2 Text conversion from one charset to another.
4 Copyright (C) 2001, 2011
5 The Free Software Foundation, Inc.
8 Walery Studennikov <despair@sama.ru>
10 This file is part of the Midnight Commander.
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 3 of the License,
15 or (at your option) any later version.
17 The Midnight Commander is distributed in the hope that it will be useful,
18 but WITHOUT ANY WARRANTY; without even the implied warranty of
19 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
20 GNU General Public License for more details.
22 You should have received a copy of the GNU General Public License
23 along with this program. If not, see <http://www.gnu.org/licenses/>.
27 * \brief Source: Text conversion from one charset to another
38 #include "lib/global.h"
39 #include "lib/strutil.h" /* utf-8 functions */
40 #include "lib/fileloc.h"
41 #include "lib/charsets.h"
43 /*** global variables ****************************************************************************/
45 GPtrArray
*codepages
= NULL
;
47 unsigned char conv_displ
[256];
48 unsigned char conv_input
[256];
50 const char *cp_display
= NULL
;
51 const char *cp_source
= NULL
;
53 /*** file scope macro definitions ****************************************************************/
55 #define OTHER_8BIT "Other_8_bit"
58 * FIXME: This assumes that ASCII is always the first encoding
63 /*** file scope type declarations ****************************************************************/
65 /*** file scope variables ************************************************************************/
67 /*** file scope functions ************************************************************************/
68 /* --------------------------------------------------------------------------------------------- */
70 static codepage_desc
*
71 new_codepage_desc (const char *id
, const char *name
)
75 desc
= g_new (codepage_desc
, 1);
76 desc
->id
= g_strdup (id
);
77 desc
->name
= g_strdup (name
);
82 /* --------------------------------------------------------------------------------------------- */
85 free_codepage_desc (gpointer data
, gpointer user_data
)
87 codepage_desc
*desc
= (codepage_desc
*) data
;
95 /* --------------------------------------------------------------------------------------------- */
96 /* returns display codepage */
99 load_codepages_list_from_file (GPtrArray
** list
, const char *fname
)
103 char buf
[BUF_MEDIUM
];
104 char *default_codepage
= NULL
;
106 f
= fopen (fname
, "r");
110 for (i
= 0; fgets (buf
, sizeof buf
, f
) != NULL
;)
112 /* split string into id and cpname */
114 size_t buflen
= strlen (buf
);
116 if (*p
== '\n' || *p
== '\0' || *p
== '#')
119 if (buflen
> 0 && buf
[buflen
- 1] == '\n')
120 buf
[buflen
- 1] = '\0';
121 while (*p
!= '\t' && *p
!= ' ' && *p
!= '\0')
131 if (strcmp (buf
, "default") == 0)
132 default_codepage
= g_strdup (p
);
135 const char *id
= buf
;
139 *list
= g_ptr_array_sized_new (16);
140 g_ptr_array_add (*list
, new_codepage_desc (id
, p
));
144 /* whether id is already present in list */
145 /* if yes, overwrite description */
146 for (i
= 0; i
< (*list
)->len
; i
++)
150 desc
= (codepage_desc
*) g_ptr_array_index (*list
, i
);
152 if (strcmp (id
, desc
->id
) == 0)
156 desc
->name
= g_strdup (p
);
162 if (i
== (*list
)->len
)
163 g_ptr_array_add (*list
, new_codepage_desc (id
, p
));
168 if (default_codepage
!= NULL
)
170 mc_global
.display_codepage
= get_codepage_index (default_codepage
);
171 g_free (default_codepage
);
178 /* --------------------------------------------------------------------------------------------- */
181 translate_character (GIConv cd
, char c
)
183 gchar
*tmp_buff
= NULL
;
184 gsize bytes_read
, bytes_written
= 0;
185 const char *ibuf
= &c
;
190 tmp_buff
= g_convert_with_iconv (ibuf
, ibuflen
, cd
, &bytes_read
, &bytes_written
, NULL
);
197 /* --------------------------------------------------------------------------------------------- */
198 /*** public functions ****************************************************************************/
199 /* --------------------------------------------------------------------------------------------- */
202 load_codepages_list (void)
206 /* 1: try load /usr/share/mc/mc.charsets */
207 fname
= g_build_filename (mc_global
.share_data_dir
, CHARSETS_LIST
, (char *) NULL
);
208 load_codepages_list_from_file (&codepages
, fname
);
211 /* 2: try load /etc/mc/mc.charsets */
212 fname
= g_build_filename (mc_global
.sysconfig_dir
, CHARSETS_LIST
, (char *) NULL
);
213 load_codepages_list_from_file (&codepages
, fname
);
216 if (codepages
== NULL
)
218 /* files are not found, add defaullt codepage */
219 fprintf (stderr
, "%s\n", _("Warning: cannot load codepages list"));
221 codepages
= g_ptr_array_new ();
222 g_ptr_array_add (codepages
, new_codepage_desc ("ASCII", _("7-bit ASCII")));
226 /* --------------------------------------------------------------------------------------------- */
229 free_codepages_list (void)
231 g_ptr_array_foreach (codepages
, free_codepage_desc
, NULL
);
232 g_ptr_array_free (codepages
, TRUE
);
235 /* --------------------------------------------------------------------------------------------- */
238 get_codepage_id (const int n
)
240 return (n
< 0) ? OTHER_8BIT
: ((codepage_desc
*) g_ptr_array_index (codepages
, n
))->id
;
243 /* --------------------------------------------------------------------------------------------- */
246 get_codepage_index (const char *id
)
249 if (strcmp (id
, OTHER_8BIT
) == 0)
251 if (codepages
== NULL
)
253 for (i
= 0; i
< codepages
->len
; i
++)
254 if (strcmp (id
, ((codepage_desc
*) g_ptr_array_index (codepages
, i
))->id
) == 0)
259 /* --------------------------------------------------------------------------------------------- */
260 /** Check if specified encoding can be used in mc.
261 * @param encoding name of encoding
262 * @returns TRUE if encoding has supported by mc, FALSE otherwise
266 is_supported_encoding (const char *encoding
)
268 gboolean result
= FALSE
;
271 for (t
= 0; t
< codepages
->len
; t
++)
273 const char *id
= ((codepage_desc
*) g_ptr_array_index (codepages
, t
))->id
;
274 result
|= (g_ascii_strncasecmp (encoding
, id
, strlen (id
)) == 0);
280 /* --------------------------------------------------------------------------------------------- */
283 init_translation_table (int cpsource
, int cpdisplay
)
288 /* Fill inpit <-> display tables */
290 if (cpsource
< 0 || cpdisplay
< 0 || cpsource
== cpdisplay
)
292 for (i
= 0; i
<= 255; ++i
)
296 cp_source
= cp_display
;
301 for (i
= 0; i
<= 127; ++i
)
306 cp_source
= ((codepage_desc
*) g_ptr_array_index (codepages
, cpsource
))->id
;
307 cp_display
= ((codepage_desc
*) g_ptr_array_index (codepages
, cpdisplay
))->id
;
309 /* display <- inpit table */
311 cd
= g_iconv_open (cp_display
, cp_source
);
312 if (cd
== INVALID_CONV
)
313 return g_strdup_printf (_("Cannot translate from %s to %s"), cp_source
, cp_display
);
315 for (i
= 128; i
<= 255; ++i
)
316 conv_displ
[i
] = translate_character (cd
, i
);
320 /* inpit <- display table */
322 cd
= g_iconv_open (cp_source
, cp_display
);
323 if (cd
== INVALID_CONV
)
324 return g_strdup_printf (_("Cannot translate from %s to %s"), cp_display
, cp_source
);
326 for (i
= 128; i
<= 255; ++i
)
329 ch
= translate_character (cd
, i
);
330 conv_input
[i
] = (ch
== UNKNCHAR
) ? i
: ch
;
338 /* --------------------------------------------------------------------------------------------- */
341 convert_to_display (char *str
)
348 *str
= conv_displ
[(unsigned char) *str
];
353 /* --------------------------------------------------------------------------------------------- */
356 str_convert_to_display (char *str
)
358 return str_nconvert_to_display (str
, -1);
362 /* --------------------------------------------------------------------------------------------- */
365 str_nconvert_to_display (char *str
, int len
)
371 return g_string_new ("");
373 if (cp_display
== cp_source
)
374 return g_string_new (str
);
376 conv
= str_crt_conv_from (cp_source
);
378 buff
= g_string_new ("");
379 str_nconvert (conv
, str
, len
, buff
);
380 str_close_conv (conv
);
384 /* --------------------------------------------------------------------------------------------- */
387 convert_from_input (char *str
)
394 *str
= conv_input
[(unsigned char) *str
];
399 /* --------------------------------------------------------------------------------------------- */
402 str_convert_to_input (char *str
)
404 return str_nconvert_to_input (str
, -1);
407 /* --------------------------------------------------------------------------------------------- */
410 str_nconvert_to_input (char *str
, int len
)
416 return g_string_new ("");
418 if (cp_display
== cp_source
)
419 return g_string_new (str
);
421 conv
= str_crt_conv_to (cp_source
);
423 buff
= g_string_new ("");
424 str_nconvert (conv
, str
, len
, buff
);
425 str_close_conv (conv
);
429 /* --------------------------------------------------------------------------------------------- */
432 convert_from_utf_to_current (const char *str
)
434 unsigned char buf_ch
[6 + 1];
435 unsigned char ch
= '.';
442 cp_to
= get_codepage_id (mc_global
.source_codepage
);
443 conv
= str_crt_conv_to (cp_to
);
445 if (conv
!= INVALID_CONV
)
447 switch (str_translate_char (conv
, str
, -1, (char *) buf_ch
, sizeof (buf_ch
)))
457 str_close_conv (conv
);
464 /* --------------------------------------------------------------------------------------------- */
467 convert_from_utf_to_current_c (const int input_char
, GIConv conv
)
469 unsigned char str
[6 + 1];
470 unsigned char buf_ch
[6 + 1];
471 unsigned char ch
= '.';
475 res
= g_unichar_to_utf8 (input_char
, (char *) str
);
482 switch (str_translate_char (conv
, (char *) str
, -1, (char *) buf_ch
, sizeof (buf_ch
)))
495 /* --------------------------------------------------------------------------------------------- */
498 convert_from_8bit_to_utf_c (const char input_char
, GIConv conv
)
500 unsigned char str
[2];
501 unsigned char buf_ch
[6 + 1];
505 str
[0] = (unsigned char) input_char
;
508 switch (str_translate_char (conv
, (char *) str
, -1, (char *) buf_ch
, sizeof (buf_ch
)))
511 res
= g_utf8_get_char_validated ((char *) buf_ch
, -1);
529 /* --------------------------------------------------------------------------------------------- */
532 convert_from_8bit_to_utf_c2 (const char input_char
)
534 unsigned char str
[2];
535 unsigned char buf_ch
[6 + 1];
541 str
[0] = (unsigned char) input_char
;
544 cp_from
= get_codepage_id (mc_global
.source_codepage
);
545 conv
= str_crt_conv_to (cp_from
);
547 if (conv
!= INVALID_CONV
)
549 switch (str_translate_char (conv
, (char *) str
, -1, (char *) buf_ch
, sizeof (buf_ch
)))
552 res
= g_utf8_get_char_validated ((char *) buf_ch
, -1);
567 str_close_conv (conv
);
573 /* --------------------------------------------------------------------------------------------- */
575 #endif /* HAVE_CHARSET */