* unicode.c: Unicode <-> locale functions
*****************************************************************************
* Copyright (C) 2005-2006 the VideoLAN team
- * Copyright © 2005-2006 Rémi Denis-Courmont
- * $Id$
+ * Copyright © 2005-2008 Rémi Denis-Courmont
*
* Authors: Rémi Denis-Courmont <rem # videolan.org>
*
/*****************************************************************************
* Preamble
*****************************************************************************/
-#include <vlc/vlc.h>
+#ifdef HAVE_CONFIG_H
+# include "config.h"
+#endif
+
+#include <vlc_common.h>
#include <vlc_charset.h>
+#include "libvlc.h" /* utf8_mkdir */
#include <assert.h>
#endif
#if defined (USE_ICONV)
-static char charset[sizeof ("CSISO11SWEDISHFORNAMES//translit")] = "";
+# include <langinfo.h>
+static char charset[sizeof ("CSISO11SWEDISHFORNAMES")] = "";
static void find_charset_once (void)
{
- char *psz_charset;
- if (vlc_current_charset (&psz_charset)
- || (psz_charset == NULL)
- || ((size_t)snprintf (charset, sizeof (charset), "%s//translit",
- psz_charset) >= sizeof (charset)))
- strcpy (charset, "UTF-8");
-
- free (psz_charset);
+ strlcpy (charset, nl_langinfo (CODESET), sizeof (charset));
+ if (!strcasecmp (charset, "ASCII")
+ || !strcasecmp (charset, "ANSI_X3.4-1968"))
+ strcpy (charset, "UTF-8"); /* superset... */
}
static int find_charset (void)
{
static pthread_once_t once = PTHREAD_ONCE_INIT;
pthread_once (&once, find_charset_once);
- return !strcmp (charset, "UTF-8");
+ return !strcasecmp (charset, "UTF-8");
}
#endif
-static char *locale_fast (const char *string, vlc_bool_t from)
+static char *locale_fast (const char *string, bool from)
{
#if defined (USE_ICONV)
if (find_charset ())
vlc_iconv_t hd = vlc_iconv_open (from ? "UTF-8" : charset,
from ? charset : "UTF-8");
if (hd == (vlc_iconv_t)(-1))
- return strdup (string); /* Uho! */
+ return NULL; /* Uho! */
const char *iptr = string;
size_t inb = strlen (string);
outb--;
iptr++;
inb--;
- vlc_iconv (hd, NULL, NULL, NULL, NULL);
+ vlc_iconv (hd, NULL, NULL, NULL, NULL); /* reset */
}
*optr = '\0';
vlc_iconv_close (hd);
0, string, -1, NULL, 0);
wchar_t wide[len];
- MultiByteToWideChar (from ? CP_UTF8 : CP_ACP, 0, string, -1, wide, len);
- len = 1 + WideCharToMultiByte (p->toCP, 0, wide, -1, NULL, 0, NULL, NULL);
+ MultiByteToWideChar (from ? CP_ACP : CP_UTF8, 0, string, -1, wide, len);
+ len = 1 + WideCharToMultiByte (from ? CP_UTF8 : CP_ACP, 0, wide, -1,
+ NULL, 0, NULL, NULL);
out = malloc (len);
if (out == NULL)
return NULL;
- WideCharToMultiByte (p->toCP, 0, wide, -1, out, len, NULL, NULL);
+ WideCharToMultiByte (from ? CP_UTF8 : CP_ACP, 0, wide, -1, out, len,
+ NULL, NULL);
return out;
#else
+ (void)from;
return (char *)string;
#endif
}
-static inline char *locale_dup (const char *string, vlc_bool_t from)
+static inline char *locale_dup (const char *string, bool from)
{
+ assert( string );
+
#if defined (USE_ICONV)
if (find_charset ())
return strdup (string);
#elif defined (USE_MB2MB)
return locale_fast (string, from);
#else
+ (void)from;
return strdup (string);
#endif
}
-
+/**
+ * Releases (if needed) a localized or uniformized string.
+ * @param str non-NULL return value from FromLocale() or ToLocale().
+ */
void LocaleFree (const char *str)
{
#if defined (USE_ICONV)
- if (find_charset ())
+ if (!find_charset ())
free ((char *)str);
#elif defined (USE_MB2MB)
free ((char *)str);
+#else
+ (void)str;
#endif
}
/**
- * FromLocale: converts a locale string to UTF-8
+ * Converts a string from the system locale character encoding to UTF-8.
*
- * @param locale nul-terminated string to be converted
+ * @param locale nul-terminated string to convert
*
* @return a nul-terminated UTF-8 string, or NULL in case of error.
* To avoid memory leak, you have to pass the result to LocaleFree()
*/
char *FromLocale (const char *locale)
{
- return locale_fast (locale, VLC_TRUE);
+ return locale_fast (locale, true);
}
+/**
+ * converts a string from the system locale character encoding to utf-8,
+ * the result is always allocated on the heap.
+ *
+ * @param locale nul-terminated string to convert
+ *
+ * @return a nul-terminated utf-8 string, or null in case of error.
+ * The result must be freed using free() - as with the strdup() function.
+ */
char *FromLocaleDup (const char *locale)
{
- return locale_dup (locale, VLC_TRUE);
+ return locale_dup (locale, true);
}
/**
- * ToLocale: converts a UTF-8 string to local system encoding.
+ * ToLocale: converts an UTF-8 string to local system encoding.
*
* @param utf8 nul-terminated string to be converted
*
*/
char *ToLocale (const char *utf8)
{
- return locale_fast (utf8, VLC_FALSE);
+ return locale_fast (utf8, false);
}
-static char *ToLocaleDup (const char *utf8)
+/**
+ * converts a string from UTF-8 to the system locale character encoding,
+ * the result is always allocated on the heap.
+ *
+ * @param utf8 nul-terminated string to convert
+ *
+ * @return a nul-terminated string, or null in case of error.
+ * The result must be freed using free() - as with the strdup() function.
+ */
+char *ToLocaleDup (const char *utf8)
{
- return locale_dup (utf8, VLC_FALSE);
+ return locale_dup (utf8, false);
}
/**
- * utf8_open: open() wrapper for UTF-8 filenames
+ * Opens a system file handle using UTF-8 paths.
+ *
+ * @param filename file path to open (with UTF-8 encoding)
+ * @param flags open() flags, see the C library open() documentation
+ * @param mode file permissions if creating a new file
+ * @return a file handle on success, -1 on error (see errno).
*/
int utf8_open (const char *filename, int flags, mode_t mode)
{
}
/**
- * utf8_fopen: fopen() wrapper for UTF-8 filenames
+ * Opens a FILE pointer using UTF-8 filenames.
+ * @param filename file path, using UTF-8 encoding
+ * @param mode fopen file open mode
+ * @return NULL on error, an open FILE pointer on success.
*/
FILE *utf8_fopen (const char *filename, const char *mode)
{
int rwflags = 0, oflags = 0;
- vlc_bool_t append = VLC_FALSE;
+ bool append = false;
for (const char *ptr = mode; *ptr; ptr++)
{
case 'a':
rwflags = O_WRONLY;
oflags |= O_CREAT;
- append = VLC_TRUE;
+ append = true;
break;
case 'w':
}
/**
- * utf8_mkdir: Calls mkdir() after conversion of file name to OS locale
+ * Creates a directory using UTF-8 paths.
*
* @param dirname a UTF-8 string with the name of the directory that you
* want to create.
- * @return A 0 return value indicates success. A -1 return value indicates an
- * error, and an error code is stored in errno
+ * @param mode directory permissions
+ * @return 0 on success, -1 on error (see errno).
*/
-int utf8_mkdir( const char *dirname )
+int utf8_mkdir( const char *dirname, mode_t mode )
{
#if defined (UNDER_CE) || defined (WIN32)
+ VLC_UNUSED( mode );
+
wchar_t wname[MAX_PATH + 1];
char mod[MAX_PATH + 1];
int i;
errno = ENOENT;
return -1;
}
- res = mkdir( locname, 0755 );
+ res = mkdir( locname, mode );
LocaleFree( locname );
return res;
}
/**
- * utf8_opendir: wrapper that converts dirname to the locale in use by the OS
+ * Opens a DIR pointer using UTF-8 paths
*
* @param dirname UTF-8 representation of the directory name
- *
- * @return a pointer to the DIR struct. Release with closedir().
+ * @return a pointer to the DIR struct, or NULL in case of error.
+ * Release with standard closedir().
*/
DIR *utf8_opendir( const char *dirname )
{
}
/**
- * utf8_readdir: a readdir wrapper that returns the name of the next entry
- * in the directory as a UTF-8 string.
+ * Reads the next file name from an open directory.
*
* @param dir The directory that is being read
*
- * @return a UTF-8 string of the directory entry. Use LocaleFree() to free this memory
+ * @return a UTF-8 string of the directory entry.
+ * Use free() to free this memory.
*/
char *utf8_readdir( DIR *dir )
{
return 1;
}
+/**
+ * Does the same as utf8_scandir(), but takes an open directory pointer
+ * instead of a directory path.
+ */
int utf8_loaddir( DIR *dir, char ***namelist,
int (*select)( const char * ),
int (*compar)( const char **, const char ** ) )
return -1;
}
+/**
+ * Selects file entries from a directory, as GNU C scandir(), yet using
+ * UTF-8 file names.
+ *
+ * @param dirname UTF-8 diretory path
+ * @param pointer [OUT] pointer set, on succesful completion, to the address
+ * of a table of UTF-8 filenames. All filenames must be freed with free().
+ * The table itself must be freed with free() as well.
+ *
+ * @return How many file names were selected (possibly 0),
+ * or -1 in case of error.
+ */
int utf8_scandir( const char *dirname, char ***namelist,
int (*select)( const char * ),
int (*compar)( const char **, const char ** ) )
}
static int utf8_statEx( const char *filename, struct stat *buf,
- vlc_bool_t deref )
+ bool deref )
{
#if defined (WIN32) || defined (UNDER_CE)
/* retrieve Windows OS version */
return -1;
}
-
+/**
+ * Finds file/inode informations, as stat().
+ * Consider usign fstat() instead, if possible.
+ *
+ * @param filename UTF-8 file path
+ */
int utf8_stat( const char *filename, struct stat *buf)
{
- return utf8_statEx( filename, buf, VLC_TRUE );
+ return utf8_statEx( filename, buf, true );
}
+/**
+ * Finds file/inode informations, as lstat().
+ * Consider usign fstat() instead, if possible.
+ *
+ * @param filename UTF-8 file path
+ */
int utf8_lstat( const char *filename, struct stat *buf)
{
- return utf8_statEx( filename, buf, VLC_FALSE );
+ return utf8_statEx( filename, buf, false );
}
/**
- * utf8_*printf: *printf with conversion from UTF-8 to local encoding
+ * Removes a file.
+ *
+ * @param filename a UTF-8 string with the name of the file you want to delete.
+ * @return A 0 return value indicates success. A -1 return value indicates an
+ * error, and an error code is stored in errno
+ */
+int utf8_unlink( const char *filename )
+{
+#if defined (WIN32) || defined (UNDER_CE)
+ if( GetVersion() < 0x80000000 )
+ {
+ /* for Windows NT and above */
+ wchar_t wpath[MAX_PATH + 1];
+
+ if( !MultiByteToWideChar( CP_UTF8, 0, filename, -1, wpath, MAX_PATH ) )
+ {
+ errno = ENOENT;
+ return -1;
+ }
+ wpath[MAX_PATH] = L'\0';
+
+ /*
+ * unlink() cannot open files with non-“ANSI” characters on Windows.
+ * We use _wunlink() instead.
+ */
+ return _wunlink( wpath );
+ }
+#endif
+ const char *local_name = ToLocale( filename );
+
+ if( local_name == NULL )
+ {
+ errno = ENOENT;
+ return -1;
+ }
+
+ int ret = unlink( local_name );
+ LocaleFree( local_name );
+ return ret;
+}
+
+
+
+/**
+ * Formats an UTF-8 string as vasprintf(), then print it to stdout, with
+ * appropriate conversion to local encoding.
*/
static int utf8_vasprintf( char **str, const char *fmt, va_list ap )
{
return res;
}
+/**
+ * Formats an UTF-8 string as vfprintf(), then print it, with
+ * appropriate conversion to local encoding.
+ */
int utf8_vfprintf( FILE *stream, const char *fmt, va_list ap )
{
char *str;
return res;
}
+/**
+ * Formats an UTF-8 string as fprintf(), then print it, with
+ * appropriate conversion to local encoding.
+ */
int utf8_fprintf( FILE *stream, const char *fmt, ... )
{
va_list ap;
}
/**
- * EnsureUTF8: replaces invalid/overlong UTF-8 sequences with question marks
+ * Replaces invalid/overlong UTF-8 sequences with question marks.
* Note that it is not possible to convert from Latin-1 to UTF-8 on the fly,
* so we don't try that, even though it would be less disruptive.
*
/**
- * IsUTF8: checks whether a string is a valid UTF-8 byte sequence.
+ * Checks whether a string is a valid UTF-8 byte sequence.
*
* @param str nul-terminated string to be checked
*