sub_cc.c

   1 /*
   2  * decoder for Closed Captions
   3  *
   4  * This decoder relies on MPlayer's OSD to display subtitles.
   5  * Be warned that decoding is somewhat preliminary, though it basically works.
   6  *
   7  * Most notably, only the text information is decoded as of now, discarding
   8  * color, background and position info (see source below).
   9  *
  10  * uses source from the xine closed captions decoder
  11  *
  12  * Copyright (C) 2002 Matteo Giani
  13  *
  14  * This file is part of MPlayer.
  15  *
  16  * MPlayer is free software; you can redistribute it and/or modify
  17  * it under the terms of the GNU General Public License as published by
  18  * the Free Software Foundation; either version 2 of the License, or
  19  * (at your option) any later version.
  20  *
  21  * MPlayer is distributed in the hope that it will be useful,
  22  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  23  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  24  * GNU General Public License for more details.
  25  *
  26  * You should have received a copy of the GNU General Public License along
  27  * with MPlayer; if not, write to the Free Software Foundation, Inc.,
  28  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
  29  */
  30
  31 #include <stdio.h>
  32 #include <stdlib.h>
  33 #include <string.h>
  34
  35 #include "config.h"
  36 #include "sub_cc.h"
  37
  38 #include "subreader.h"
  39
  40 #include "libvo/video_out.h"
  41 #include "libvo/sub.h"
  42
  43
  44 #define CC_MAX_LINE_LENGTH 64
  45
  46 static char chartbl[128];
  47
  48 static subtitle buf1,buf2;
  49 static subtitle *fb,*bb;
  50
  51 static unsigned int cursor_pos=0;
  52
  53 static int initialized=0;
  54
  55 #define CC_ROLLON 1
  56 #define CC_ROLLUP 2
  57
  58 static int cc_mode=CC_ROLLON;
  59 static int cc_lines=4; ///< number of visible rows in CC roll-up mode, not used in CC roll-on mode
  60
  61 static void build_char_table(void)
  62 {
  63   int i;
  64   /* first the normal ASCII codes */
  65   for (i = 0; i < 128; i++)
  66     chartbl[i] = (char) i;
  67   /* now the special codes */
  68   chartbl[0x2a] = 'á';
  69   chartbl[0x5c] = 'é';
  70   chartbl[0x5e] = 'í';
  71   chartbl[0x5f] = 'ó';
  72   chartbl[0x60] = 'ú';
  73   chartbl[0x7b] = 'ç';
  74   chartbl[0x7c] = '÷';
  75   chartbl[0x7d] = 'Ñ';
  76   chartbl[0x7e] = 'ñ';
  77   chartbl[0x7f] = '¤';    /* FIXME: this should be a solid block */
  78 }
  79
  80 static void clear_buffer(subtitle *buf)
  81 {
  82         int i;
  83         buf->lines=0;
  84         for (i = 0; i < SUB_MAX_TEXT; i++) {
  85                 free(buf->text[i]);
  86                 buf->text[i] = NULL;
  87         }
  88 }
  89
  90
  91 /**
  92  \brief scroll buffer one line up
  93  \param buf buffer to scroll
  94 */
  95 static void scroll_buffer(subtitle* buf)
  96 {
  97         int i;
  98
  99         while(buf->lines > cc_lines)
 100         {
 101                 free(buf->text[0]);
 102
 103                 for(i = 0; i < buf->lines - 1; i++) buf->text[i] = buf->text[i+1];
 104
 105                 buf->text[buf->lines-1] = NULL;
 106                 buf->lines--;
 107         }
 108 }
 109
 110 static int channel;
 111
 112 void subcc_init(void)
 113 {
 114         int i;
 115         //printf("subcc_init(): initing...\n");
 116         build_char_table();
 117         for(i=0;i<SUB_MAX_TEXT;i++) {buf1.text[i]=buf2.text[i]=NULL;}
 118         buf1.lines=buf2.lines=0;
 119         fb=&buf1;
 120         bb=&buf2;
 121         channel = -1;
 122
 123         initialized=1;
 124 }
 125
 126
 127 static void display_buffer(subtitle *buf)
 128 {
 129         vo_sub = buf;
 130         vo_osd_changed(OSDTYPE_SUBTITLE);
 131 }
 132
 133
 134 static void append_char(char c)
 135 {
 136         if(!bb->lines) {bb->lines++; cursor_pos=0;}
 137         if(bb->text[bb->lines - 1]==NULL)
 138         {
 139                 bb->text[bb->lines - 1] = calloc(1, CC_MAX_LINE_LENGTH);
 140                 cursor_pos=0;
 141         }
 142
 143         if(c=='\n')
 144         {
 145                 if(cursor_pos>0 && bb->lines < SUB_MAX_TEXT)
 146                 {
 147                         bb->lines++;cursor_pos=0;
 148                         if(cc_mode==CC_ROLLUP){ //Carriage return - scroll buffer one line up
 149                                 bb->text[bb->lines - 1]=calloc(1, CC_MAX_LINE_LENGTH);
 150                                 scroll_buffer(bb);
 151                         }
 152                 }
 153         }
 154         else
 155         {
 156                 if(cursor_pos==CC_MAX_LINE_LENGTH-1)
 157                 {
 158                         fprintf(stderr,"CC: append_char() reached CC_MAX_LINE_LENGTH!\n");
 159                         return;
 160                 }
 161                 bb->text[bb->lines - 1][cursor_pos++]=c;
 162         }
 163         //In CC roll-up mode data should be shown immediately
 164         if(cc_mode==CC_ROLLUP) display_buffer(bb);
 165 }
 166
 167
 168 static void swap_buffers(void)
 169 {
 170         subtitle *foo;
 171         foo=fb;
 172         fb=bb;
 173         bb=foo;
 174 }
 175
 176 static int selected_channel(void)
 177 {
 178     return subcc_enabled - 1;
 179 }
 180
 181 static void cc_decode_EIA608(unsigned short int data)
 182 {
 183
 184   static unsigned short int lastcode=0x0000;
 185   unsigned char c1 = data & 0x7f;
 186   unsigned char c2 = (data >> 8) & 0x7f;
 187
 188   if (c1 & 0x60) {              /* normal character, 0x20 <= c1 <= 0x7f */
 189            if (channel != (selected_channel() & 1))
 190                    return;
 191            append_char(chartbl[c1]);
 192            if(c2 & 0x60)        /*c2 might not be a normal char even if c1 is*/
 193                    append_char(chartbl[c2]);
 194   }
 195   else if (c1 & 0x10)           // control code / special char
 196   {
 197           channel = (c1 & 0x08) >> 3;
 198           if (channel != (selected_channel() & 1))
 199                 return;
 200           c1&=~0x08;
 201           if(data!=lastcode)
 202           {
 203                 if(c2 & 0x40) { /*PAC, Preamble Address Code */
 204                         append_char('\n'); /*FIXME properly interpret PACs*/
 205                 }
 206                 else
 207                         switch(c1)
 208                         {
 209                                 case 0x10:      break; // ext attribute
 210                                 case 0x11:
 211                                         if((c2 & 0x30)==0x30)
 212                                         {
 213                                                 //printf("[debug]:Special char (ignored)\n");
 214                                                 /*cc_decode_special_char()*/;
 215                                         }
 216                                         else if (c2 & 0x20)
 217                                         {
 218                                                 //printf("[debug]: midrow_attr (ignored)\n");
 219                                                 /*cc_decode_midrow_attr()*/;
 220                                         }
 221                                         break;
 222                                 case 0x14:
 223                                         switch(c2)
 224                                         {
 225                                                 case 0x00: //CC roll-on mode
 226                                                            cc_mode=CC_ROLLON;
 227                                                            break;
 228                                                 case 0x25: //CC roll-up, 2 rows
 229                                                 case 0x26: //CC roll-up, 3 rows
 230                                                 case 0x27: //CC roll-up, 4 rows
 231                                                            cc_lines=c2-0x23;
 232                                                            cc_mode=CC_ROLLUP;
 233                                                            break;
 234                                                 case 0x2C: display_buffer(NULL); //EDM
 235                                                            clear_buffer(fb); break;
 236                                                 case 0x2d: append_char('\n');   //carriage return
 237                                                            break;
 238                                                 case 0x2e: clear_buffer(bb);    //ENM
 239                                                            break;
 240                                                 case 0x2f: swap_buffers();      //Swap buffers
 241                                                            display_buffer(fb);
 242                                                            clear_buffer(bb);
 243                                                            break;
 244                                         }
 245                                         break;
 246                                 case 0x17:
 247                                         if( c2>=0x21 && c2<=0x23) //TAB
 248                                         {
 249                                                 break;
 250                                         }
 251                         }
 252           }
 253   }
 254   lastcode=data;
 255 }
 256
 257 static void subcc_decode(unsigned char *inputbuffer, unsigned int inputlength)
 258 {
 259   /* The first number may denote a channel number. I don't have the
 260    * EIA-708 standard, so it is hard to say.
 261    * From what I could figure out so far, the general format seems to be:
 262    *
 263    * repeat
 264    *
 265    *   0xfe starts 2 byte sequence of unknown purpose. It might denote
 266    *        field #2 in line 21 of the VBI.
 267    *        Treating it identical of 0xff fixes
 268    *        http://samples.mplayerhq.hu/MPEG-VOB/ClosedCaptions/Starship_Troopers.vob
 269    *
 270    *   0xff starts 2 byte EIA-608 sequence, field #1 in line 21 of the VBI.
 271    *        Followed by a 3-code triplet that starts either with 0xff or
 272    *        0xfe. In either case, the following triplet needs to be ignored
 273    *        for line 21, field 1.
 274    *
 275    *   0x00 is padding, followed by 2 more 0x00.
 276    *
 277    *   0x01 always seems to appear at the beginning, always seems to
 278    *        be followed by 0xf8, 8-bit number.
 279    *        The lower 7 bits of this 8-bit number seem to denote the
 280    *        number of code triplets that follow.
 281    *        The most significant bit denotes whether the Line 21 field 1
 282    *        captioning information is at odd or even triplet offsets from this
 283    *        beginning triplet. 1 denotes odd offsets, 0 denotes even offsets.
 284    *
 285    *        Most captions are encoded with odd offsets, so this is what we
 286    *        will assume.
 287    *
 288    * until end of packet
 289    */
 290   unsigned char *current = inputbuffer;
 291   unsigned int curbytes = 0;
 292   unsigned char data1, data2;
 293   unsigned char cc_code;
 294   int odd_offset = 1;
 295
 296   while (curbytes < inputlength) {
 297     cc_code = current[0];
 298
 299     if (inputlength - curbytes < 2) {
 300 #ifdef LOG_DEBUG
 301       fprintf(stderr, "Not enough data for 2-byte CC encoding\n");
 302 #endif
 303       break;
 304     }
 305
 306     data1 = current[1];
 307     data2 = current[2];
 308     current += 3; curbytes += 3;
 309
 310     switch (cc_code) {
 311     case 0xfe:
 312     case 0xff:
 313       odd_offset ^= 1;
 314       if (odd_offset != selected_channel() >> 1)
 315           break;
 316       /* expect EIA-608 CC1/CC2 encoding */
 317       // FIXME check parity!
 318       // Parity check omitted assuming we are reading from a DVD and therefore
 319       // we should encounter no "transmission errors".
 320       cc_decode_EIA608(data1 | (data2 << 8));
 321       break;
 322
 323     case 0x00:
 324       /* This seems to be just padding */
 325       break;
 326
 327     case 0x01:
 328       odd_offset = data2 >> 7;
 329       break;
 330
 331     default:
 332 //#ifdef LOG_DEBUG
 333       fprintf(stderr, "Unknown CC encoding: %x\n", cc_code);
 334 //#endif
 335       break;
 336     }
 337   }
 338 }
 339
 340
 341 void subcc_process_data(unsigned char *inputdata,unsigned int len)
 342 {
 343         if(!subcc_enabled) return;
 344         if(!initialized) subcc_init();
 345
 346         subcc_decode(inputdata, len);
 347 }