case MC_SEARCH_T_GLOB:
 261         ret = mc_search__run_glob (lc_mc_search, user_data, start_search, end_search, found_len);
 262         break;
 263     case MC_SEARCH_T_HEX:
 264         ret = mc_search__run_hex (lc_mc_search, user_data, start_search, end_search, found_len);
 265         break;
 266     default:
 267         break;
 268     }
 269     return ret;
 270 }
 271
 272 /* --------------------------------------------------------------------------------------------- */
 273
 274 gboolean
 275 mc_search_is_type_avail (mc_search_type_t search_type)
 276 {
 277     switch (search_type) {
 278     case MC_SEARCH_T_GLOB:
 279     case MC_SEARCH_T_NORMAL:
 280     case MC_SEARCH_T_REGEX:
 281     case MC_SEARCH_T_HEX:
 282         return TRUE;
 283         break;
 284     default:
 285         break;
 286     }
 287     return FALSE;
 288 }
 289
 290 /* --------------------------------------------------------------------------------------------- */
 291
 292 const mc_search_type_str_t *
 293 mc_search_types_list_get (size_t *num)
 294 {
 295     /* don't count last NULL item */
 296     if (num != NULL)
 297         *num = sizeof (mc_search__list_types) / sizeof (mc_search__list_types[0]) - 1;
 298
 299     return mc_search__list_types;
 300 }
 301
 302 /* --------------------------------------------------------------------------------------------- */
 303
 304 GString *
 305 mc_search_prepare_replace_str (mc_search_t * lc_mc_search, GString * replace_str)
 306 {
 307     GString *ret;
 308
 309     if (lc_mc_search == NULL)
 310         return g_string_new_len (replace_str->str, replace_str->len);
 311
 312     if (replace_str == NULL || replace_str->str == NULL || replace_str->len == 0)
 313         return g_string_new ("");
 314
 315     switch (lc_mc_search->search_type) {
 316     case MC_SEARCH_T_REGEX:
 317         ret = mc_search_regex_prepare_replace_str (lc_mc_search, replace_str);
 318         break;
 319     case MC_SEARCH_T_GLOB:
 320         ret = mc_search_glob_prepare_replace_str (lc_mc_search, replace_str);
 321         break;
 322     case MC_SEARCH_T_NORMAL:
 323         ret = mc_search_normal_prepare_replace_str (lc_mc_search, replace_str);
 324         break;
 325     case MC_SEARCH_T_HEX:
 326         ret = mc_search_hex_prepare_replace_str (lc_mc_search, replace_str);
 327         break;
 328     default:
 329         ret = g_string_new_len (replace_str->str, replace_str->len);
 330         break;
 331     }
 332     return ret;
 333 }
 334
 335 /* --------------------------------------------------------------------------------------------- */
 336
 337 char *
 338 mc_search_prepare_replace_str2 (mc_search_t * lc_mc_search, char *replace_str)
 339 {
 340     GString *ret;
 341     GString *replace_str2 = g_string_new (replace_str);
 342     ret = mc_search_prepare_replace_str (lc_mc_search, replace_str2);
 343     g_string_free (replace_str2, TRUE);
 344     if (ret)
 345         return g_string_free (ret, FALSE);
 346     return NULL;
 347 }
 348
 349 /* --------------------------------------------------------------------------------------------- */
 350
 351 gboolean
 352 mc_search_is_fixed_search_str (mc_search_t * lc_mc_search)
 353 {
 354     if (lc_mc_search == NULL)
 355         return FALSE;
 356     switch (lc_mc_search->search_type) {
 357     case MC_SEARCH_T_REGEX:
 358     case MC_SEARCH_T_GLOB:
 359         return FALSE;
 360         break;
 361     default:
 362         return TRUE;
 363         break;
 364     }
 365 }
 366
 367 /* --------------------------------------------------------------------------------------------- */
 368
 369 gboolean
 370 mc_search (const gchar * pattern, const gchar * str, mc_search_type_t type)
 371 {
 372     gboolean ret;
 373     mc_search_t *search = mc_search_new (pattern, -1);
 374     if (search == NULL)
 375         return FALSE;
 376     search->search_type = type;
 377     search->is_case_sentitive = TRUE;
 378
 379     if (type == MC_SEARCH_T_GLOB)
 380         search->is_entire_line = TRUE;
 381
 382     ret = mc_search_run (search, str, 0, strlen (str), NULL);
 383     mc_search_free (search);
 384     return ret;
 385 }
 386
 387 /* --------------------------------------------------------------------------------------------- */
 388
 389 int
 390 mc_search_getstart_rezult_by_num (mc_search_t * lc_mc_search, int lc_index)
 391 {
 392     if (!lc_mc_search)
 393         return 0;
 394     if (lc_mc_search->search_type == MC_SEARCH_T_NORMAL)
 395         return 0;
 396 #ifdef SEARCH_TYPE_GLIB
 397     {
 398         gint start_pos;
 399         gint end_pos;
 400         g_match_info_fetch_pos (lc_mc_search->regex_match_info, lc_index, &start_pos, &end_pos);
 401         return (int) start_pos;
 402     }
 403 #else /* SEARCH_TYPE_GLIB */
 404     return lc_mc_search->iovector[lc_index * 2];
 405 #endif /* SEARCH_TYPE_GLIB */
 406
 407 }
 408
 409 /* --------------------------------------------------------------------------------------------- */
 410
 411 int
 412 mc_search_getend_rezult_by_num (mc_search_t * lc_mc_search, int lc_index)
 413 {
 414     if (!lc_mc_search)
 415         return 0;
 416     if (lc_mc_search->search_type == MC_SEARCH_T_NORMAL)
 417         return 0;
 418 #ifdef SEARCH_TYPE_GLIB
 419     {
 420         gint start_pos;
 421         gint end_pos;
 422         g_match_info_fetch_pos (lc_mc_search->regex_match_info, lc_index, &start_pos, &end_pos);
 423         return (int) end_pos;
 424     }
 425 #else /* SEARCH_TYPE_GLIB */
 426     return lc_mc_search->iovector[lc_index * 2 + 1];
 427 #endif /* SEARCH_TYPE_GLIB */
 428
 429 }
 430
 431 /* --------------------------------------------------------------------------------------------- */