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.
31 #ifdef G_PLATFORM_WIN32
40 g_convert_error_quark()
44 quark
= g_quark_from_static_string ("g_convert_error");
49 #if defined(USE_LIBICONV) && !defined (_LIBICONV_H)
50 #error libiconv in use but included iconv.h not from libiconv
52 #if !defined(USE_LIBICONV) && defined (_LIBICONV_H)
53 #error libiconv not in use but included iconv.h is from libiconv
57 try_conversion (const char *to_codeset
,
58 const char *from_codeset
,
61 *cd
= iconv_open (to_codeset
, from_codeset
);
63 if (*cd
== (iconv_t
)-1 && errno
== EINVAL
)
70 try_to_aliases (const char **to_aliases
,
71 const char *from_codeset
,
76 const char **p
= to_aliases
;
79 if (try_conversion (*p
, from_codeset
, cd
))
89 extern const char **_g_charset_get_aliases (const char *canonical_name
);
93 * @to_codeset: destination codeset
94 * @from_codeset: source codeset
96 * Same as the standard UNIX routine <function>iconv_open()</function>, but
97 * may be implemented via libiconv on UNIX flavors that lack
98 * a native implementation.
100 * GLib provides g_convert() and g_locale_to_utf8() which are likely
101 * more convenient than the raw iconv wrappers.
103 * Return value: a "conversion descriptor"
106 g_iconv_open (const gchar
*to_codeset
,
107 const gchar
*from_codeset
)
111 if (!try_conversion (to_codeset
, from_codeset
, &cd
))
113 const char **to_aliases
= _g_charset_get_aliases (to_codeset
);
114 const char **from_aliases
= _g_charset_get_aliases (from_codeset
);
118 const char **p
= from_aliases
;
121 if (try_conversion (to_codeset
, *p
, &cd
))
124 if (try_to_aliases (to_aliases
, *p
, &cd
))
131 if (try_to_aliases (to_aliases
, from_codeset
, &cd
))
140 * @converter: conversion descriptor from g_iconv_open()
141 * @inbuf: bytes to convert
142 * @inbytes_left: inout parameter, bytes remaining to convert in @inbuf
143 * @outbuf: converted output bytes
144 * @outbytes_left: inout parameter, bytes available to fill in @outbuf
146 * Same as the standard UNIX routine <function>iconv()</function>, but
147 * may be implemented via libiconv on UNIX flavors that lack
148 * a native implementation.
150 * GLib provides g_convert() and g_locale_to_utf8() which are likely
151 * more convenient than the raw iconv wrappers.
153 * Return value: count of non-reversible conversions, or -1 on error
156 g_iconv (GIConv converter
,
160 gsize
*outbytes_left
)
162 iconv_t cd
= (iconv_t
)converter
;
164 return iconv (cd
, inbuf
, inbytes_left
, outbuf
, outbytes_left
);
169 * @converter: a conversion descriptor from g_iconv_open()
171 * Same as the standard UNIX routine <function>iconv_close()</function>, but
172 * may be implemented via libiconv on UNIX flavors that lack
173 * a native implementation. Should be called to clean up
174 * the conversion descriptor from g_iconv_open() when
175 * you are done converting things.
177 * GLib provides g_convert() and g_locale_to_utf8() which are likely
178 * more convenient than the raw iconv wrappers.
180 * Return value: -1 on error, 0 on success
183 g_iconv_close (GIConv converter
)
185 iconv_t cd
= (iconv_t
)converter
;
187 return iconv_close (cd
);
191 open_converter (const gchar
*to_codeset
,
192 const gchar
*from_codeset
,
195 GIConv cd
= g_iconv_open (to_codeset
, from_codeset
);
197 if (cd
== (iconv_t
) -1)
199 /* Something went wrong. */
201 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_NO_CONVERSION
,
202 _("Conversion from character set '%s' to '%s' is not supported"),
203 from_codeset
, to_codeset
);
205 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_FAILED
,
206 _("Could not open converter from '%s' to '%s': %s"),
207 from_codeset
, to_codeset
, strerror (errno
));
216 * @str: the string to convert
217 * @len: the length of the string
218 * @to_codeset: name of character set into which to convert @str
219 * @from_codeset: character set of @str.
220 * @bytes_read: location to store the number of bytes in the
221 * input string that were successfully converted, or %NULL.
222 * Even if the conversion was successful, this may be
223 * less than @len if there were partial characters
224 * at the end of the input. If the error
225 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
226 * stored will the byte offset after the last valid
228 * @bytes_written: the number of bytes stored in the output buffer (not
229 * including the terminating nul).
230 * @error: location to store the error occuring, or %NULL to ignore
231 * errors. Any of the errors in #GConvertError may occur.
233 * Converts a string from one character set to another.
235 * Return value: If the conversion was successful, a newly allocated
236 * nul-terminated string, which must be freed with
237 * g_free(). Otherwise %NULL and @error will be set.
240 g_convert (const gchar
*str
,
242 const gchar
*to_codeset
,
243 const gchar
*from_codeset
,
245 gsize
*bytes_written
,
251 g_return_val_if_fail (str
!= NULL
, NULL
);
252 g_return_val_if_fail (to_codeset
!= NULL
, NULL
);
253 g_return_val_if_fail (from_codeset
!= NULL
, NULL
);
255 cd
= open_converter (to_codeset
, from_codeset
, error
);
257 if (cd
== (GIConv
) -1)
268 res
= g_convert_with_iconv (str
, len
, cd
,
269 bytes_read
, bytes_written
,
278 * g_convert_with_iconv:
279 * @str: the string to convert
280 * @len: the length of the string
281 * @converter: conversion descriptor from g_iconv_open()
282 * @bytes_read: location to store the number of bytes in the
283 * input string that were successfully converted, or %NULL.
284 * Even if the conversion was successful, this may be
285 * less than @len if there were partial characters
286 * at the end of the input. If the error
287 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
288 * stored will the byte offset after the last valid
290 * @bytes_written: the number of bytes stored in the output buffer (not
291 * including the terminating nul).
292 * @error: location to store the error occuring, or %NULL to ignore
293 * errors. Any of the errors in #GConvertError may occur.
295 * Converts a string from one character set to another.
297 * Return value: If the conversion was successful, a newly allocated
298 * nul-terminated string, which must be freed with
299 * g_free(). Otherwise %NULL and @error will be set.
302 g_convert_with_iconv (const gchar
*str
,
306 gsize
*bytes_written
,
312 gsize inbytes_remaining
;
313 gsize outbytes_remaining
;
316 gboolean have_error
= FALSE
;
318 g_return_val_if_fail (str
!= NULL
, NULL
);
319 g_return_val_if_fail (converter
!= (GIConv
) -1, NULL
);
325 inbytes_remaining
= len
;
326 outbuf_size
= len
+ 1; /* + 1 for nul in case len == 1 */
328 outbytes_remaining
= outbuf_size
- 1; /* -1 for nul */
329 outp
= dest
= g_malloc (outbuf_size
);
333 err
= g_iconv (converter
, (char **)&p
, &inbytes_remaining
, &outp
, &outbytes_remaining
);
335 if (err
== (size_t) -1)
340 /* Incomplete text, do not report an error */
344 size_t used
= outp
- dest
;
347 dest
= g_realloc (dest
, outbuf_size
);
350 outbytes_remaining
= outbuf_size
- used
- 1; /* -1 for nul */
355 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_ILLEGAL_SEQUENCE
,
356 _("Invalid byte sequence in conversion input"));
360 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_FAILED
,
361 _("Error during conversion: %s"),
371 *bytes_read
= p
- str
;
374 if ((p
- str
) != len
)
378 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_PARTIAL_INPUT
,
379 _("Partial character sequence at end of input"));
386 *bytes_written
= outp
- dest
; /* Doesn't include '\0' */
398 * g_convert_with_fallback:
399 * @str: the string to convert
400 * @len: the length of the string
401 * @to_codeset: name of character set into which to convert @str
402 * @from_codeset: character set of @str.
403 * @fallback: UTF-8 string to use in place of character not
404 * present in the target encoding. (This must be
405 * in the target encoding), if %NULL, characters
406 * not in the target encoding will be represented
407 * as Unicode escapes \x{XXXX} or \x{XXXXXX}.
408 * @bytes_read: location to store the number of bytes in the
409 * input string that were successfully converted, or %NULL.
410 * Even if the conversion was successful, this may be
411 * less than @len if there were partial characters
412 * at the end of the input.
413 * @bytes_written: the number of bytes stored in the output buffer (not
414 * including the terminating nul).
415 * @error: location to store the error occuring, or %NULL to ignore
416 * errors. Any of the errors in #GConvertError may occur.
418 * Converts a string from one character set to another, possibly
419 * including fallback sequences for characters not representable
420 * in the output. Note that it is not guaranteed that the specification
421 * for the fallback sequences in @fallback will be honored. Some
422 * systems may do a approximate conversion from @from_codeset
423 * to @to_codeset in their <function>iconv()</function> functions,
424 * in which case GLib will simply return that approximate conversion.
426 * Return value: If the conversion was successful, a newly allocated
427 * nul-terminated string, which must be freed with
428 * g_free(). Otherwise %NULL and @error will be set.
431 g_convert_with_fallback (const gchar
*str
,
433 const gchar
*to_codeset
,
434 const gchar
*from_codeset
,
437 gsize
*bytes_written
,
443 const gchar
*insert_str
= NULL
;
445 gsize inbytes_remaining
;
446 const gchar
*save_p
= NULL
;
447 gsize save_inbytes
= 0;
448 gsize outbytes_remaining
;
452 gboolean have_error
= FALSE
;
453 gboolean done
= FALSE
;
455 GError
*local_error
= NULL
;
457 g_return_val_if_fail (str
!= NULL
, NULL
);
458 g_return_val_if_fail (to_codeset
!= NULL
, NULL
);
459 g_return_val_if_fail (from_codeset
!= NULL
, NULL
);
464 /* Try an exact conversion; we only proceed if this fails
465 * due to an illegal sequence in the input string.
467 dest
= g_convert (str
, len
, to_codeset
, from_codeset
,
468 bytes_read
, bytes_written
, &local_error
);
472 if (!g_error_matches (local_error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_ILLEGAL_SEQUENCE
))
474 g_propagate_error (error
, local_error
);
478 g_error_free (local_error
);
482 /* No go; to proceed, we need a converter from "UTF-8" to
483 * to_codeset, and the string as UTF-8.
485 cd
= open_converter (to_codeset
, "UTF-8", error
);
486 if (cd
== (GIConv
) -1)
497 utf8
= g_convert (str
, len
, "UTF-8", from_codeset
,
498 bytes_read
, &inbytes_remaining
, error
);
502 /* Now the heart of the code. We loop through the UTF-8 string, and
503 * whenever we hit an offending character, we form fallback, convert
504 * the fallback to the target codeset, and then go back to
505 * converting the original string after finishing with the fallback.
507 * The variables save_p and save_inbytes store the input state
508 * for the original string while we are converting the fallback
512 outbuf_size
= len
+ 1; /* + 1 for nul in case len == 1 */
513 outbytes_remaining
= outbuf_size
- 1; /* -1 for nul */
514 outp
= dest
= g_malloc (outbuf_size
);
516 while (!done
&& !have_error
)
518 size_t inbytes_tmp
= inbytes_remaining
;
519 err
= g_iconv (cd
, (char **)&p
, &inbytes_tmp
, &outp
, &outbytes_remaining
);
520 inbytes_remaining
= inbytes_tmp
;
522 if (err
== (size_t) -1)
527 g_assert_not_reached();
531 size_t used
= outp
- dest
;
534 dest
= g_realloc (dest
, outbuf_size
);
537 outbytes_remaining
= outbuf_size
- used
- 1; /* -1 for nul */
544 /* Error converting fallback string - fatal
546 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_ILLEGAL_SEQUENCE
,
547 _("Cannot convert fallback '%s' to codeset '%s'"),
548 insert_str
, to_codeset
);
556 gunichar ch
= g_utf8_get_char (p
);
557 insert_str
= g_strdup_printf ("\\x{%0*X}",
558 (ch
< 0x10000) ? 4 : 6,
562 insert_str
= fallback
;
564 save_p
= g_utf8_next_char (p
);
565 save_inbytes
= inbytes_remaining
- (save_p
- p
);
567 inbytes_remaining
= strlen (p
);
571 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_FAILED
,
572 _("Error during conversion: %s"),
583 g_free ((gchar
*)insert_str
);
585 inbytes_remaining
= save_inbytes
;
600 *bytes_written
= outp
- str
; /* Doesn't include '\0' */
606 if (save_p
&& !fallback
)
607 g_free ((gchar
*)insert_str
);
622 strdup_len (const gchar
*string
,
624 gsize
*bytes_written
,
631 if (!g_utf8_validate (string
, -1, NULL
))
638 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_ILLEGAL_SEQUENCE
,
639 _("Invalid byte sequence in conversion input"));
644 real_len
= strlen (string
);
649 while (real_len
< len
&& string
[real_len
])
654 *bytes_read
= real_len
;
656 *bytes_written
= real_len
;
658 return g_strndup (string
, real_len
);
663 * @opsysstring: a string in the encoding of the current locale
664 * @len: the length of the string, or -1 if the string is
666 * @bytes_read: location to store the number of bytes in the
667 * input string that were successfully converted, or %NULL.
668 * Even if the conversion was successful, this may be
669 * less than @len if there were partial characters
670 * at the end of the input. If the error
671 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
672 * stored will the byte offset after the last valid
674 * @bytes_written: the number of bytes stored in the output buffer (not
675 * including the terminating nul).
676 * @error: location to store the error occuring, or %NULL to ignore
677 * errors. Any of the errors in #GConvertError may occur.
679 * Converts a string which is in the encoding used for strings by
680 * the C runtime (usually the same as that used by the operating
681 * system) in the current locale into a UTF-8 string.
683 * Return value: The converted string, or %NULL on an error.
686 g_locale_to_utf8 (const gchar
*opsysstring
,
689 gsize
*bytes_written
,
692 #ifdef G_PLATFORM_WIN32
694 gint i
, clen
, total_len
, wclen
, first
;
700 len
= strlen (opsysstring
);
702 wcs
= g_new (wchar_t, len
);
703 wclen
= MultiByteToWideChar (CP_ACP
, 0, opsysstring
, len
, wcs
, len
);
707 for (i
= 0; i
< wclen
; i
++)
715 else if (wc
< 0x10000)
717 else if (wc
< 0x200000)
719 else if (wc
< 0x4000000)
725 result
= g_malloc (total_len
+ 1);
729 for (i
= 0; i
< wclen
; i
++)
743 else if (wc
< 0x10000)
748 else if (wc
< 0x200000)
753 else if (wc
< 0x4000000)
767 case 6: bp
[5] = (wc
& 0x3f) | 0x80; wc
>>= 6; /* Fall through */
768 case 5: bp
[4] = (wc
& 0x3f) | 0x80; wc
>>= 6; /* Fall through */
769 case 4: bp
[3] = (wc
& 0x3f) | 0x80; wc
>>= 6; /* Fall through */
770 case 3: bp
[2] = (wc
& 0x3f) | 0x80; wc
>>= 6; /* Fall through */
771 case 2: bp
[1] = (wc
& 0x3f) | 0x80; wc
>>= 6; /* Fall through */
772 case 1: bp
[0] = wc
| first
;
784 *bytes_written
= total_len
;
788 #else /* !G_PLATFORM_WIN32 */
792 if (g_get_charset (&charset
))
793 return strdup_len (opsysstring
, len
, bytes_read
, bytes_written
, error
);
795 return g_convert (opsysstring
, len
,
796 "UTF-8", charset
, bytes_read
, bytes_written
, error
);
798 #endif /* !G_PLATFORM_WIN32 */
802 * g_locale_from_utf8:
803 * @utf8string: a UTF-8 encoded string
804 * @len: the length of the string, or -1 if the string is
806 * @bytes_read: location to store the number of bytes in the
807 * input string that were successfully converted, or %NULL.
808 * Even if the conversion was successful, this may be
809 * less than @len if there were partial characters
810 * at the end of the input. If the error
811 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
812 * stored will the byte offset after the last valid
814 * @bytes_written: the number of bytes stored in the output buffer (not
815 * including the terminating nul).
816 * @error: location to store the error occuring, or %NULL to ignore
817 * errors. Any of the errors in #GConvertError may occur.
819 * Converts a string from UTF-8 to the encoding used for strings by
820 * the C runtime (usually the same as that used by the operating
821 * system) in the current locale.
823 * Return value: The converted string, or %NULL on an error.
826 g_locale_from_utf8 (const gchar
*utf8string
,
829 gsize
*bytes_written
,
832 #ifdef G_PLATFORM_WIN32
834 gint i
, mask
, clen
, mblen
;
841 len
= strlen (utf8string
);
843 /* First convert to wide chars */
844 cp
= (guchar
*) utf8string
;
847 wcs
= g_new (wchar_t, len
+ 1);
859 else if ((c
& 0xe0) == 0xc0)
864 else if ((c
& 0xf0) == 0xe0)
869 else if ((c
& 0xf8) == 0xf0)
874 else if ((c
& 0xfc) == 0xf8)
879 else if ((c
& 0xfc) == 0xfc)
896 *wcp
= (cp
[0] & mask
);
897 for (i
= 1; i
< clen
; i
++)
899 if ((cp
[i
] & 0xc0) != 0x80)
905 *wcp
|= (cp
[i
] & 0x3f);
918 /* n is the number of wide chars constructed */
920 /* Convert to a string in the current ANSI codepage */
922 result
= g_new (gchar
, 3 * n
+ 1);
923 mblen
= WideCharToMultiByte (CP_ACP
, 0, wcs
, n
, result
, 3*n
, NULL
, NULL
);
930 *bytes_written
= mblen
;
934 #else /* !G_PLATFORM_WIN32 */
936 const gchar
*charset
;
938 if (g_get_charset (&charset
))
939 return strdup_len (utf8string
, len
, bytes_read
, bytes_written
, error
);
941 return g_convert (utf8string
, len
,
942 charset
, "UTF-8", bytes_read
, bytes_written
, error
);
944 #endif /* !G_PLATFORM_WIN32 */
948 * g_filename_to_utf8:
949 * @opsysstring: a string in the encoding for filenames
950 * @len: the length of the string, or -1 if the string is
952 * @bytes_read: location to store the number of bytes in the
953 * input string that were successfully converted, or %NULL.
954 * Even if the conversion was successful, this may be
955 * less than @len if there were partial characters
956 * at the end of the input. If the error
957 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
958 * stored will the byte offset after the last valid
960 * @bytes_written: the number of bytes stored in the output buffer (not
961 * including the terminating nul).
962 * @error: location to store the error occuring, or %NULL to ignore
963 * errors. Any of the errors in #GConvertError may occur.
965 * Converts a string which is in the encoding used for filenames
966 * into a UTF-8 string.
968 * Return value: The converted string, or %NULL on an error.
971 g_filename_to_utf8 (const gchar
*opsysstring
,
974 gsize
*bytes_written
,
977 #ifdef G_PLATFORM_WIN32
978 return g_locale_to_utf8 (opsysstring
, len
,
979 bytes_read
, bytes_written
,
981 #else /* !G_PLATFORM_WIN32 */
983 if (getenv ("G_BROKEN_FILENAMES"))
984 return g_locale_to_utf8 (opsysstring
, len
,
985 bytes_read
, bytes_written
,
988 return strdup_len (opsysstring
, len
, bytes_read
, bytes_written
, error
);
989 #endif /* !G_PLATFORM_WIN32 */
993 * g_filename_from_utf8:
994 * @utf8string: a UTF-8 encoded string.
995 * @len: the length of the string, or -1 if the string is
997 * @bytes_read: location to store the number of bytes in the
998 * input string that were successfully converted, or %NULL.
999 * Even if the conversion was successful, this may be
1000 * less than @len if there were partial characters
1001 * at the end of the input. If the error
1002 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
1003 * stored will the byte offset after the last valid
1005 * @bytes_written: the number of bytes stored in the output buffer (not
1006 * including the terminating nul).
1007 * @error: location to store the error occuring, or %NULL to ignore
1008 * errors. Any of the errors in #GConvertError may occur.
1010 * Converts a string from UTF-8 to the encoding used for filenames.
1012 * Return value: The converted string, or %NULL on an error.
1015 g_filename_from_utf8 (const gchar
*utf8string
,
1018 gsize
*bytes_written
,
1021 #ifdef G_PLATFORM_WIN32
1022 return g_locale_from_utf8 (utf8string
, len
,
1023 bytes_read
, bytes_written
,
1025 #else /* !G_PLATFORM_WIN32 */
1026 if (getenv ("G_BROKEN_FILENAMES"))
1027 return g_locale_from_utf8 (utf8string
, len
,
1028 bytes_read
, bytes_written
,
1031 return strdup_len (utf8string
, len
, bytes_read
, bytes_written
, error
);
1032 #endif /* !G_PLATFORM_WIN32 */
1035 /* Test of haystack has the needle prefix, comparing case
1036 * insensitive. haystack may be UTF-8, but needle must
1037 * contain only ascii. */
1039 has_case_prefix (const gchar
*haystack
, const gchar
*needle
)
1043 /* Eat one character at a time. */
1048 g_ascii_tolower (*n
) == g_ascii_tolower (*h
))
1058 UNSAFE_ALL
= 0x1, /* Escape all unsafe characters */
1059 UNSAFE_ALLOW_PLUS
= 0x2, /* Allows '+' */
1060 UNSAFE_PATH
= 0x4, /* Allows '/' and '?' and '&' and '=' */
1061 UNSAFE_DOS_PATH
= 0x8, /* Allows '/' and '?' and '&' and '=' and ':' */
1062 UNSAFE_HOST
= 0x10, /* Allows '/' and ':' and '@' */
1063 UNSAFE_SLASHES
= 0x20 /* Allows all characters except for '/' and '%' */
1064 } UnsafeCharacterSet
;
1066 static const guchar acceptable
[96] = {
1067 /* X0 X1 X2 X3 X4 X5 X6 X7 X8 X9 XA XB XC XD XE XF */
1068 0x00,0x3F,0x20,0x20,0x20,0x00,0x2C,0x3F,0x3F,0x3F,0x3F,0x22,0x20,0x3F,0x3F,0x1C, /* 2X !"#$%&'()*+,-./ */
1069 0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x38,0x20,0x20,0x2C,0x20,0x2C, /* 3X 0123456789:;<=>? */
1070 0x30,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F, /* 4X @ABCDEFGHIJKLMNO */
1071 0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x20,0x20,0x20,0x20,0x3F, /* 5X PQRSTUVWXYZ[\]^_ */
1072 0x20,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F, /* 6X `abcdefghijklmno */
1073 0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x20,0x20,0x20,0x3F,0x20 /* 7X pqrstuvwxyz{|}~DEL */
1076 static const gchar hex
[16] = "0123456789ABCDEF";
1078 /* Note: This escape function works on file: URIs, but if you want to
1079 * escape something else, please read RFC-2396 */
1081 g_escape_uri_string (const gchar
*string
,
1082 UnsafeCharacterSet mask
)
1084 #define ACCEPTABLE(a) ((a)>=32 && (a)<128 && (acceptable[(a)-32] & use_mask))
1091 UnsafeCharacterSet use_mask
;
1093 g_return_val_if_fail (mask
== UNSAFE_ALL
1094 || mask
== UNSAFE_ALLOW_PLUS
1095 || mask
== UNSAFE_PATH
1096 || mask
== UNSAFE_DOS_PATH
1097 || mask
== UNSAFE_HOST
1098 || mask
== UNSAFE_SLASHES
, NULL
);
1102 for (p
= string
; *p
!= '\0'; p
++)
1105 if (!ACCEPTABLE (c
))
1109 result
= g_malloc (p
- string
+ unacceptable
* 2 + 1);
1112 for (q
= result
, p
= string
; *p
!= '\0'; p
++)
1114 c
= (unsigned char)*p
;
1116 if (!ACCEPTABLE (c
))
1118 *q
++ = '%'; /* means hex coming */
1133 g_escape_file_uri (const gchar
*hostname
,
1134 const gchar
*pathname
)
1136 char *escaped_hostname
= NULL
;
1140 if (hostname
&& *hostname
!= '\0')
1142 escaped_hostname
= g_escape_uri_string (hostname
, UNSAFE_HOST
);
1145 escaped_path
= g_escape_uri_string (pathname
, UNSAFE_DOS_PATH
);
1147 res
= g_strconcat ("file://",
1148 (escaped_hostname
) ? escaped_hostname
: "",
1149 (*escaped_path
!= '/') ? "/" : "",
1153 g_free (escaped_hostname
);
1154 g_free (escaped_path
);
1160 unescape_character (const char *scanner
)
1165 first_digit
= g_ascii_xdigit_value (*scanner
++);
1167 if (first_digit
< 0)
1170 second_digit
= g_ascii_xdigit_value (*scanner
++);
1171 if (second_digit
< 0)
1174 return (first_digit
<< 4) | second_digit
;
1178 g_unescape_uri_string (const gchar
*escaped
,
1179 const gchar
*illegal_characters
,
1182 const gchar
*in
, *in_end
;
1183 gchar
*out
, *result
;
1186 if (escaped
== NULL
)
1190 len
= strlen (escaped
);
1192 result
= g_malloc (len
+ 1);
1195 for (in
= escaped
, in_end
= escaped
+ len
; in
< in_end
&& *in
!= '\0'; in
++)
1198 if (character
== '%')
1200 character
= unescape_character (in
+ 1);
1202 /* Check for an illegal character. We consider '\0' illegal here. */
1204 || (illegal_characters
!= NULL
1205 && strchr (illegal_characters
, (char)character
) != NULL
))
1217 g_assert (out
- result
<= strlen (escaped
));
1219 if (!g_utf8_validate (result
, -1, NULL
))
1229 * g_filename_from_uri:
1230 * @uri: a uri describing a filename (escaped, encoded in UTF-8).
1231 * @hostname: Location to store hostname for the URI, or %NULL.
1232 * If there is no hostname in the URI, %NULL will be
1233 * stored in this location.
1234 * @error: location to store the error occuring, or %NULL to ignore
1235 * errors. Any of the errors in #GConvertError may occur.
1237 * Converts an escaped UTF-8 encoded URI to a local filename in the
1238 * encoding used for filenames.
1240 * Return value: a newly-allocated string holding the resulting
1241 * filename, or %NULL on an error.
1244 g_filename_from_uri (const char *uri
,
1248 const char *path_part
;
1249 const char *host_part
;
1250 char *unescaped_hostname
;
1258 if (!has_case_prefix (uri
, "file:/"))
1260 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_BAD_URI
,
1261 _("The URI '%s' is not an absolute URI using the file scheme"),
1266 path_part
= uri
+ strlen ("file:");
1268 if (strchr (path_part
, '#') != NULL
)
1270 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_BAD_URI
,
1271 _("The local file URI '%s' may not include a '#'"),
1276 if (has_case_prefix (path_part
, "///"))
1278 else if (has_case_prefix (path_part
, "//"))
1281 host_part
= path_part
;
1283 path_part
= strchr (path_part
, '/');
1285 if (path_part
== NULL
)
1287 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_BAD_URI
,
1288 _("The URI '%s' is invalid"),
1293 unescaped_hostname
= g_unescape_uri_string (host_part
, "", path_part
- host_part
);
1294 if (unescaped_hostname
== NULL
)
1296 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_BAD_URI
,
1297 _("The hostname of the URI '%s' contains invalidly escaped characters"),
1303 *hostname
= unescaped_hostname
;
1305 g_free (unescaped_hostname
);
1308 filename
= g_unescape_uri_string (path_part
, "/", -1);
1310 if (filename
== NULL
)
1312 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_BAD_URI
,
1313 _("The URI '%s' contains invalidly escaped characters"),
1318 /* DOS uri's are like "file://host/c:\foo", so we need to check if we need to
1319 * drop the initial slash */
1321 if (g_path_is_absolute (filename
+1))
1324 result
= g_filename_from_utf8 (filename
+ offs
, -1, NULL
, NULL
, error
);
1331 * g_filename_to_uri:
1332 * @filename: an absolute filename specified in the encoding
1333 * used for filenames by the operating system.
1334 * @hostname: A UTF-8 encoded hostname, or %NULL for none.
1335 * @error: location to store the error occuring, or %NULL to ignore
1336 * errors. Any of the errors in #GConvertError may occur.
1338 * Converts an absolute filename to an escaped UTF-8 encoded URI.
1340 * Return value: a newly-allocated string holding the resulting
1341 * URI, or %NULL on an error.
1344 g_filename_to_uri (const char *filename
,
1349 char *utf8_filename
;
1351 g_return_val_if_fail (filename
!= NULL
, NULL
);
1353 if (!g_path_is_absolute (filename
))
1355 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_NOT_ABSOLUTE_PATH
,
1356 _("The pathname '%s' is not an absolute path"),
1361 utf8_filename
= g_filename_to_utf8 (filename
, -1, NULL
, NULL
, error
);
1362 if (utf8_filename
== NULL
)
1366 !g_utf8_validate (hostname
, -1, NULL
))
1368 g_free (utf8_filename
);
1369 g_set_error (error
, G_CONVERT_ERROR
, G_CONVERT_ERROR_ILLEGAL_SEQUENCE
,
1370 _("Invalid byte sequence in hostname"));
1374 escaped_uri
= g_escape_file_uri (hostname
,
1376 g_free (utf8_filename
);