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, see <http://www.gnu.org/licenses/>.
22 #include "glibconfig.h"
33 #include "win_iconv.c"
36 #ifdef G_PLATFORM_WIN32
44 #include "gcharsetprivate.h"
46 #include "gstrfuncs.h"
47 #include "gtestutils.h"
50 #include "gfileutils.h"
54 #if defined(USE_LIBICONV_GNU) && !defined (_LIBICONV_H)
55 #error GNU libiconv in use but included iconv.h not from libiconv
57 #if !defined(USE_LIBICONV_GNU) && defined (_LIBICONV_H) \
58 && !defined (__APPLE_CC__) && !defined (__LP_64__)
59 #error GNU libiconv not in use but included iconv.h is from libiconv
65 * @title: Character Set Conversion
66 * @short_description: convert strings between different character sets
68 * The g_convert() family of function wraps the functionality of iconv().
69 * In addition to pure character set conversions, GLib has functions to
70 * deal with the extra complications of encodings for file names.
72 * ## File Name Encodings
74 * Historically, UNIX has not had a defined encoding for file names:
75 * a file name is valid as long as it does not have path separators
76 * in it ("/"). However, displaying file names may require conversion:
77 * from the character set in which they were created, to the character
78 * set in which the application operates. Consider the Spanish file name
79 * "Presentación.sxi". If the application which created it uses
80 * ISO-8859-1 for its encoding,
82 * Character: P r e s e n t a c i ó n . s x i
83 * Hex code: 50 72 65 73 65 6e 74 61 63 69 f3 6e 2e 73 78 69
85 * However, if the application use UTF-8, the actual file name on
86 * disk would look like this:
87 * <programlisting id="filename-utf-8">
88 * Character: P r e s e n t a c i ó n . s x i
89 * Hex code: 50 72 65 73 65 6e 74 61 63 69 c3 b3 6e 2e 73 78 69
91 * Glib uses UTF-8 for its strings, and GUI toolkits like GTK+ that use
92 * Glib do the same thing. If you get a file name from the file system,
93 * for example, from readdir() or from g_dir_read_name(), and you wish
94 * to display the file name to the user, you will need to convert it
95 * into UTF-8. The opposite case is when the user types the name of a
96 * file he wishes to save: the toolkit will give you that string in
97 * UTF-8 encoding, and you will need to convert it to the character
98 * set used for file names before you can create the file with open()
101 * By default, Glib assumes that file names on disk are in UTF-8
102 * encoding. This is a valid assumption for file systems which
103 * were created relatively recently: most applications use UTF-8
104 * encoding for their strings, and that is also what they use for
105 * the file names they create. However, older file systems may
106 * still contain file names created in "older" encodings, such as
107 * ISO-8859-1. In this case, for compatibility reasons, you may
108 * want to instruct Glib to use that particular encoding for file
109 * names rather than UTF-8. You can do this by specifying the
110 * encoding for file names in the <link
111 * linkend="G_FILENAME_ENCODING"><envar>G_FILENAME_ENCODING</envar></link>
112 * environment variable. For example, if your installation uses
113 * ISO-8859-1 for file names, you can put this in your
114 * <filename>~/.profile</filename>:
116 * export G_FILENAME_ENCODING=ISO-8859-1
118 * Glib provides the functions g_filename_to_utf8() and
119 * g_filename_from_utf8() to perform the necessary conversions.
120 * These functions convert file names from the encoding specified
121 * in <envar>G_FILENAME_ENCODING</envar> to UTF-8 and vice-versa.
122 * <xref linkend="file-name-encodings-diagram"/> illustrates how
123 * these functions are used to convert between UTF-8 and the
124 * encoding for file names in the file system.
126 * <figure id="file-name-encodings-diagram">
127 * <title>Conversion between File Name Encodings</title>
128 * <graphic fileref="file-name-encodings.png" format="PNG"/>
131 * ## Checklist for Application Writers
133 * This section is a practical summary of the detailed
135 * things to do to make sure your applications process file
136 * name encodings correctly.
138 * 1. If you get a file name from the file system from a function
139 * such as readdir() or gtk_file_chooser_get_filename(), you do
140 * not need to do any conversion to pass that file name to
141 * functions like open(), rename(), or fopen() -- those are "raw"
142 * file names which the file system understands.
144 * 2. If you need to display a file name, convert it to UTF-8 first
145 * by using g_filename_to_utf8(). If conversion fails, display a
146 * string like "Unknown file name". Do not convert this string back
147 * into the encoding used for file names if you wish to pass it to
148 * the file system; use the original file name instead.
150 * For example, the document window of a word processor could display
151 * "Unknown file name" in its title bar but still let the user save
152 * the file, as it would keep the raw file name internally. This can
153 * happen if the user has not set the <envar>G_FILENAME_ENCODING</envar>
154 * environment variable even though he has files whose names are not
157 * 3. If your user interface lets the user type a file name for saving
158 * or renaming, convert it to the encoding used for file names in
159 * the file system by using g_filename_from_utf8(). Pass the converted
160 * file name to functions like fopen(). If conversion fails, ask the
161 * user to enter a different file name. This can happen if the user
162 * types Japanese characters when <envar>G_FILENAME_ENCODING</envar>
163 * is set to <literal>ISO-8859-1</literal>, for example.
166 /* We try to terminate strings in unknown charsets with this many zero bytes
167 * to ensure that multibyte strings really are nul-terminated when we return
168 * them from g_convert() and friends.
170 #define NUL_TERMINATOR_LENGTH 4
172 G_DEFINE_QUARK (g_convert_error, g_convert_error)
175 try_conversion (const char *to_codeset,
176 const char *from_codeset,
179 *cd = iconv_open (to_codeset, from_codeset);
181 if (*cd == (iconv_t)-1 && errno == EINVAL)
188 try_to_aliases (const char **to_aliases,
189 const char *from_codeset,
194 const char **p = to_aliases;
197 if (try_conversion (*p, from_codeset, cd))
209 * @to_codeset: destination codeset
210 * @from_codeset: source codeset
212 * Same as the standard UNIX routine iconv_open(), but
213 * may be implemented via libiconv on UNIX flavors that lack
214 * a native implementation.
216 * GLib provides g_convert() and g_locale_to_utf8() which are likely
217 * more convenient than the raw iconv wrappers.
219 * Return value: a "conversion descriptor", or (GIConv)-1 if
220 * opening the converter failed.
223 g_iconv_open (const gchar *to_codeset,
224 const gchar *from_codeset)
228 if (!try_conversion (to_codeset, from_codeset, &cd))
230 const char **to_aliases = _g_charset_get_aliases (to_codeset);
231 const char **from_aliases = _g_charset_get_aliases (from_codeset);
235 const char **p = from_aliases;
238 if (try_conversion (to_codeset, *p, &cd))
241 if (try_to_aliases (to_aliases, *p, &cd))
248 if (try_to_aliases (to_aliases, from_codeset, &cd))
253 return (cd == (iconv_t)-1) ? (GIConv)-1 : (GIConv)cd;
258 * @converter: conversion descriptor from g_iconv_open()
259 * @inbuf: bytes to convert
260 * @inbytes_left: inout parameter, bytes remaining to convert in @inbuf
261 * @outbuf: converted output bytes
262 * @outbytes_left: inout parameter, bytes available to fill in @outbuf
264 * Same as the standard UNIX routine iconv(), but
265 * may be implemented via libiconv on UNIX flavors that lack
266 * a native implementation.
268 * GLib provides g_convert() and g_locale_to_utf8() which are likely
269 * more convenient than the raw iconv wrappers.
271 * Return value: count of non-reversible conversions, or -1 on error
274 g_iconv (GIConv converter,
278 gsize *outbytes_left)
280 iconv_t cd = (iconv_t)converter;
282 return iconv (cd, inbuf, inbytes_left, outbuf, outbytes_left);
287 * @converter: a conversion descriptor from g_iconv_open()
289 * Same as the standard UNIX routine iconv_close(), but
290 * may be implemented via libiconv on UNIX flavors that lack
291 * a native implementation. Should be called to clean up
292 * the conversion descriptor from g_iconv_open() when
293 * you are done converting things.
295 * GLib provides g_convert() and g_locale_to_utf8() which are likely
296 * more convenient than the raw iconv wrappers.
298 * Return value: -1 on error, 0 on success
301 g_iconv_close (GIConv converter)
303 iconv_t cd = (iconv_t)converter;
305 return iconv_close (cd);
309 open_converter (const gchar *to_codeset,
310 const gchar *from_codeset,
315 cd = g_iconv_open (to_codeset, from_codeset);
317 if (cd == (GIConv) -1)
319 /* Something went wrong. */
323 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_NO_CONVERSION,
324 _("Conversion from character set '%s' to '%s' is not supported"),
325 from_codeset, to_codeset);
327 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_FAILED,
328 _("Could not open converter from '%s' to '%s'"),
329 from_codeset, to_codeset);
337 close_converter (GIConv cd)
339 if (cd == (GIConv) -1)
342 return g_iconv_close (cd);
346 * g_convert_with_iconv:
347 * @str: the string to convert
348 * @len: the length of the string, or -1 if the string is
349 * nul-terminated (Note that some encodings may allow nul
350 * bytes to occur inside strings. In that case, using -1
351 * for the @len parameter is unsafe)
352 * @converter: conversion descriptor from g_iconv_open()
353 * @bytes_read: location to store the number of bytes in the
354 * input string that were successfully converted, or %NULL.
355 * Even if the conversion was successful, this may be
356 * less than @len if there were partial characters
357 * at the end of the input. If the error
358 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
359 * stored will the byte offset after the last valid
361 * @bytes_written: the number of bytes stored in the output buffer (not
362 * including the terminating nul).
363 * @error: location to store the error occurring, or %NULL to ignore
364 * errors. Any of the errors in #GConvertError may occur.
366 * Converts a string from one character set to another.
368 * Note that you should use g_iconv() for streaming conversions.
369 * Despite the fact that @byes_read can return information about partial
370 * characters, the g_convert_... functions are not generally suitable
371 * for streaming. If the underlying converter maintains internal state,
372 * then this won't be preserved across successive calls to g_convert(),
373 * g_convert_with_iconv() or g_convert_with_fallback(). (An example of
374 * this is the GNU C converter for CP1255 which does not emit a base
375 * character until it knows that the next character is not a mark that
376 * could combine with the base character.)
378 * Return value: If the conversion was successful, a newly allocated
379 * nul-terminated string, which must be freed with
380 * g_free(). Otherwise %NULL and @error will be set.
383 g_convert_with_iconv (const gchar *str,
387 gsize *bytes_written,
393 gsize inbytes_remaining;
394 gsize outbytes_remaining;
397 gboolean have_error = FALSE;
398 gboolean done = FALSE;
399 gboolean reset = FALSE;
401 g_return_val_if_fail (converter != (GIConv) -1, NULL);
407 inbytes_remaining = len;
408 outbuf_size = len + NUL_TERMINATOR_LENGTH;
410 outbytes_remaining = outbuf_size - NUL_TERMINATOR_LENGTH;
411 outp = dest = g_malloc (outbuf_size);
413 while (!done && !have_error)
416 err = g_iconv (converter, NULL, &inbytes_remaining, &outp, &outbytes_remaining);
418 err = g_iconv (converter, (char **)&p, &inbytes_remaining, &outp, &outbytes_remaining);
420 if (err == (gsize) -1)
425 /* Incomplete text, do not report an error */
430 gsize used = outp - dest;
433 dest = g_realloc (dest, outbuf_size);
436 outbytes_remaining = outbuf_size - used - NUL_TERMINATOR_LENGTH;
440 g_set_error_literal (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
441 _("Invalid byte sequence in conversion input"));
448 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_FAILED,
449 _("Error during conversion: %s"),
460 /* call g_iconv with NULL inbuf to cleanup shift state */
462 inbytes_remaining = 0;
469 memset (outp, 0, NUL_TERMINATOR_LENGTH);
472 *bytes_read = p - str;
475 if ((p - str) != len)
479 g_set_error_literal (error, G_CONVERT_ERROR, G_CONVERT_ERROR_PARTIAL_INPUT,
480 _("Partial character sequence at end of input"));
487 *bytes_written = outp - dest; /* Doesn't include '\0' */
500 * @str: the string to convert
501 * @len: the length of the string, or -1 if the string is
502 * nul-terminated (Note that some encodings may allow nul
503 * bytes to occur inside strings. In that case, using -1
504 * for the @len parameter is unsafe)
505 * @to_codeset: name of character set into which to convert @str
506 * @from_codeset: character set of @str.
507 * @bytes_read: (out): location to store the number of bytes in the
508 * input string that were successfully converted, or %NULL.
509 * Even if the conversion was successful, this may be
510 * less than @len if there were partial characters
511 * at the end of the input. If the error
512 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
513 * stored will the byte offset after the last valid
515 * @bytes_written: (out): the number of bytes stored in the output buffer (not
516 * including the terminating nul).
517 * @error: location to store the error occurring, or %NULL to ignore
518 * errors. Any of the errors in #GConvertError may occur.
520 * Converts a string from one character set to another.
522 * Note that you should use g_iconv() for streaming conversions.
523 * Despite the fact that @byes_read can return information about partial
524 * characters, the g_convert_... functions are not generally suitable
525 * for streaming. If the underlying converter maintains internal state,
526 * then this won't be preserved across successive calls to g_convert(),
527 * g_convert_with_iconv() or g_convert_with_fallback(). (An example of
528 * this is the GNU C converter for CP1255 which does not emit a base
529 * character until it knows that the next character is not a mark that
530 * could combine with the base character.)
532 * Return value: If the conversion was successful, a newly allocated
533 * nul-terminated string, which must be freed with
534 * g_free(). Otherwise %NULL and @error will be set.
537 g_convert (const gchar *str,
539 const gchar *to_codeset,
540 const gchar *from_codeset,
542 gsize *bytes_written,
548 g_return_val_if_fail (str != NULL, NULL);
549 g_return_val_if_fail (to_codeset != NULL, NULL);
550 g_return_val_if_fail (from_codeset != NULL, NULL);
552 cd = open_converter (to_codeset, from_codeset, error);
554 if (cd == (GIConv) -1)
565 res = g_convert_with_iconv (str, len, cd,
566 bytes_read, bytes_written,
569 close_converter (cd);
575 * g_convert_with_fallback:
576 * @str: the string to convert
577 * @len: the length of the string, or -1 if the string is
578 * nul-terminated (Note that some encodings may allow nul
579 * bytes to occur inside strings. In that case, using -1
580 * for the @len parameter is unsafe)
581 * @to_codeset: name of character set into which to convert @str
582 * @from_codeset: character set of @str.
583 * @fallback: UTF-8 string to use in place of character not
584 * present in the target encoding. (The string must be
585 * representable in the target encoding).
586 If %NULL, characters not in the target encoding will
587 be represented as Unicode escapes \uxxxx or \Uxxxxyyyy.
588 * @bytes_read: location to store the number of bytes in the
589 * input string that were successfully converted, or %NULL.
590 * Even if the conversion was successful, this may be
591 * less than @len if there were partial characters
592 * at the end of the input.
593 * @bytes_written: the number of bytes stored in the output buffer (not
594 * including the terminating nul).
595 * @error: location to store the error occurring, or %NULL to ignore
596 * errors. Any of the errors in #GConvertError may occur.
598 * Converts a string from one character set to another, possibly
599 * including fallback sequences for characters not representable
600 * in the output. Note that it is not guaranteed that the specification
601 * for the fallback sequences in @fallback will be honored. Some
602 * systems may do an approximate conversion from @from_codeset
603 * to @to_codeset in their iconv() functions,
604 * in which case GLib will simply return that approximate conversion.
606 * Note that you should use g_iconv() for streaming conversions.
607 * Despite the fact that @byes_read can return information about partial
608 * characters, the g_convert_... functions are not generally suitable
609 * for streaming. If the underlying converter maintains internal state,
610 * then this won't be preserved across successive calls to g_convert(),
611 * g_convert_with_iconv() or g_convert_with_fallback(). (An example of
612 * this is the GNU C converter for CP1255 which does not emit a base
613 * character until it knows that the next character is not a mark that
614 * could combine with the base character.)
616 * Return value: If the conversion was successful, a newly allocated
617 * nul-terminated string, which must be freed with
618 * g_free(). Otherwise %NULL and @error will be set.
621 g_convert_with_fallback (const gchar *str,
623 const gchar *to_codeset,
624 const gchar *from_codeset,
625 const gchar *fallback,
627 gsize *bytes_written,
633 const gchar *insert_str = NULL;
635 gsize inbytes_remaining;
636 const gchar *save_p = NULL;
637 gsize save_inbytes = 0;
638 gsize outbytes_remaining;
642 gboolean have_error = FALSE;
643 gboolean done = FALSE;
645 GError *local_error = NULL;
647 g_return_val_if_fail (str != NULL, NULL);
648 g_return_val_if_fail (to_codeset != NULL, NULL);
649 g_return_val_if_fail (from_codeset != NULL, NULL);
654 /* Try an exact conversion; we only proceed if this fails
655 * due to an illegal sequence in the input string.
657 dest = g_convert (str, len, to_codeset, from_codeset,
658 bytes_read, bytes_written, &local_error);
662 if (!g_error_matches (local_error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE))
664 g_propagate_error (error, local_error);
668 g_error_free (local_error);
672 /* No go; to proceed, we need a converter from "UTF-8" to
673 * to_codeset, and the string as UTF-8.
675 cd = open_converter (to_codeset, "UTF-8", error);
676 if (cd == (GIConv) -1)
687 utf8 = g_convert (str, len, "UTF-8", from_codeset,
688 bytes_read, &inbytes_remaining, error);
691 close_converter (cd);
697 /* Now the heart of the code. We loop through the UTF-8 string, and
698 * whenever we hit an offending character, we form fallback, convert
699 * the fallback to the target codeset, and then go back to
700 * converting the original string after finishing with the fallback.
702 * The variables save_p and save_inbytes store the input state
703 * for the original string while we are converting the fallback
707 outbuf_size = len + NUL_TERMINATOR_LENGTH;
708 outbytes_remaining = outbuf_size - NUL_TERMINATOR_LENGTH;
709 outp = dest = g_malloc (outbuf_size);
711 while (!done && !have_error)
713 gsize inbytes_tmp = inbytes_remaining;
714 err = g_iconv (cd, (char **)&p, &inbytes_tmp, &outp, &outbytes_remaining);
715 inbytes_remaining = inbytes_tmp;
717 if (err == (gsize) -1)
722 g_assert_not_reached();
726 gsize used = outp - dest;
729 dest = g_realloc (dest, outbuf_size);
732 outbytes_remaining = outbuf_size - used - NUL_TERMINATOR_LENGTH;
739 /* Error converting fallback string - fatal
741 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
742 _("Cannot convert fallback '%s' to codeset '%s'"),
743 insert_str, to_codeset);
751 gunichar ch = g_utf8_get_char (p);
752 insert_str = g_strdup_printf (ch < 0x10000 ? "\\u%04x" : "\\U%08x",
756 insert_str = fallback;
758 save_p = g_utf8_next_char (p);
759 save_inbytes = inbytes_remaining - (save_p - p);
761 inbytes_remaining = strlen (p);
764 /* fall thru if p is NULL */
769 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_FAILED,
770 _("Error during conversion: %s"),
783 g_free ((gchar *)insert_str);
785 inbytes_remaining = save_inbytes;
790 /* call g_iconv with NULL inbuf to cleanup shift state */
792 inbytes_remaining = 0;
801 memset (outp, 0, NUL_TERMINATOR_LENGTH);
803 close_converter (cd);
806 *bytes_written = outp - dest; /* Doesn't include '\0' */
812 if (save_p && !fallback)
813 g_free ((gchar *)insert_str);
828 strdup_len (const gchar *string,
830 gsize *bytes_written,
837 if (!g_utf8_validate (string, len, NULL))
844 g_set_error_literal (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
845 _("Invalid byte sequence in conversion input"));
850 real_len = strlen (string);
855 while (real_len < len && string[real_len])
860 *bytes_read = real_len;
862 *bytes_written = real_len;
864 return g_strndup (string, real_len);
869 * @opsysstring: a string in the encoding of the current locale. On Windows
870 * this means the system codepage.
871 * @len: the length of the string, or -1 if the string is
872 * nul-terminated (Note that some encodings may allow nul
873 * bytes to occur inside strings. In that case, using -1
874 * for the @len parameter is unsafe)
875 * @bytes_read: location to store the number of bytes in the
876 * input string that were successfully converted, or %NULL.
877 * Even if the conversion was successful, this may be
878 * less than @len if there were partial characters
879 * at the end of the input. If the error
880 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
881 * stored will the byte offset after the last valid
883 * @bytes_written: the number of bytes stored in the output buffer (not
884 * including the terminating nul).
885 * @error: location to store the error occurring, or %NULL to ignore
886 * errors. Any of the errors in #GConvertError may occur.
888 * Converts a string which is in the encoding used for strings by
889 * the C runtime (usually the same as that used by the operating
890 * system) in the <link linkend="setlocale">current locale</link> into a
893 * Return value: A newly-allocated buffer containing the converted string,
894 * or %NULL on an error, and error will be set.
897 g_locale_to_utf8 (const gchar *opsysstring,
900 gsize *bytes_written,
905 if (g_get_charset (&charset))
906 return strdup_len (opsysstring, len, bytes_read, bytes_written, error);
908 return g_convert (opsysstring, len,
909 "UTF-8", charset, bytes_read, bytes_written, error);
913 * g_locale_from_utf8:
914 * @utf8string: a UTF-8 encoded string
915 * @len: the length of the string, or -1 if the string is
916 * nul-terminated (Note that some encodings may allow nul
917 * bytes to occur inside strings. In that case, using -1
918 * for the @len parameter is unsafe)
919 * @bytes_read: location to store the number of bytes in the
920 * input string that were successfully converted, or %NULL.
921 * Even if the conversion was successful, this may be
922 * less than @len if there were partial characters
923 * at the end of the input. If the error
924 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
925 * stored will the byte offset after the last valid
927 * @bytes_written: the number of bytes stored in the output buffer (not
928 * including the terminating nul).
929 * @error: location to store the error occurring, or %NULL to ignore
930 * errors. Any of the errors in #GConvertError may occur.
932 * Converts a string from UTF-8 to the encoding used for strings by
933 * the C runtime (usually the same as that used by the operating
934 * system) in the <link linkend="setlocale">current locale</link>. On
935 * Windows this means the system codepage.
937 * Return value: A newly-allocated buffer containing the converted string,
938 * or %NULL on an error, and error will be set.
941 g_locale_from_utf8 (const gchar *utf8string,
944 gsize *bytes_written,
947 const gchar *charset;
949 if (g_get_charset (&charset))
950 return strdup_len (utf8string, len, bytes_read, bytes_written, error);
952 return g_convert (utf8string, len,
953 charset, "UTF-8", bytes_read, bytes_written, error);
956 #ifndef G_PLATFORM_WIN32
958 typedef struct _GFilenameCharsetCache GFilenameCharsetCache;
960 struct _GFilenameCharsetCache {
963 gchar **filename_charsets;
967 filename_charset_cache_free (gpointer data)
969 GFilenameCharsetCache *cache = data;
970 g_free (cache->charset);
971 g_strfreev (cache->filename_charsets);
976 * g_get_filename_charsets:
977 * @charsets: return location for the %NULL-terminated list of encoding names
979 * Determines the preferred character sets used for filenames.
980 * The first character set from the @charsets is the filename encoding, the
981 * subsequent character sets are used when trying to generate a displayable
982 * representation of a filename, see g_filename_display_name().
984 * On Unix, the character sets are determined by consulting the
985 * environment variables <envar>G_FILENAME_ENCODING</envar> and
986 * <envar>G_BROKEN_FILENAMES</envar>. On Windows, the character set
987 * used in the GLib API is always UTF-8 and said environment variables
990 * <envar>G_FILENAME_ENCODING</envar> may be set to a comma-separated list
991 * of character set names. The special token "@locale" is taken to
992 * mean the character set for the <link linkend="setlocale">current
993 * locale</link>. If <envar>G_FILENAME_ENCODING</envar> is not set, but
994 * <envar>G_BROKEN_FILENAMES</envar> is, the character set of the current
995 * locale is taken as the filename encoding. If neither environment variable
996 * is set, UTF-8 is taken as the filename encoding, but the character
997 * set of the current locale is also put in the list of encodings.
999 * The returned @charsets belong to GLib and must not be freed.
1001 * Note that on Unix, regardless of the locale character set or
1002 * <envar>G_FILENAME_ENCODING</envar> value, the actual file names present
1003 * on a system might be in any random encoding or just gibberish.
1005 * Return value: %TRUE if the filename encoding is UTF-8.
1010 g_get_filename_charsets (const gchar ***filename_charsets)
1012 static GPrivate cache_private = G_PRIVATE_INIT (filename_charset_cache_free);
1013 GFilenameCharsetCache *cache = g_private_get (&cache_private);
1014 const gchar *charset;
1018 cache = g_new0 (GFilenameCharsetCache, 1);
1019 g_private_set (&cache_private, cache);
1022 g_get_charset (&charset);
1024 if (!(cache->charset && strcmp (cache->charset, charset) == 0))
1026 const gchar *new_charset;
1030 g_free (cache->charset);
1031 g_strfreev (cache->filename_charsets);
1032 cache->charset = g_strdup (charset);
1034 p = getenv ("G_FILENAME_ENCODING");
1035 if (p != NULL && p[0] != '\0')
1037 cache->filename_charsets = g_strsplit (p, ",", 0);
1038 cache->is_utf8 = (strcmp (cache->filename_charsets[0], "UTF-8") == 0);
1040 for (i = 0; cache->filename_charsets[i]; i++)
1042 if (strcmp ("@locale", cache->filename_charsets[i]) == 0)
1044 g_get_charset (&new_charset);
1045 g_free (cache->filename_charsets[i]);
1046 cache->filename_charsets[i] = g_strdup (new_charset);
1050 else if (getenv ("G_BROKEN_FILENAMES") != NULL)
1052 cache->filename_charsets = g_new0 (gchar *, 2);
1053 cache->is_utf8 = g_get_charset (&new_charset);
1054 cache->filename_charsets[0] = g_strdup (new_charset);
1058 cache->filename_charsets = g_new0 (gchar *, 3);
1059 cache->is_utf8 = TRUE;
1060 cache->filename_charsets[0] = g_strdup ("UTF-8");
1061 if (!g_get_charset (&new_charset))
1062 cache->filename_charsets[1] = g_strdup (new_charset);
1066 if (filename_charsets)
1067 *filename_charsets = (const gchar **)cache->filename_charsets;
1069 return cache->is_utf8;
1072 #else /* G_PLATFORM_WIN32 */
1075 g_get_filename_charsets (const gchar ***filename_charsets)
1077 static const gchar *charsets[] = {
1083 /* On Windows GLib pretends that the filename charset is UTF-8 */
1084 if (filename_charsets)
1085 *filename_charsets = charsets;
1091 /* Cygwin works like before */
1092 result = g_get_charset (&(charsets[0]));
1094 if (filename_charsets)
1095 *filename_charsets = charsets;
1101 #endif /* G_PLATFORM_WIN32 */
1104 get_filename_charset (const gchar **filename_charset)
1106 const gchar **charsets;
1109 is_utf8 = g_get_filename_charsets (&charsets);
1111 if (filename_charset)
1112 *filename_charset = charsets[0];
1118 * g_filename_to_utf8:
1119 * @opsysstring: a string in the encoding for filenames
1120 * @len: the length of the string, or -1 if the string is
1121 * nul-terminated (Note that some encodings may allow nul
1122 * bytes to occur inside strings. In that case, using -1
1123 * for the @len parameter is unsafe)
1124 * @bytes_read: location to store the number of bytes in the
1125 * input string that were successfully converted, or %NULL.
1126 * Even if the conversion was successful, this may be
1127 * less than @len if there were partial characters
1128 * at the end of the input. If the error
1129 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
1130 * stored will the byte offset after the last valid
1132 * @bytes_written: the number of bytes stored in the output buffer (not
1133 * including the terminating nul).
1134 * @error: location to store the error occurring, or %NULL to ignore
1135 * errors. Any of the errors in #GConvertError may occur.
1137 * Converts a string which is in the encoding used by GLib for
1138 * filenames into a UTF-8 string. Note that on Windows GLib uses UTF-8
1139 * for filenames; on other platforms, this function indirectly depends on
1140 * the <link linkend="setlocale">current locale</link>.
1142 * Return value: The converted string, or %NULL on an error.
1145 g_filename_to_utf8 (const gchar *opsysstring,
1148 gsize *bytes_written,
1151 const gchar *charset;
1153 g_return_val_if_fail (opsysstring != NULL, NULL);
1155 if (get_filename_charset (&charset))
1156 return strdup_len (opsysstring, len, bytes_read, bytes_written, error);
1158 return g_convert (opsysstring, len,
1159 "UTF-8", charset, bytes_read, bytes_written, error);
1162 #if defined (G_OS_WIN32) && !defined (_WIN64)
1164 #undef g_filename_to_utf8
1166 /* Binary compatibility version. Not for newly compiled code. Also not needed for
1167 * 64-bit versions as there should be no old deployed binaries that would use
1172 g_filename_to_utf8 (const gchar *opsysstring,
1175 gsize *bytes_written,
1178 const gchar *charset;
1180 g_return_val_if_fail (opsysstring != NULL, NULL);
1182 if (g_get_charset (&charset))
1183 return strdup_len (opsysstring, len, bytes_read, bytes_written, error);
1185 return g_convert (opsysstring, len,
1186 "UTF-8", charset, bytes_read, bytes_written, error);
1192 * g_filename_from_utf8:
1193 * @utf8string: a UTF-8 encoded string.
1194 * @len: the length of the string, or -1 if the string is
1196 * @bytes_read: (out) (allow-none): location to store the number of bytes in
1197 * the input string that were successfully converted, or %NULL.
1198 * Even if the conversion was successful, this may be
1199 * less than @len if there were partial characters
1200 * at the end of the input. If the error
1201 * #G_CONVERT_ERROR_ILLEGAL_SEQUENCE occurs, the value
1202 * stored will the byte offset after the last valid
1204 * @bytes_written: (out): the number of bytes stored in the output buffer (not
1205 * including the terminating nul).
1206 * @error: location to store the error occurring, or %NULL to ignore
1207 * errors. Any of the errors in #GConvertError may occur.
1209 * Converts a string from UTF-8 to the encoding GLib uses for
1210 * filenames. Note that on Windows GLib uses UTF-8 for filenames;
1211 * on other platforms, this function indirectly depends on the
1212 * <link linkend="setlocale">current locale</link>.
1214 * Return value: (array length=bytes_written) (element-type guint8) (transfer full):
1215 * The converted string, or %NULL on an error.
1218 g_filename_from_utf8 (const gchar *utf8string,
1221 gsize *bytes_written,
1224 const gchar *charset;
1226 if (get_filename_charset (&charset))
1227 return strdup_len (utf8string, len, bytes_read, bytes_written, error);
1229 return g_convert (utf8string, len,
1230 charset, "UTF-8", bytes_read, bytes_written, error);
1233 #if defined (G_OS_WIN32) && !defined (_WIN64)
1235 #undef g_filename_from_utf8
1237 /* Binary compatibility version. Not for newly compiled code. */
1240 g_filename_from_utf8 (const gchar *utf8string,
1243 gsize *bytes_written,
1246 const gchar *charset;
1248 if (g_get_charset (&charset))
1249 return strdup_len (utf8string, len, bytes_read, bytes_written, error);
1251 return g_convert (utf8string, len,
1252 charset, "UTF-8", bytes_read, bytes_written, error);
1257 /* Test of haystack has the needle prefix, comparing case
1258 * insensitive. haystack may be UTF-8, but needle must
1259 * contain only ascii. */
1261 has_case_prefix (const gchar *haystack, const gchar *needle)
1265 /* Eat one character at a time. */
1270 g_ascii_tolower (*n) == g_ascii_tolower (*h))
1280 UNSAFE_ALL = 0x1, /* Escape all unsafe characters */
1281 UNSAFE_ALLOW_PLUS = 0x2, /* Allows '+' */
1282 UNSAFE_PATH = 0x8, /* Allows '/', '&', '=', ':', '@', '+', '$' and ',' */
1283 UNSAFE_HOST = 0x10, /* Allows '/' and ':' and '@' */
1284 UNSAFE_SLASHES = 0x20 /* Allows all characters except for '/' and '%' */
1285 } UnsafeCharacterSet;
1287 static const guchar acceptable[96] = {
1288 /* A table of the ASCII chars from space (32) to DEL (127) */
1289 /* ! " # $ % & ' ( ) * + , - . / */
1290 0x00,0x3F,0x20,0x20,0x28,0x00,0x2C,0x3F,0x3F,0x3F,0x3F,0x2A,0x28,0x3F,0x3F,0x1C,
1291 /* 0 1 2 3 4 5 6 7 8 9 : ; < = > ? */
1292 0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x38,0x20,0x20,0x2C,0x20,0x20,
1293 /* @ A B C D E F G H I J K L M N O */
1294 0x38,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,
1295 /* P Q R S T U V W X Y Z [ \ ] ^ _ */
1296 0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x20,0x20,0x20,0x20,0x3F,
1297 /* ` a b c d e f g h i j k l m n o */
1298 0x20,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,
1299 /* p q r s t u v w x y z { | } ~ DEL */
1300 0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x3F,0x20,0x20,0x20,0x3F,0x20
1303 static const gchar hex[16] = "0123456789ABCDEF";
1305 /* Note: This escape function works on file: URIs, but if you want to
1306 * escape something else, please read RFC-2396 */
1308 g_escape_uri_string (const gchar *string,
1309 UnsafeCharacterSet mask)
1311 #define ACCEPTABLE(a) ((a)>=32 && (a)<128 && (acceptable[(a)-32] & use_mask))
1318 UnsafeCharacterSet use_mask;
1320 g_return_val_if_fail (mask == UNSAFE_ALL
1321 || mask == UNSAFE_ALLOW_PLUS
1322 || mask == UNSAFE_PATH
1323 || mask == UNSAFE_HOST
1324 || mask == UNSAFE_SLASHES, NULL);
1328 for (p = string; *p != '\0'; p++)
1331 if (!ACCEPTABLE (c))
1335 result = g_malloc (p - string + unacceptable * 2 + 1);
1338 for (q = result, p = string; *p != '\0'; p++)
1342 if (!ACCEPTABLE (c))
1344 *q++ = '%'; /* means hex coming */
1359 g_escape_file_uri (const gchar *hostname,
1360 const gchar *pathname)
1362 char *escaped_hostname = NULL;
1367 char *p, *backslash;
1369 /* Turn backslashes into forward slashes. That's what Netscape
1370 * does, and they are actually more or less equivalent in Windows.
1373 pathname = g_strdup (pathname);
1374 p = (char *) pathname;
1376 while ((backslash = strchr (p, '\\')) != NULL)
1383 if (hostname && *hostname != '\0')
1385 escaped_hostname = g_escape_uri_string (hostname, UNSAFE_HOST);
1388 escaped_path = g_escape_uri_string (pathname, UNSAFE_PATH);
1390 res = g_strconcat ("file://",
1391 (escaped_hostname) ? escaped_hostname : "",
1392 (*escaped_path != '/') ? "/" : "",
1397 g_free ((char *) pathname);
1400 g_free (escaped_hostname);
1401 g_free (escaped_path);
1407 unescape_character (const char *scanner)
1412 first_digit = g_ascii_xdigit_value (scanner[0]);
1413 if (first_digit < 0)
1416 second_digit = g_ascii_xdigit_value (scanner[1]);
1417 if (second_digit < 0)
1420 return (first_digit << 4) | second_digit;
1424 g_unescape_uri_string (const char *escaped,
1426 const char *illegal_escaped_characters,
1427 gboolean ascii_must_not_be_escaped)
1429 const gchar *in, *in_end;
1430 gchar *out, *result;
1433 if (escaped == NULL)
1437 len = strlen (escaped);
1439 result = g_malloc (len + 1);
1442 for (in = escaped, in_end = escaped + len; in < in_end; in++)
1448 /* catch partial escape sequences past the end of the substring */
1449 if (in + 3 > in_end)
1452 c = unescape_character (in + 1);
1454 /* catch bad escape sequences and NUL characters */
1458 /* catch escaped ASCII */
1459 if (ascii_must_not_be_escaped && c <= 0x7F)
1462 /* catch other illegal escaped characters */
1463 if (strchr (illegal_escaped_characters, c) != NULL)
1472 g_assert (out - result <= len);
1485 is_asciialphanum (gunichar c)
1487 return c <= 0x7F && g_ascii_isalnum (c);
1491 is_asciialpha (gunichar c)
1493 return c <= 0x7F && g_ascii_isalpha (c);
1496 /* allows an empty string */
1498 hostname_validate (const char *hostname)
1501 gunichar c, first_char, last_char;
1508 /* read in a label */
1509 c = g_utf8_get_char (p);
1510 p = g_utf8_next_char (p);
1511 if (!is_asciialphanum (c))
1517 c = g_utf8_get_char (p);
1518 p = g_utf8_next_char (p);
1520 while (is_asciialphanum (c) || c == '-');
1521 if (last_char == '-')
1524 /* if that was the last label, check that it was a toplabel */
1525 if (c == '\0' || (c == '.' && *p == '\0'))
1526 return is_asciialpha (first_char);
1533 * g_filename_from_uri:
1534 * @uri: a uri describing a filename (escaped, encoded in ASCII).
1535 * @hostname: (out) (allow-none): Location to store hostname for the URI, or %NULL.
1536 * If there is no hostname in the URI, %NULL will be
1537 * stored in this location.
1538 * @error: location to store the error occurring, or %NULL to ignore
1539 * errors. Any of the errors in #GConvertError may occur.
1541 * Converts an escaped ASCII-encoded URI to a local filename in the
1542 * encoding used for filenames.
1544 * Return value: (type filename): a newly-allocated string holding
1545 * the resulting filename, or %NULL on an error.
1548 g_filename_from_uri (const gchar *uri,
1552 const char *path_part;
1553 const char *host_part;
1554 char *unescaped_hostname;
1565 if (!has_case_prefix (uri, "file:/"))
1567 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_BAD_URI,
1568 _("The URI '%s' is not an absolute URI using the \"file\" scheme"),
1573 path_part = uri + strlen ("file:");
1575 if (strchr (path_part, '#') != NULL)
1577 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_BAD_URI,
1578 _("The local file URI '%s' may not include a '#'"),
1583 if (has_case_prefix (path_part, "///"))
1585 else if (has_case_prefix (path_part, "//"))
1588 host_part = path_part;
1590 path_part = strchr (path_part, '/');
1592 if (path_part == NULL)
1594 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_BAD_URI,
1595 _("The URI '%s' is invalid"),
1600 unescaped_hostname = g_unescape_uri_string (host_part, path_part - host_part, "", TRUE);
1602 if (unescaped_hostname == NULL ||
1603 !hostname_validate (unescaped_hostname))
1605 g_free (unescaped_hostname);
1606 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_BAD_URI,
1607 _("The hostname of the URI '%s' is invalid"),
1613 *hostname = unescaped_hostname;
1615 g_free (unescaped_hostname);
1618 filename = g_unescape_uri_string (path_part, -1, "/", FALSE);
1620 if (filename == NULL)
1622 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_BAD_URI,
1623 _("The URI '%s' contains invalidly escaped characters"),
1630 /* Drop localhost */
1631 if (hostname && *hostname != NULL &&
1632 g_ascii_strcasecmp (*hostname, "localhost") == 0)
1638 /* Turn slashes into backslashes, because that's the canonical spelling */
1640 while ((slash = strchr (p, '/')) != NULL)
1646 /* Windows URIs with a drive letter can be like "file://host/c:/foo"
1647 * or "file://host/c|/foo" (some Netscape versions). In those cases, start
1648 * the filename from the drive letter.
1650 if (g_ascii_isalpha (filename[1]))
1652 if (filename[2] == ':')
1654 else if (filename[2] == '|')
1662 result = g_strdup (filename + offs);
1668 #if defined (G_OS_WIN32) && !defined (_WIN64)
1670 #undef g_filename_from_uri
1673 g_filename_from_uri (const gchar *uri,
1677 gchar *utf8_filename;
1678 gchar *retval = NULL;
1680 utf8_filename = g_filename_from_uri_utf8 (uri, hostname, error);
1683 retval = g_locale_from_utf8 (utf8_filename, -1, NULL, NULL, error);
1684 g_free (utf8_filename);
1692 * g_filename_to_uri:
1693 * @filename: an absolute filename specified in the GLib file name encoding,
1694 * which is the on-disk file name bytes on Unix, and UTF-8 on
1696 * @hostname: (allow-none): A UTF-8 encoded hostname, or %NULL for none.
1697 * @error: location to store the error occurring, or %NULL to ignore
1698 * errors. Any of the errors in #GConvertError may occur.
1700 * Converts an absolute filename to an escaped ASCII-encoded URI, with the path
1701 * component following Section 3.3. of RFC 2396.
1703 * Return value: a newly-allocated string holding the resulting
1704 * URI, or %NULL on an error.
1707 g_filename_to_uri (const gchar *filename,
1708 const gchar *hostname,
1713 g_return_val_if_fail (filename != NULL, NULL);
1715 if (!g_path_is_absolute (filename))
1717 g_set_error (error, G_CONVERT_ERROR, G_CONVERT_ERROR_NOT_ABSOLUTE_PATH,
1718 _("The pathname '%s' is not an absolute path"),
1724 !(g_utf8_validate (hostname, -1, NULL)
1725 && hostname_validate (hostname)))
1727 g_set_error_literal (error, G_CONVERT_ERROR, G_CONVERT_ERROR_ILLEGAL_SEQUENCE,
1728 _("Invalid hostname"));
1733 /* Don't use localhost unnecessarily */
1734 if (hostname && g_ascii_strcasecmp (hostname, "localhost") == 0)
1738 escaped_uri = g_escape_file_uri (hostname, filename);
1743 #if defined (G_OS_WIN32) && !defined (_WIN64)
1745 #undef g_filename_to_uri
1748 g_filename_to_uri (const gchar *filename,
1749 const gchar *hostname,
1752 gchar *utf8_filename;
1753 gchar *retval = NULL;
1755 utf8_filename = g_locale_to_utf8 (filename, -1, NULL, NULL, error);
1759 retval = g_filename_to_uri_utf8 (utf8_filename, hostname, error);
1760 g_free (utf8_filename);
1769 * g_uri_list_extract_uris:
1770 * @uri_list: an URI list
1772 * Splits an URI list conforming to the text/uri-list
1773 * mime type defined in RFC 2483 into individual URIs,
1774 * discarding any comments. The URIs are not validated.
1776 * Returns: (transfer full): a newly allocated %NULL-terminated list
1777 * of strings holding the individual URIs. The array should be freed
1778 * with g_strfreev().
1783 g_uri_list_extract_uris (const gchar *uri_list)
1794 /* We don't actually try to validate the URI according to RFC
1795 * 2396, or even check for allowed characters - we just ignore
1796 * comments and trim whitespace off the ends. We also
1797 * allow LF delimination as well as the specified CRLF.
1799 * We do allow comments like specified in RFC 2483.
1805 while (g_ascii_isspace (*p))
1809 while (*q && (*q != '\n') && (*q != '\r'))
1815 while (q > p && g_ascii_isspace (*q))
1820 uris = g_slist_prepend (uris, g_strndup (p, q - p + 1));
1825 p = strchr (p, '\n');
1830 result = g_new (gchar *, n_uris + 1);
1832 result[n_uris--] = NULL;
1833 for (u = uris; u; u = u->next)
1834 result[n_uris--] = u->data;
1836 g_slist_free (uris);
1842 * g_filename_display_basename:
1843 * @filename: an absolute pathname in the GLib file name encoding
1845 * Returns the display basename for the particular filename, guaranteed
1846 * to be valid UTF-8. The display name might not be identical to the filename,
1847 * for instance there might be problems converting it to UTF-8, and some files
1848 * can be translated in the display.
1850 * If GLib cannot make sense of the encoding of @filename, as a last resort it
1851 * replaces unknown characters with U+FFFD, the Unicode replacement character.
1852 * You can search the result for the UTF-8 encoding of this character (which is
1853 * "\357\277\275" in octal notation) to find out if @filename was in an invalid
1856 * You must pass the whole absolute pathname to this functions so that
1857 * translation of well known locations can be done.
1859 * This function is preferred over g_filename_display_name() if you know the
1860 * whole path, as it allows translation.
1862 * Return value: a newly allocated string containing
1863 * a rendition of the basename of the filename in valid UTF-8
1868 g_filename_display_basename (const gchar *filename)
1873 g_return_val_if_fail (filename != NULL, NULL);
1875 basename = g_path_get_basename (filename);
1876 display_name = g_filename_display_name (basename);
1878 return display_name;
1882 * g_filename_display_name:
1883 * @filename: a pathname hopefully in the GLib file name encoding
1885 * Converts a filename into a valid UTF-8 string. The conversion is
1886 * not necessarily reversible, so you should keep the original around
1887 * and use the return value of this function only for display purposes.
1888 * Unlike g_filename_to_utf8(), the result is guaranteed to be non-%NULL
1889 * even if the filename actually isn't in the GLib file name encoding.
1891 * If GLib cannot make sense of the encoding of @filename, as a last resort it
1892 * replaces unknown characters with U+FFFD, the Unicode replacement character.
1893 * You can search the result for the UTF-8 encoding of this character (which is
1894 * "\357\277\275" in octal notation) to find out if @filename was in an invalid
1897 * If you know the whole pathname of the file you should use
1898 * g_filename_display_basename(), since that allows location-based
1899 * translation of filenames.
1901 * Return value: a newly allocated string containing
1902 * a rendition of the filename in valid UTF-8
1907 g_filename_display_name (const gchar *filename)
1910 const gchar **charsets;
1911 gchar *display_name = NULL;
1914 is_utf8 = g_get_filename_charsets (&charsets);
1918 if (g_utf8_validate (filename, -1, NULL))
1919 display_name = g_strdup (filename);
1924 /* Try to convert from the filename charsets to UTF-8.
1925 * Skip the first charset if it is UTF-8.
1927 for (i = is_utf8 ? 1 : 0; charsets[i]; i++)
1929 display_name = g_convert (filename, -1, "UTF-8", charsets[i],
1937 /* if all conversions failed, we replace invalid UTF-8
1938 * by a question mark
1941 display_name = _g_utf8_make_valid (filename);
1943 return display_name;