root/src/text/unicode.c

/*****************************************************************************
 * unicode.c: Unicode <-> locale functions
 *****************************************************************************
 * Copyright (C) 2005-2006 the VideoLAN team
 * Copyright © 2005-2008 Rémi Denis-Courmont
 *
 * Authors: Rémi Denis-Courmont <rem # videolan.org>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston MA 02110-1301, USA.
 *****************************************************************************/

/*****************************************************************************
 * Preamble
 *****************************************************************************/
#ifdef HAVE_CONFIG_H
# include "config.h"
#endif

#include <vlc_common.h>
#include <vlc_charset.h>
#include "libvlc.h" /* utf8_mkdir */

#include <assert.h>

#include <stdio.h>
#include <stdarg.h>
#include <stdlib.h>
#include <errno.h>
#include <sys/types.h>
#ifdef HAVE_DIRENT_H
#  include <dirent.h>
#endif
#ifdef UNDER_CE
#  include <tchar.h>
#endif
#ifdef HAVE_SYS_STAT_H
# include <sys/stat.h>
#endif
#ifdef HAVE_FCNTL_H
# include <fcntl.h>
#endif
#ifdef WIN32
# include <io.h>
#else
# include <unistd.h>
#endif

#ifndef HAVE_LSTAT
# define lstat( a, b ) stat(a, b)
#endif

#ifdef __APPLE__
/* Define this if the OS always use UTF-8 internally */
# define ASSUME_UTF8 1
#endif

#if defined (ASSUME_UTF8)
/* Cool */
#elif defined (WIN32) || defined (UNDER_CE)
# define USE_MB2MB 1
#elif defined (HAVE_ICONV)
# define USE_ICONV 1
#else
# error No UTF8 charset conversion implemented on this platform!
#endif

#if defined (USE_ICONV)
# include <langinfo.h>
static char charset[sizeof ("CSISO11SWEDISHFORNAMES")] = "";

static void find_charset_once (void)
{
    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 !strcasecmp (charset, "UTF-8");
}
#endif


static char *locale_fast (const char *string, bool from)
{
#if defined (USE_ICONV)
    if (find_charset ())
        return (char *)string;

    vlc_iconv_t hd = vlc_iconv_open (from ? "UTF-8" : charset,
                                     from ? charset : "UTF-8");
    if (hd == (vlc_iconv_t)(-1))
        return NULL; /* Uho! */

    const char *iptr = string;
    size_t inb = strlen (string);
    size_t outb = inb * 6 + 1;
    char output[outb], *optr = output;

    if (string == NULL)
        return NULL;

    while (vlc_iconv (hd, &iptr, &inb, &optr, &outb) == (size_t)(-1))
    {
        *optr++ = '?';
        outb--;
        iptr++;
        inb--;
        vlc_iconv (hd, NULL, NULL, NULL, NULL); /* reset */
    }
    *optr = '\0';
    vlc_iconv_close (hd);

    assert (inb == 0);
    assert (*iptr == '\0');
    assert (*optr == '\0');
    assert (strlen (output) == (size_t)(optr - output));
    return strdup (output);
#elif defined (USE_MB2MB)
    char *out;
    int len;

    if (string == NULL)
        return NULL;

    len = 1 + MultiByteToWideChar (from ? CP_ACP : CP_UTF8,
                                   0, string, -1, NULL, 0);
    wchar_t wide[len];

    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 (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, bool from)
{
    assert( string );

#if defined (USE_ICONV)
    if (find_charset ())
        return strdup (string);
    return locale_fast (string, from);
#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 ())
        free ((char *)str);
#elif defined (USE_MB2MB)
    free ((char *)str);
#else
    (void)str;
#endif
}


/**
 * Converts a string from the system locale character encoding to UTF-8.
 *
 * @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()
 * when it is no longer needed.
 */
char *FromLocale (const char *locale)
{
    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, true);
}


/**
 * ToLocale: converts an UTF-8 string to local system encoding.
 *
 * @param utf8 nul-terminated string to be converted
 *
 * @return a nul-terminated string, or NULL in case of error.
 * To avoid memory leak, you have to pass the result to LocaleFree()
 * when it is no longer needed.
 */
char *ToLocale (const char *utf8)
{
    return locale_fast (utf8, false);
}


/**
 * 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, false);
}


/**
 * 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)
{
#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';

        /*
         * open() cannot open files with non-“ANSI” characters on Windows.
         * We use _wopen() instead. Same thing for mkdir() and stat().
         */
        return _wopen (wpath, flags, mode);
    }
#endif
    const char *local_name = ToLocale (filename);

    if (local_name == NULL)
    {
        errno = ENOENT;
        return -1;
    }

    int fd = open (local_name, flags, mode);
    LocaleFree (local_name);
    return fd;
}

/**
 * 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;
    bool append = false;

    for (const char *ptr = mode; *ptr; ptr++)
    {
        switch (*ptr)
        {
            case 'r':
                rwflags = O_RDONLY;
                break;

            case 'a':
                rwflags = O_WRONLY;
                oflags |= O_CREAT;
                append = true;
                break;

            case 'w':
                rwflags = O_WRONLY;
                oflags |= O_CREAT | O_TRUNC;
                break;

            case '+':
                rwflags = O_RDWR;
                break;

#ifdef O_TEXT
            case 't':
                oflags |= O_TEXT;
                break;
#endif
        }
    }

    int fd = utf8_open (filename, rwflags | oflags, 0666);
    if (fd == -1)
        return NULL;

    if (append && (lseek (fd, 0, SEEK_END) == -1))
    {
        close (fd);
        return NULL;
    }

    FILE *stream = fdopen (fd, mode);
    if (stream == NULL)
        close (fd);

    return stream;
}

/**
 * Creates a directory using UTF-8 paths.
 *
 * @param dirname a UTF-8 string with the name of the directory that you
 *        want to create.
 * @param mode directory permissions
 * @return 0 on success, -1 on error (see errno).
 */
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;

    /* Convert '/' into '\' */
    for( i = 0; *dirname; i++ )
    {
        if( i == MAX_PATH )
            return -1; /* overflow */

        if( *dirname == '/' )
            mod[i] = '\\';
        else
            mod[i] = *dirname;
        dirname++;

    }
    mod[i] = 0;

    if( MultiByteToWideChar( CP_UTF8, 0, mod, -1, wname, MAX_PATH ) == 0 )
    {
        errno = ENOENT;
        return -1;
    }
    wname[MAX_PATH] = L'\0';

    if( CreateDirectoryW( wname, NULL ) == 0 )
    {
        if( GetLastError( ) == ERROR_ALREADY_EXISTS )
            errno = EEXIST;
        else
            errno = ENOENT;
        return -1;
    }
    return 0;
#else
    char *locname = ToLocale( dirname );
    int res;

    if( locname == NULL )
    {
        errno = ENOENT;
        return -1;
    }
    res = mkdir( locname, mode );

    LocaleFree( locname );
    return res;
#endif
}

/**
 * Opens a DIR pointer using UTF-8 paths
 *
 * @param dirname UTF-8 representation of the directory name
 * @return a pointer to the DIR struct, or NULL in case of error.
 * Release with standard closedir().
 */
DIR *utf8_opendir( const char *dirname )
{
#ifdef WIN32
    wchar_t wname[MAX_PATH + 1];

    if (MultiByteToWideChar (CP_UTF8, 0, dirname, -1, wname, MAX_PATH))
    {
        wname[MAX_PATH] = L'\0';
        return (DIR *)vlc_wopendir (wname);
    }
#else
    const char *local_name = ToLocale( dirname );

    if( local_name != NULL )
    {
        DIR *dir = opendir( local_name );
        LocaleFree( local_name );
        return dir;
    }
#endif

    errno = ENOENT;
    return NULL;
}

/**
 * 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 free() to free this memory.
 */
char *utf8_readdir( DIR *dir )
{
#ifdef WIN32
    struct _wdirent *ent = vlc_wreaddir (dir);
    if (ent == NULL)
        return NULL;

    return FromWide (ent->d_name);
#else
    struct dirent *ent;

    ent = readdir( (DIR *)dir );
    if( ent == NULL )
        return NULL;

    return vlc_fix_readdir( ent->d_name );
#endif
}

static int dummy_select( const char *str )
{
    (void)str;
    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 ** ) )
{
    if( select == NULL )
        select = dummy_select;

    if( dir == NULL )
        return -1;
    else
    {
        char **tab = NULL;
        char *entry;
        unsigned num = 0;

        rewinddir( dir );

        while( ( entry = utf8_readdir( dir ) ) != NULL )
        {
            char **newtab;

            if( !select( entry ) )
            {
                free( entry );
                continue;
            }

            newtab = realloc( tab, sizeof( char * ) * (num + 1) );
            if( newtab == NULL )
            {
                free( entry );
                goto error;
            }
            tab = newtab;
            tab[num++] = entry;
        }

        if( compar != NULL )
            qsort( tab, num, sizeof( tab[0] ),
                   (int (*)( const void *, const void *))compar );

        *namelist = tab;
        return num;

    error:{
        unsigned i;

        for( i = 0; i < num; i++ )
            free( tab[i] );
        if( tab != NULL )
            free( tab );
        }
    }
    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 ** ) )
{
    DIR *dir = utf8_opendir (dirname);
    int val = -1;

    if (dir != NULL)
    {
        val = utf8_loaddir (dir, namelist, select, compar);
        closedir (dir);
    }
    return val;
}

static int utf8_statEx( const char *filename, struct stat *buf,
                        bool deref )
{
#if defined (WIN32) || defined (UNDER_CE)
    /* retrieve Windows OS version */
    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';

        return _wstati64( wpath, buf );
    }
#endif
#ifdef HAVE_SYS_STAT_H
    const char *local_name = ToLocale( filename );

    if( local_name != NULL )
    {
        int res = deref ? stat( local_name, buf )
                       : lstat( local_name, buf );
        LocaleFree( local_name );
        return res;
    }
    errno = ENOENT;
#endif
    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, 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, false );
}

/**
 * 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 )
{
    char *utf8;
    int res = vasprintf( &utf8, fmt, ap );
    if( res == -1 )
        return -1;

    *str = ToLocaleDup( utf8 );
    free( utf8 );
    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;
    int res = utf8_vasprintf( &str, fmt, ap );
    if( res == -1 )
        return -1;

    fputs( str, stream );
    free( 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;
    int res;

    va_start( ap, fmt );
    res = utf8_vfprintf( stream, fmt, ap );
    va_end( ap );
    return res;
}


static char *CheckUTF8( char *str, char rep )
{
    uint8_t *ptr = (uint8_t *)str;
    assert (str != NULL);

    for (;;)
    {
        uint8_t c = ptr[0];
        int charlen = -1;

        if (c == '\0')
            break;

        for (int i = 0; i < 7; i++)
            if ((c >> (7 - i)) == ((0xff >> (7 - i)) ^ 1))
            {
                charlen = i;
                break;
            }

        switch (charlen)
        {
            case 0: // 7-bit ASCII character -> OK
                ptr++;
                continue;

            case -1: // 1111111x -> error
            case 1: // continuation byte -> error
                goto error;
        }

        assert (charlen >= 2);

        uint32_t cp = c & ~((0xff >> (7 - charlen)) << (7 - charlen));
        for (int i = 1; i < charlen; i++)
        {
            assert (cp < (1 << 26));
            c = ptr[i];

            if ((c == '\0') // unexpected end of string
             || ((c >> 6) != 2)) // not a continuation byte
                goto error;

            cp = (cp << 6) | (ptr[i] & 0x3f);
        }

        if (cp < 128) // overlong (special case for ASCII)
            goto error;
        if (cp < (1u << (5 * charlen - 3))) // overlong
            goto error;

        ptr += charlen;
        continue;

    error:
        if (rep == 0)
            return NULL;
        *ptr++ = rep;
        str = NULL;
    }

    return str;
}

/**
 * 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.
 *
 * @return str if it was valid UTF-8, NULL if not.
 */
char *EnsureUTF8( char *str )
{
    return CheckUTF8( str, '?' );
}


/**
 * Checks whether a string is a valid UTF-8 byte sequence.
 *
 * @param str nul-terminated string to be checked
 *
 * @return str if it was valid UTF-8, NULL if not.
 */
const char *IsUTF8( const char *str )
{
    return CheckUTF8( (char *)str, 0 );
}