1 /* GLIB - Library of useful routines for C programming
3 * gconvert.c: Convert between character sets using iconv
4 * Copyright Red Hat Inc., 2000
5 * Authors: Havoc Pennington <hp@redhat.com>, Owen Taylor <otaylor@redhat.com
7 * This library is free software; you can redistribute it and/or
8 * modify it under the terms of the GNU Lesser General Public
9 * License as published by the Free Software Foundation; either
10 * version 2 of the License, or (at your option) any later version.
12 * This library is distributed in the hope that it will be useful,
13 * but WITHOUT ANY WARRANTY; without even the implied warranty of
14 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15 * Lesser General Public License for more details.
17 * You should have received a copy of the GNU Lesser General Public
18 * License along with this library; if not, write to the
19 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
20 * Boston, MA 02111-1307, USA.
38 g_convert_error_quark()
42 quark = g_quark_from_static_string ("g_convert_error");
47 #if defined(USE_LIBICONV) && !defined (_LIBICONV_H)
48 #error libiconv in use but included iconv.h not from libiconv
50 #if !defined(USE_LIBICONV) && defined (_LIBICONV_H)
51 #error libiconv not in use but included iconv.h is from libiconv
55 g_iconv_open (const gchar *to_codeset,
56 const gchar *from_codeset)
58 iconv_t cd = iconv_open (to_codeset, from_codeset);
64 g_iconv (GIConv converter,
68 size_t *outbytes_left)
70 iconv_t cd = (iconv_t)converter;
72 return iconv (cd, inbuf, inbytes_left, outbuf, outbytes_left);
76 g_iconv_close (GIConv converter)
78 iconv_t cd = (iconv_t)converter;
80 return iconv_close (cd);
84 open_converter (const gchar *to_codeset,
85 const gchar *from_codeset,
88 GIConv cd = g_iconv_open (to_codeset, from_codeset);
90 if (cd == (iconv_t) -1)
92 /* Something went wrong. */
94 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_NO_CONVERSION,
95 _("Conversion from character set `%s' to `%s' is not suppo\rted"),
96 from_codeset, to_codeset);
98 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_FAILED,
99 _("Could not open converter from `%s' to `%s': %s"),
100 from_codeset, to_codeset, strerror (errno));
109 * @str: the string to convert
110 * @len: the length of the string
111 * @to_codeset: name of character set into which to convert @str
112 * @from_codeset: character set of @str.
113 * @bytes_read: location to store the number of bytes in the
114 * input string that were successfully converted, or %NULL.
115 * Even if the conversion was succesful, this may be
116 * less than len if there were partial characters
117 * at the end of the input. If the error
118 * G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
119 * stored will the byte fofset after the last valid
121 * @bytes_written: the stored in the output buffer (not including the
123 * @error: location to store the error occuring, or %NULL to ignore
124 * errors. Any of the errors in #GConvertError may occur.
126 * Convert a string from one character set to another.
128 * Return value: If the conversion was successful, a newly allocated
129 * NUL-terminated string, which must be freed with
130 * g_free. Otherwise %NULL and @error will be set.
133 g_convert (const gchar *str,
135 const gchar *to_codeset,
136 const gchar *from_codeset,
144 size_t inbytes_remaining;
145 size_t outbytes_remaining;
149 gboolean have_error = FALSE;
151 g_return_val_if_fail (str != NULL, NULL);
152 g_return_val_if_fail (to_codeset != NULL, NULL);
153 g_return_val_if_fail (from_codeset != NULL, NULL);
155 cd = open_converter (to_codeset, from_codeset, error);
157 if (cd == (GIConv) -1)
172 inbytes_remaining = len;
173 outbuf_size = len + 1; /* + 1 for nul in case len == 1 */
174 outbytes_remaining = outbuf_size - 1; /* -1 for nul */
175 outp = dest = g_malloc (outbuf_size);
179 err = g_iconv (cd, (char **)&p, &inbytes_remaining, &outp, &outbytes_remaining);
181 if (err == (size_t) -1)
186 /* Incomplete text, do not report an error */
190 size_t used = outp - dest;
192 dest = g_realloc (dest, outbuf_size);
195 outbytes_remaining = outbuf_size - used - 1; /* -1 for nul */
200 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
201 _("Invalid byte sequence in conversion input"));
205 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_FAILED,
206 _("Error during conversion: %s"),
218 *bytes_read = p - str;
221 if ((p - str) != len)
223 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,
224 _("Partial character sequence at end of input"));
230 *bytes_written = outp - dest; /* Doesn't include '\0' */
242 * g_convert_with_fallback:
243 * @str: the string to convert
244 * @len: the length of the string
245 * @to_codeset: name of character set into which to convert @str
246 * @from_codeset: character set of @str.
247 * @fallback: UTF-8 string to use in place of character not
248 * present in the target encoding. (This must be
249 * in the target encoding), if %NULL, characters
250 * not in the target encoding will be represented
251 * as Unicode escapes \x{XXXX} or \x{XXXXXX}.
252 * @bytes_read: location to store the number of bytes in the
253 * input string that were successfully converted, or %NULL.
254 * Even if the conversion was succesful, this may be
255 * less than len if there were partial characters
256 * at the end of the input. If the error
257 * G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
258 * stored will the byte fofset after the last valid
260 * @bytes_written: the stored in the output buffer (not including the
262 * @error: location to store the error occuring, or %NULL to ignore
263 * errors. Any of the errors in #GConvertError may occur.
265 * Convert a string from one character set to another, possibly
266 * including fallback sequences for characters not representable
267 * in the output. Note that it is not guaranteed that the specification
268 * for the fallback sequences in @fallback will be honored. Some
269 * systems may do a approximate conversion from @from_codeset
270 * to @to_codeset in their iconv() functions, in which case GLib
271 * will simply return that approximate conversion.
273 * Return value: If the conversion was successful, a newly allocated
274 * NUL-terminated string, which must be freed with
275 * g_free. Otherwise %NULL and @error will be set.
278 g_convert_with_fallback (const gchar *str,
280 const gchar *to_codeset,
281 const gchar *from_codeset,
290 const gchar *insert_str = NULL;
292 int inbytes_remaining;
293 const gchar *save_p = NULL;
294 size_t save_inbytes = 0;
295 size_t outbytes_remaining;
299 gboolean have_error = FALSE;
300 gboolean done = FALSE;
302 GError *local_error = NULL;
304 g_return_val_if_fail (str != NULL, NULL);
305 g_return_val_if_fail (to_codeset != NULL, NULL);
306 g_return_val_if_fail (from_codeset != NULL, NULL);
311 /* Try an exact conversion; we only proceed if this fails
312 * due to an illegal sequence in the input string.
314 dest = g_convert (str, len, to_codeset, from_codeset,
315 bytes_read, bytes_written, &local_error);
319 if (!g_error_matches (local_error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE))
321 g_propagate_error (error, local_error);
325 g_error_free (local_error);
327 /* No go; to proceed, we need a converter from "UTF-8" to
328 * to_codeset, and the string as UTF-8.
330 cd = open_converter (to_codeset, "UTF-8", error);
331 if (cd == (GIConv) -1)
342 utf8 = g_convert (str, len, "UTF-8", from_codeset,
343 bytes_read, &inbytes_remaining, error);
347 /* Now the heart of the code. We loop through the UTF-8 string, and
348 * whenever we hit an offending character, we form fallback, convert
349 * the fallback to the target codeset, and then go back to
350 * converting the original string after finishing with the fallback.
352 * The variables save_p and save_inbytes store the input state
353 * for the original string while we are converting the fallback
356 outbuf_size = len + 1; /* + 1 for nul in case len == 1 */
357 outbytes_remaining = outbuf_size - 1; /* -1 for nul */
358 outp = dest = g_malloc (outbuf_size);
360 while (!done && !have_error)
362 size_t inbytes_tmp = inbytes_remaining;
363 err = g_iconv (cd, (char **)&p, &inbytes_tmp, &outp, &outbytes_remaining);
364 inbytes_remaining = inbytes_tmp;
366 if (err == (size_t) -1)
371 g_assert_not_reached();
375 size_t used = outp - dest;
377 dest = g_realloc (dest, outbuf_size);
380 outbytes_remaining = outbuf_size - used - 1; /* -1 for nul */
387 /* Error converting fallback string - fatal
389 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
390 _("Cannot convert fallback '%s' to codeset '%s'"),
391 insert_str, to_codeset);
399 gunichar ch = g_utf8_get_char (p);
400 insert_str = g_strdup_printf ("\\x{%0*X}",
401 (ch < 0x10000) ? 4 : 6,
405 insert_str = fallback;
407 save_p = g_utf8_next_char (p);
408 save_inbytes = inbytes_remaining - (save_p - p);
410 inbytes_remaining = strlen (p);
414 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_FAILED,
415 _("Error during conversion: %s"),
426 g_free ((gchar *)insert_str);
428 inbytes_remaining = save_inbytes;
443 *bytes_written = outp - str; /* Doesn't include '\0' */
449 if (save_p && !fallback)
450 g_free ((gchar *)insert_str);
461 * Converts a string which is in the encoding used for strings by
462 * the C runtime (usually the same as that used by the operating
463 * system) in the current locale into a UTF-8 string.
467 g_locale_to_utf8 (const gchar *opsysstring, GError **error)
471 gint i, clen, wclen, first;
472 const gint len = strlen (opsysstring);
477 wcs = g_new (wchar_t, len);
478 wclen = MultiByteToWideChar (CP_ACP, 0, opsysstring, len, wcs, len);
482 for (i = 0; i < wclen; i++)
490 else if (wc < 0x10000)
492 else if (wc < 0x200000)
494 else if (wc < 0x4000000)
500 result = g_malloc (clen + 1);
504 for (i = 0; i < wclen; i++)
518 else if (wc < 0x10000)
523 else if (wc < 0x200000)
528 else if (wc < 0x4000000)
542 case 6: bp[5] = (wc & 0x3f) | 0x80; wc >>= 6; /* Fall through */
543 case 5: bp[4] = (wc & 0x3f) | 0x80; wc >>= 6; /* Fall through */
544 case 4: bp[3] = (wc & 0x3f) | 0x80; wc >>= 6; /* Fall through */
545 case 3: bp[2] = (wc & 0x3f) | 0x80; wc >>= 6; /* Fall through */
546 case 2: bp[1] = (wc & 0x3f) | 0x80; wc >>= 6; /* Fall through */
547 case 1: bp[0] = wc | first;
562 if (g_get_charset (&charset))
563 return g_strdup (opsysstring);
565 str = g_convert (opsysstring, strlen (opsysstring),
566 "UTF-8", charset, NULL, NULL, error);
575 * The reverse of g_locale_to_utf8.
579 g_locale_from_utf8 (const gchar *utf8string, GError **error)
583 gint i, mask, clen, mblen;
584 const gint len = strlen (utf8string);
590 /* First convert to wide chars */
591 cp = (guchar *) utf8string;
594 wcs = g_new (wchar_t, len + 1);
606 else if ((c & 0xe0) == 0xc0)
611 else if ((c & 0xf0) == 0xe0)
616 else if ((c & 0xf8) == 0xf0)
621 else if ((c & 0xfc) == 0xf8)
626 else if ((c & 0xfc) == 0xfc)
643 *wcp = (cp[0] & mask);
644 for (i = 1; i < clen; i++)
646 if ((cp[i] & 0xc0) != 0x80)
652 *wcp |= (cp[i] & 0x3f);
665 /* n is the number of wide chars constructed */
667 /* Convert to a string in the current ANSI codepage */
669 result = g_new (gchar, 3 * n + 1);
670 mblen = WideCharToMultiByte (CP_ACP, 0, wcs, n, result, 3*n, NULL, NULL);
678 gchar *charset, *str;
680 if (g_get_charset (&charset))
681 return g_strdup (utf8string);
683 str = g_convert (utf8string, strlen (utf8string),
684 charset, "UTF-8", NULL, NULL, error);
691 /* Filenames are in UTF-8 unless specificially requested otherwise */
694 g_filename_to_utf8 (const gchar *string, GError **error)
698 return g_locale_to_utf8 (string, error);
700 if (getenv ("G_BROKEN_FILENAMES"))
701 return g_locale_to_utf8 (string, error);
703 return g_strdup (string);
708 g_filename_from_utf8 (const gchar *string, GError **error)
711 return g_locale_from_utf8 (string, error);
713 if (getenv ("G_BROKEN_FILENAMES"))
714 return g_locale_from_utf8 (string, error);
716 return g_strdup (string);