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->len = str_len;
  68     mc_search_cond->charset = g_strdup (charset);
  69
  70     switch (lc_mc_search->search_type) {
  71     case MC_SEARCH_T_GLOB:
  72         mc_search__cond_struct_new_init_glob (charset, lc_mc_search, mc_search_cond);
  73         break;
  74     case MC_SEARCH_T_NORMAL:
  75         mc_search__cond_struct_new_init_normal (charset, lc_mc_search, mc_search_cond);
  76         break;
  77     case MC_SEARCH_T_REGEX:
  78         mc_search__cond_struct_new_init_regex (charset, lc_mc_search, mc_search_cond);
  79         break;
  80     case MC_SEARCH_T_HEX:
  81         mc_search__cond_struct_new_init_hex (charset, lc_mc_search, mc_search_cond);
  82         break;
  83     default:
  84         break;
  85     }
  86     return mc_search_cond;
  87 }
  88
  89 /* --------------------------------------------------------------------------------------------- */
  90
  91 static void
  92 mc_search__cond_struct_free (mc_search_cond_t * mc_search_cond)
  93 {
  94     if (mc_search_cond->upper)
  95         g_string_free (mc_search_cond->upper, TRUE);
  96
  97     if (mc_search_cond->lower)
  98         g_string_free (mc_search_cond->lower, TRUE);
  99
 100     g_string_free (mc_search_cond->str, TRUE);
 101     g_free (mc_search_cond->charset);
 102
 103 #ifdef SEARCH_TYPE_GLIB
 104     if (mc_search_cond->regex_handle)
 105         g_regex_unref (mc_search_cond->regex_handle);
 106 #else /* SEARCH_TYPE_GLIB */
 107     g_free (mc_search_cond->regex_handle);
 108 #endif /* SEARCH_TYPE_GLIB */
 109
 110     g_free (mc_search_cond);
 111 }
 112
 113 /* --------------------------------------------------------------------------------------------- */
 114
 115 static void
 116 mc_search__conditions_free (GPtrArray * array)
 117 {
 118     gsize loop1;
 119     mc_search_cond_t *lc_mc_search;
 120
 121     for (loop1 = 0; loop1 < array->len; loop1++) {
 122         lc_mc_search = (mc_search_cond_t *) g_ptr_array_index (array, loop1);
 123         mc_search__cond_struct_free (lc_mc_search);
 124     }
 125     g_ptr_array_free (array, TRUE);
 126 }
 127
 128 /* --------------------------------------------------------------------------------------------- */
 129
 130
 131
 132 /*** public functions ****************************************************************************/
 133
 134 mc_search_t *
 135 mc_search_new (const gchar * original, gsize str_len)
 136 {
 137     mc_search_t *lc_mc_search;
 138     if (!original)
 139         return NULL;
 140
 141     if ((gssize) str_len == -1) {
 142         str_len = strlen (original);
 143         if (str_len == 0)
 144             return NULL;
 145     }
 146
 147     lc_mc_search = g_malloc0 (sizeof (mc_search_t));
 148     lc_mc_search->original = g_strndup (original, str_len);
 149     lc_mc_search->original_len = str_len;
 150     return lc_mc_search;
 151 }
 152
 153 /* --------------------------------------------------------------------------------------------- */
 154
 155 void
 156 mc_search_free (mc_search_t * lc_mc_search)
 157 {
 158     if (!lc_mc_search)
 159         return;
 160
 161     g_free (lc_mc_search->original);
 162     g_free (lc_mc_search->error_str);
 163
 164     if (lc_mc_search->conditions)
 165         mc_search__conditions_free (lc_mc_search->conditions);
 166
 167 #ifdef SEARCH_TYPE_GLIB
 168     if (lc_mc_search->regex_match_info)
 169         g_match_info_free (lc_mc_search->regex_match_info);
 170 #else /* SEARCH_TYPE_GLIB */
 171     g_free (lc_mc_search->regex_match_info);
 172 #endif /* SEARCH_TYPE_GLIB */
 173
 174     if (lc_mc_search->regex_buffer != NULL)
 175         g_string_free (lc_mc_search->regex_buffer, TRUE);
 176
 177     g_free (lc_mc_search);
 178 }
 179
 180 /* --------------------------------------------------------------------------------------------- */
 181
 182 gboolean
 183 mc_search_prepare (mc_search_t * lc_mc_search)
 184 {
 185     GPtrArray *ret;
 186     ret = g_ptr_array_new ();
 187 #ifdef HAVE_CHARSET
 188     if (lc_mc_search->is_all_charsets) {
 189         gsize loop1, recoded_str_len;
 190         gchar *buffer;
 191         for (loop1 = 0; loop1 < (gsize) n_codepages; loop1++) {
 192             if (!g_ascii_strcasecmp (codepages[loop1].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                                        codepages[loop1].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, codepages[loop1].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         break;
 286     default:
 287         break;
 288     }
 289     return FALSE;
 290 }
 291
 292 /* --------------------------------------------------------------------------------------------- */
 293
 294 const mc_search_type_str_t *
 295 mc_search_types_list_get (size_t *num)
 296 {
 297     /* don't count last NULL item */
 298     if (num != NULL)
 299         *num = sizeof (mc_search__list_types) / sizeof (mc_search__list_types[0]) - 1;
 300
 301     return mc_search__list_types;
 302 }
 303
 304 /* --------------------------------------------------------------------------------------------- */
 305
 306 GString *
 307 mc_search_prepare_replace_str (mc_search_t * lc_mc_search, GString * replace_str)
 308 {
 309     GString *ret;
 310
 311     if (lc_mc_search == NULL)
 312         return g_string_new_len (replace_str->str, replace_str->len);
 313
 314     if (replace_str == NULL || replace_str->str == NULL || replace_str->len == 0)
 315         return g_string_new ("");
 316
 317     switch (lc_mc_search->search_type) {
 318     case MC_SEARCH_T_REGEX:
 319         ret = mc_search_regex_prepare_replace_str (lc_mc_search, replace_str);
 320         break;
 321     case MC_SEARCH_T_GLOB:
 322         ret = mc_search_glob_prepare_replace_str (lc_mc_search, replace_str);
 323         break;
 324     case MC_SEARCH_T_NORMAL:
 325         ret = mc_search_normal_prepare_replace_str (lc_mc_search, replace_str);
 326         break;
 327     case MC_SEARCH_T_HEX:
 328         ret = mc_search_hex_prepare_replace_str (lc_mc_search, replace_str);
 329         break;
 330     default:
 331         ret = g_string_new_len (replace_str->str, replace_str->len);
 332         break;
 333     }
 334     return ret;
 335 }
 336
 337 /* --------------------------------------------------------------------------------------------- */
 338
 339 char *
 340 mc_search_prepare_replace_str2 (mc_search_t * lc_mc_search, char *replace_str)
 341 {
 342     GString *ret;
 343     GString *replace_str2 = g_string_new (replace_str);
 344     ret = mc_search_prepare_replace_str (lc_mc_search, replace_str2);
 345     g_string_free (replace_str2, TRUE);
 346     if (ret)
 347         return g_string_free (ret, FALSE);
 348     return NULL;
 349 }
 350
 351 /* --------------------------------------------------------------------------------------------- */
 352
 353 gboolean
 354 mc_search_is_fixed_search_str (mc_search_t * lc_mc_search)
 355 {
 356     if (lc_mc_search == NULL)
 357         return FALSE;
 358     switch (lc_mc_search->search_type) {
 359     case MC_SEARCH_T_REGEX:
 360     case MC_SEARCH_T_GLOB:
 361         return FALSE;
 362         break;
 363     default:
 364         return TRUE;
 365         break;
 366     }
 367 }
 368
 369 /* --------------------------------------------------------------------------------------------- */
 370
 371 gboolean
 372 mc_search (const gchar * pattern, const gchar * str, mc_search_type_t type)
 373 {
 374     gboolean ret;
 375     mc_search_t *search = mc_search_new (pattern, -1);
 376     if (search == NULL)
 377         return FALSE;
 378     search->search_type = type;
 379     search->is_case_sentitive = TRUE;
 380
 381     if (type == MC_SEARCH_T_GLOB)
 382         search->is_entire_line = TRUE;
 383
 384     ret = mc_search_run (search, str, 0, strlen (str), NULL);
 385     mc_search_free (search);
 386     return ret;
 387 }
 388
 389 /* --------------------------------------------------------------------------------------------- */
 390
 391 int
 392 mc_search_getstart_rezult_by_num (mc_search_t * lc_mc_search, int lc_index)
 393 {
 394     if (!lc_mc_search)
 395         return 0;
 396     if (lc_mc_search->search_type == MC_SEARCH_T_NORMAL)
 397         return 0;
 398 #ifdef SEARCH_TYPE_GLIB
 399     {
 400         gint start_pos;
 401         gint end_pos;
 402         g_match_info_fetch_pos (lc_mc_search->regex_match_info, lc_index, &start_pos, &end_pos);
 403         return (int) start_pos;
 404     }
 405 #else /* SEARCH_TYPE_GLIB */
 406     return lc_mc_search->iovector[lc_index * 2];
 407 #endif /* SEARCH_TYPE_GLIB */
 408
 409 }
 410
 411 /* --------------------------------------------------------------------------------------------- */
 412
 413 int
 414 mc_search_getend_rezult_by_num (mc_search_t * lc_mc_search, int lc_index)
 415 {
 416     if (!lc_mc_search)
 417         return 0;
 418     if (lc_mc_search->search_type == MC_SEARCH_T_NORMAL)
 419         return 0;
 420 #ifdef SEARCH_TYPE_GLIB
 421     {
 422         gint start_pos;
 423         gint end_pos;
 424         g_match_info_fetch_pos (lc_mc_search->regex_match_info, lc_index, &start_pos, &end_pos);
 425         return (int) end_pos;
 426     }
 427 #else /* SEARCH_TYPE_GLIB */
 428     return lc_mc_search->iovector[lc_index * 2 + 1];
 429 #endif /* SEARCH_TYPE_GLIB */
 430
 431 }
 432
 433 /* --------------------------------------------------------------------------------------------- */