libavformat/ttmlenc.c

   1 /*
   2  * TTML subtitle muxer
   3  * Copyright (c) 2020 24i
   4  *
   5  * This file is part of FFmpeg.
   6  *
   7  * FFmpeg 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 2.1 of the License, or (at your option) any later version.
  11  *
  12  * FFmpeg 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.
  16  *
  17  * You should have received a copy of the GNU Lesser General Public
  18  * License along with FFmpeg; if not, write to the Free Software
  19  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
  20  */
  21
  22 /**
  23  * @file
  24  * TTML subtitle muxer
  25  * @see https://www.w3.org/TR/ttml1/
  26  * @see https://www.w3.org/TR/ttml2/
  27  * @see https://www.w3.org/TR/ttml-imsc/rec
  28  */
  29
  30 #include "libavutil/avstring.h"
  31 #include "avformat.h"
  32 #include "internal.h"
  33 #include "mux.h"
  34 #include "ttmlenc.h"
  35 #include "libavcodec/ttmlenc.h"
  36 #include "libavutil/internal.h"
  37
  38 enum TTMLPacketType {
  39     PACKET_TYPE_PARAGRAPH,
  40     PACKET_TYPE_DOCUMENT,
  41 };
  42
  43 struct TTMLHeaderParameters {
  44     const char *tt_element_params;
  45     const char *pre_body_elements;
  46 };
  47
  48 typedef struct TTMLMuxContext {
  49     enum TTMLPacketType input_type;
  50     unsigned int document_written;
  51 } TTMLMuxContext;
  52
  53 static const char ttml_header_text[] =
  54 "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n"
  55 "<tt\n"
  56 "%s"
  57 "  xml:lang=\"%s\">\n"
  58 "%s"
  59 "  <body>\n"
  60 "    <div>\n";
  61
  62 static const char ttml_footer_text[] =
  63 "    </div>\n"
  64 "  </body>\n"
  65 "</tt>\n";
  66
  67 static void ttml_write_time(AVIOContext *pb, const char tag[],
  68                             int64_t millisec)
  69 {
  70     int64_t sec, min, hour;
  71     sec = millisec / 1000;
  72     millisec -= 1000 * sec;
  73     min = sec / 60;
  74     sec -= 60 * min;
  75     hour = min / 60;
  76     min -= 60 * hour;
  77
  78     avio_printf(pb, "%s=\"%02"PRId64":%02"PRId64":%02"PRId64".%03"PRId64"\"",
  79                 tag, hour, min, sec, millisec);
  80 }
  81
  82 static int ttml_set_header_values_from_extradata(
  83     AVCodecParameters *par, struct TTMLHeaderParameters *header_params)
  84 {
  85     size_t additional_data_size =
  86         par->extradata_size - TTMLENC_EXTRADATA_SIGNATURE_SIZE;
  87     char *value =
  88         (char *)par->extradata + TTMLENC_EXTRADATA_SIGNATURE_SIZE;
  89     size_t value_size = av_strnlen(value, additional_data_size);
  90     struct TTMLHeaderParameters local_params = { 0 };
  91
  92     if (!additional_data_size) {
  93         // simple case, we don't have to go through local_params and just
  94         // set default fall-back values (for old extradata format).
  95         header_params->tt_element_params = TTML_DEFAULT_NAMESPACING;
  96         header_params->pre_body_elements = "";
  97
  98         return 0;
  99     }
 100
 101     if (value_size == additional_data_size ||
 102         value[value_size] != '\0')
 103         return AVERROR_INVALIDDATA;
 104
 105     local_params.tt_element_params = value;
 106
 107     additional_data_size -= value_size + 1;
 108     value += value_size + 1;
 109     if (!additional_data_size)
 110         return AVERROR_INVALIDDATA;
 111
 112     value_size = av_strnlen(value, additional_data_size);
 113     if (value_size == additional_data_size ||
 114         value[value_size] != '\0')
 115         return AVERROR_INVALIDDATA;
 116
 117     local_params.pre_body_elements = value;
 118
 119     *header_params = local_params;
 120
 121     return 0;
 122 }
 123
 124 static int ttml_write_header(AVFormatContext *ctx)
 125 {
 126     TTMLMuxContext *ttml_ctx = ctx->priv_data;
 127     AVStream    *st = ctx->streams[0];
 128     AVIOContext *pb = ctx->pb;
 129
 130     const AVDictionaryEntry *lang = av_dict_get(st->metadata, "language", NULL,
 131                                                 0);
 132     const char *printed_lang = (lang && lang->value) ? lang->value : "";
 133
 134     ttml_ctx->document_written = 0;
 135     ttml_ctx->input_type = ff_is_ttml_stream_paragraph_based(st->codecpar) ?
 136                            PACKET_TYPE_PARAGRAPH :
 137                            PACKET_TYPE_DOCUMENT;
 138
 139     avpriv_set_pts_info(st, 64, 1, 1000);
 140
 141     if (ttml_ctx->input_type == PACKET_TYPE_PARAGRAPH) {
 142         struct TTMLHeaderParameters header_params;
 143         int ret = ttml_set_header_values_from_extradata(
 144             st->codecpar, &header_params);
 145         if (ret < 0) {
 146             av_log(ctx, AV_LOG_ERROR,
 147                    "Failed to parse TTML header values from extradata: "
 148                    "%s!\n", av_err2str(ret));
 149             return ret;
 150         }
 151
 152         avio_printf(pb, ttml_header_text,
 153                     header_params.tt_element_params,
 154                     printed_lang,
 155                     header_params.pre_body_elements);
 156     }
 157
 158     return 0;
 159 }
 160
 161 static int ttml_write_packet(AVFormatContext *ctx, AVPacket *pkt)
 162 {
 163     TTMLMuxContext *ttml_ctx = ctx->priv_data;
 164     AVIOContext    *pb       = ctx->pb;
 165
 166     switch (ttml_ctx->input_type) {
 167     case PACKET_TYPE_PARAGRAPH:
 168         // write out a paragraph element with the given contents.
 169         avio_printf(pb,     "      <p\n");
 170         ttml_write_time(pb, "        begin", pkt->pts);
 171         avio_w8(pb, '\n');
 172         ttml_write_time(pb, "        end",   pkt->pts + pkt->duration);
 173         avio_printf(pb, ">");
 174         avio_write(pb, pkt->data, pkt->size);
 175         avio_printf(pb, "</p>\n");
 176         break;
 177     case PACKET_TYPE_DOCUMENT:
 178         // dump the given document out as-is.
 179         if (ttml_ctx->document_written) {
 180             av_log(ctx, AV_LOG_ERROR,
 181                    "Attempting to write multiple TTML documents into a "
 182                    "single document! The XML specification forbids this "
 183                    "as there has to be a single root tag.\n");
 184             return AVERROR(EINVAL);
 185         }
 186         avio_write(pb, pkt->data, pkt->size);
 187         ttml_ctx->document_written = 1;
 188         break;
 189     default:
 190         av_log(ctx, AV_LOG_ERROR,
 191                "Internal error: invalid TTML input packet type: %d!\n",
 192                ttml_ctx->input_type);
 193         return AVERROR_BUG;
 194     }
 195
 196     return 0;
 197 }
 198
 199 static int ttml_write_trailer(AVFormatContext *ctx)
 200 {
 201     TTMLMuxContext *ttml_ctx = ctx->priv_data;
 202     AVIOContext    *pb       = ctx->pb;
 203
 204     if (ttml_ctx->input_type == PACKET_TYPE_PARAGRAPH)
 205         avio_printf(pb, ttml_footer_text);
 206
 207     return 0;
 208 }
 209
 210 const FFOutputFormat ff_ttml_muxer = {
 211     .p.name            = "ttml",
 212     .p.long_name       = NULL_IF_CONFIG_SMALL("TTML subtitle"),
 213     .p.extensions      = "ttml",
 214     .p.mime_type       = "text/ttml",
 215     .priv_data_size    = sizeof(TTMLMuxContext),
 216     .p.flags           = AVFMT_GLOBALHEADER | AVFMT_VARIABLE_FPS |
 217                          AVFMT_TS_NONSTRICT,
 218     .p.video_codec     = AV_CODEC_ID_NONE,
 219     .p.audio_codec     = AV_CODEC_ID_NONE,
 220     .p.subtitle_codec  = AV_CODEC_ID_TTML,
 221     .flags_internal    = FF_OFMT_FLAG_MAX_ONE_OF_EACH |
 222                          FF_OFMT_FLAG_ONLY_DEFAULT_CODECS,
 223     .write_header      = ttml_write_header,
 224     .write_packet      = ttml_write_packet,
 225     .write_trailer     = ttml_write_trailer,
 226 };