X-Git-Url: https://git.sesse.net/?a=blobdiff_plain;f=modules%2Fcodec%2Fsubtitles%2Fsubsdec.c;h=c49e8f1dae08de4ec555e6295e984095472a93be;hb=22da31efd99ff625cfd7e74d2231667a74c582d7;hp=f217f0b77e244effe9cd8e718a68939c6f68a15d;hpb=9d5e75a7a5643a2144dc561fc246a7a3ca7e5fb0;p=vlc

diff --git a/modules/codec/subtitles/subsdec.c b/modules/codec/subtitles/subsdec.c
index f217f0b77e..c49e8f1dae 100644
--- a/modules/codec/subtitles/subsdec.c
+++ b/modules/codec/subtitles/subsdec.c
@@ -68,7 +68,7 @@ static const char *const ppsz_encodings[] = {
     "ISO-8859-6",
     "Windows-1256",
     "ISO-8859-7",
-    "Windows-1256",
+    "Windows-1253",
     "ISO-8859-8",
     "Windows-1255",
     "ISO-8859-9",
@@ -84,8 +84,8 @@ static const char *const ppsz_encodings[] = {
     "ISO-2022-JP-2",
     "EUC-JP",
     "Shift_JIS",
+    "CP949",
     "ISO-2022-KR",
-    "EUC-KR",
     "Big5",
     "ISO-2022-TW",
     "Big5-HKSCS",
@@ -94,7 +94,11 @@ static const char *const ppsz_encodings[] = {
 };
 
 static const char *const ppsz_encoding_names[] = {
-    N_("Auto"),
+    /* xgettext:
+      The character encoding name in parenthesis corresponds to that used for
+      the GetACP translation. "Windows-1252" applies to Western European
+      languages using the Latin alphabet. */
+    N_("Default (Windows-1252)"),
     N_("Universal (UTF-8)"),
     N_("Universal (UTF-16)"),
     N_("Universal (big endian UTF-16)"),
@@ -121,7 +125,7 @@ static const char *const ppsz_encoding_names[] = {
     N_("Arabic (Windows-1256)"),
     /* 7 */
     N_("Greek (ISO 8859-7)"),
-    N_("Greek (Windows-1256)"),
+    N_("Greek (Windows-1253)"),
     /* 8 */
     N_("Hebrew (ISO 8859-8)"),
     N_("Hebrew (Windows-1255)"),
@@ -147,8 +151,8 @@ static const char *const ppsz_encoding_names[] = {
     N_("Japanese (7-bits JIS/ISO-2022-JP-2)"),
     N_("Japanese Unix (EUC-JP)"),
     N_("Japanese (Shift JIS)"),
+    N_("Korean (EUC-KR/CP949)"),
     N_("Korean (ISO-2022-KR)"),
-    N_("Korean Unix (EUC-KR)"),
     N_("Traditional Chinese (Big5)"),
     N_("Traditional Chinese Unix (EUC-TW)"),
     N_("Hong-Kong Supplementary (HKSCS)"),
@@ -228,7 +232,6 @@ static int OpenDecoder( vlc_object_t *p_this )
 {
     decoder_t     *p_dec = (decoder_t*)p_this;
     decoder_sys_t *p_sys;
-    vlc_value_t    val;
 
     switch( p_dec->fmt_in.i_codec )
     {
@@ -276,7 +279,7 @@ static int OpenDecoder( vlc_object_t *p_this )
     /* Second, try configured encoding */
     if (psz_charset == NULL)
     {
-        psz_charset = var_CreateGetNonEmptyString (p_dec, "subsdec-encoding");
+        psz_charset = var_InheritString (p_dec, "subsdec-encoding");
         msg_Dbg (p_dec, "trying configured character encoding: %s",
                  psz_charset ? psz_charset : "not specified");
     }
@@ -284,11 +287,23 @@ static int OpenDecoder( vlc_object_t *p_this )
     /* Third, try "local" encoding with optional UTF-8 autodetection */
     if (psz_charset == NULL)
     {
-        psz_charset = strdup (GetFallbackEncoding ());
+        /* xgettext:
+           The Windows ANSI code page most commonly used for this language.
+           VLC uses this as a guess of the subtitle files character set
+           (if UTF-8 and UTF-16 autodetection fails).
+           Western European languages normally use "CP1252", which is a
+           Microsoft-variant of ISO 8859-1. That suits the Latin alphabet.
+           Other scripts use other code pages.
+
+           This MUST be a valid iconv character set. If unsure, please refer
+           the VideoLAN translators mailing list. */
+        const char *acp = vlc_pgettext("GetACP", "CP1252");
+
+        psz_charset = strdup (acp);
         msg_Dbg (p_dec, "trying default character encoding: %s",
                  psz_charset ? psz_charset : "not specified");
 
-        if (var_CreateGetBool (p_dec, "subsdec-autodetect-utf8"))
+        if (var_InheritBool (p_dec, "subsdec-autodetect-utf8"))
         {
             msg_Dbg (p_dec, "using automatic UTF-8 detection");
             p_sys->b_autodetect_utf8 = true;
@@ -312,12 +327,10 @@ static int OpenDecoder( vlc_object_t *p_this )
     }
     free (psz_charset);
 
-    var_Create( p_dec, "subsdec-align", VLC_VAR_INTEGER | VLC_VAR_DOINHERIT );
-    var_Get( p_dec, "subsdec-align", &val );
-    p_sys->i_align = val.i_int;
+    p_sys->i_align = var_InheritInteger( p_dec, "subsdec-align" );
 
     if( p_dec->fmt_in.i_codec == VLC_CODEC_SSA
-     && var_CreateGetBool( p_dec, "subsdec-formatted" ) )
+     && var_InheritBool( p_dec, "subsdec-formatted" ) )
     {
         if( p_dec->fmt_in.i_extra > 0 )
             ParseSSAHeader( p_dec );
@@ -410,7 +423,7 @@ static subpicture_t *ParseText( decoder_t *p_dec, block_t *p_block )
     video_format_t fmt;
 
     /* We cannot display a subpicture with no date */
-    if( p_block->i_pts == 0 )
+    if( p_block->i_pts <= VLC_TS_INVALID )
     {
         msg_Warn( p_dec, "subtitle without a date" );
         return NULL;
@@ -426,10 +439,11 @@ static subpicture_t *ParseText( decoder_t *p_dec, block_t *p_block )
     }
 
     /* Should be resiliant against bad subtitles */
-    psz_subtitle = strndup( (const char *)p_block->p_buffer,
-                            p_block->i_buffer );
+    psz_subtitle = malloc( p_block->i_buffer + 1 );
     if( psz_subtitle == NULL )
         return NULL;
+    memcpy( psz_subtitle, p_block->p_buffer, p_block->i_buffer );
+    psz_subtitle[p_block->i_buffer] = '\0';
 
     if( p_sys->iconv_handle == (vlc_iconv_t)-1 )
     {
@@ -457,7 +471,7 @@ static subpicture_t *ParseText( decoder_t *p_dec, block_t *p_block )
         {
             size_t inbytes_left = strlen( psz_subtitle );
             size_t outbytes_left = 6 * inbytes_left;
-            char *psz_new_subtitle = malloc( outbytes_left + 1 );
+            char *psz_new_subtitle = xmalloc( outbytes_left + 1 );
             char *psz_convert_buffer_out = psz_new_subtitle;
             const char *psz_convert_buffer_in = psz_subtitle;
 
@@ -479,11 +493,13 @@ static subpicture_t *ParseText( decoder_t *p_dec, block_t *p_block )
 
             psz_subtitle = realloc( psz_new_subtitle,
                                     psz_convert_buffer_out - psz_new_subtitle );
+            if( !psz_subtitle )
+                psz_subtitle = psz_new_subtitle;
         }
     }
 
     /* Create the subpicture unit */
-    p_spu = decoder_NewSubpicture( p_dec );
+    p_spu = decoder_NewSubpicture( p_dec, NULL );
     if( !p_spu )
     {
         msg_Warn( p_dec, "can't get spu buffer" );
@@ -494,7 +510,6 @@ static subpicture_t *ParseText( decoder_t *p_dec, block_t *p_block )
     /* Create a new subpicture region */
     memset( &fmt, 0, sizeof(video_format_t) );
     fmt.i_chroma = VLC_CODEC_TEXT;
-    fmt.i_aspect = 0;
     fmt.i_width = fmt.i_height = 0;
     fmt.i_x_offset = fmt.i_y_offset = 0;
     p_spu->p_region = subpicture_region_New( &fmt );
@@ -517,7 +532,7 @@ static subpicture_t *ParseText( decoder_t *p_dec, block_t *p_block )
         /* Remove formatting from string */
 
         p_spu->p_region->psz_text = StripTags( psz_subtitle );
-        if( var_CreateGetBool( p_dec, "subsdec-formatted" ) )
+        if( var_InheritBool( p_dec, "subsdec-formatted" ) )
         {
             p_spu->p_region->psz_html = CreateHtmlSubtitle( &p_spu->p_region->i_align, psz_subtitle );
         }
@@ -620,7 +635,8 @@ static char *StripTags( char *psz_subtitle )
         psz_subtitle++;
     }
     *psz_text = '\0';
-    psz_text_start = realloc( psz_text_start, strlen( psz_text_start ) + 1 );
+    char *psz = realloc( psz_text_start, strlen( psz_text_start ) + 1 );
+    if( psz ) psz_text_start = psz;
 
     return psz_text_start;
 }
@@ -635,7 +651,7 @@ static char *StripTags( char *psz_subtitle )
  * to be carrying style information. Over time people have used them that way.
  * In the absence of specifications from which to work, the tags supported
  * have been restricted to the simple set permitted by the USF DTD, ie. :
- *  Basic: <br>, <i>, <b>, <u>
+ *  Basic: <br>, <i>, <b>, <u>, <s>
  *  Extended: <font>
  *    Attributes: face
  *                family
@@ -685,6 +701,7 @@ static char *CreateHtmlSubtitle( int *pi_align, char *psz_subtitle )
     psz_tag[ 0 ] = '\0';
 
     /* */
+    //Oo + 100 ???
     size_t i_buf_size = strlen( psz_subtitle ) + 100;
     char   *psz_html_start = malloc( i_buf_size );
     char   *psz_html = psz_html_start;
@@ -728,6 +745,11 @@ static char *CreateHtmlSubtitle( int *pi_align, char *psz_subtitle )
                 HtmlCopy( &psz_html, &psz_subtitle, "<u>" );
                 strcat( psz_tag, "u" );
             }
+            else if( !strncasecmp( psz_subtitle, "<s>", 3 ) )
+            {
+                HtmlCopy( &psz_html, &psz_subtitle, "<s>" );
+                strcat( psz_tag, "s" );
+            }
             else if( !strncasecmp( psz_subtitle, "<font ", 6 ))
             {
                 const char *psz_attribs[] = { "face=", "family=", "size=",
@@ -777,9 +799,13 @@ static char *CreateHtmlSubtitle( int *pi_align, char *psz_subtitle )
                     if( psz_attribs[ k ] == NULL )
                     {
                         /* Jump over unrecognised tag */
-                        int i_len = strcspn( psz_subtitle, "\"" ) + 1;
-
-                        i_len += strcspn( psz_subtitle + i_len, "\"" ) + 1;
+                        int i_len = strcspn( psz_subtitle, "\"" );
+                        if( psz_subtitle[i_len] == '\"' )
+                        {
+                            i_len += 1 + strcspn( &psz_subtitle[i_len + 1], "\"" );
+                            if( psz_subtitle[i_len] == '\"' )
+                                i_len++;
+                        }
                         psz_subtitle += i_len;
                     }
                     while (*psz_subtitle == ' ')
@@ -813,6 +839,10 @@ static char *CreateHtmlSubtitle( int *pi_align, char *psz_subtitle )
                         b_match = !strncasecmp( psz_subtitle, "</u>", 4 );
                         i_len   = 4;
                         break;
+                    case 's':
+                        b_match = !strncasecmp( psz_subtitle, "</s>", 4 );
+                        i_len   = 4;
+                        break;
                     case 'f':
                         b_match = !strncasecmp( psz_subtitle, "</font>", 7 );
                         i_len   = 7;
@@ -940,6 +970,31 @@ static char *CreateHtmlSubtitle( int *pi_align, char *psz_subtitle )
             /* Hide {\stupidity} */
             psz_subtitle = strchr( psz_subtitle, '}' ) + 1;
         }
+        else if( psz_subtitle[0] == '{' && psz_subtitle[1] == 'Y'
+                && psz_subtitle[2] == ':' && strchr( psz_subtitle, '}' ) )
+        {
+            /* Hide {Y:stupidity} */
+            psz_subtitle = strchr( psz_subtitle, '}' ) + 1;
+        }
+        else if( psz_subtitle[0] == '\\' && psz_subtitle[1] )
+        {
+            if( psz_subtitle[1] == 'N' || psz_subtitle[1] == 'n' )
+            {
+                HtmlPut( &psz_html, "<br/>" );
+                psz_subtitle += 2;
+            }
+            else if( psz_subtitle[1] == 'h' )
+            {
+                /* Non breakable space */
+                HtmlPut( &psz_html, NO_BREAKING_SPACE );
+                psz_subtitle += 2;
+            }
+            else
+            {
+                HtmlPut( &psz_html, "\\" );
+                psz_subtitle++;
+            }
+        }
         else
         {
             *psz_html = *psz_subtitle;
@@ -1000,6 +1055,9 @@ static char *CreateHtmlSubtitle( int *pi_align, char *psz_subtitle )
                 case 'u':
                     HtmlPut( &psz_html, "</u>" );
                     break;
+                case 's':
+                    HtmlPut( &psz_html, "</s>" );
+                    break;
                 case 'f':
                     HtmlPut( &psz_html, "/font>" );
                     break;