1/*
2 * Copyright (c) 2012 Clément Bœsch
3 *
4 * This file is part of FFmpeg.
5 *
6 * FFmpeg is free software; you can redistribute it and/or
7 * modify it under the terms of the GNU Lesser General Public
8 * License as published by the Free Software Foundation; either
9 * version 2.1 of the License, or (at your option) any later version.
10 *
11 * FFmpeg is distributed in the hope that it will be useful,
12 * but WITHOUT ANY WARRANTY; without even the implied warranty of
13 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
14 * Lesser General Public License for more details.
15 *
16 * You should have received a copy of the GNU Lesser General Public
17 * License along with FFmpeg; if not, write to the Free Software
18 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
19 */
20
21/**
22 * @file
23 * WebVTT subtitle demuxer
24 * @see http://dev.w3.org/html5/webvtt/
25 */
26
27#include "avformat.h"
28#include "internal.h"
29#include "subtitles.h"
30#include "libavutil/bprint.h"
31#include "libavutil/intreadwrite.h"
32#include "libavutil/opt.h"
33
34typedef struct {
35    const AVClass *class;
36    FFDemuxSubtitlesQueue q;
37    int kind;
38} WebVTTContext;
39
40static int webvtt_probe(const AVProbeData *p)
41{
42    const uint8_t *ptr = p->buf;
43
44    if (AV_RB24(ptr) == 0xEFBBBF)
45        ptr += 3;  /* skip UTF-8 BOM */
46    if (!strncmp(ptr, "WEBVTT", 6) &&
47        (!ptr[6] || strchr("\n\r\t ", ptr[6])))
48        return AVPROBE_SCORE_MAX;
49    return 0;
50}
51
52static int64_t read_ts(const char *s)
53{
54    int hh, mm, ss, ms;
55    if (sscanf(s, "%u:%u:%u.%u", &hh, &mm, &ss, &ms) == 4) return (hh*3600LL + mm*60LL + ss) * 1000LL + ms;
56    if (sscanf(s,    "%u:%u.%u",      &mm, &ss, &ms) == 3) return (            mm*60LL + ss) * 1000LL + ms;
57    return AV_NOPTS_VALUE;
58}
59
60static int webvtt_read_header(AVFormatContext *s)
61{
62    WebVTTContext *webvtt = s->priv_data;
63    AVBPrint cue;
64    int res = 0;
65    AVStream *st = avformat_new_stream(s, NULL);
66
67    if (!st)
68        return AVERROR(ENOMEM);
69    avpriv_set_pts_info(st, 64, 1, 1000);
70    st->codecpar->codec_type = AVMEDIA_TYPE_SUBTITLE;
71    st->codecpar->codec_id   = AV_CODEC_ID_WEBVTT;
72    st->disposition |= webvtt->kind;
73
74    av_bprint_init(&cue,    0, AV_BPRINT_SIZE_UNLIMITED);
75
76    for (;;) {
77        int i;
78        int64_t pos;
79        AVPacket *sub;
80        const char *p, *identifier, *settings;
81        size_t identifier_len, settings_len;
82        int64_t ts_start, ts_end;
83
84        ff_subtitles_read_chunk(s->pb, &cue);
85
86        if (!cue.len)
87            break;
88
89        p = identifier = cue.str;
90        pos = avio_tell(s->pb);
91
92        /* ignore header chunk */
93        if (!strncmp(p, "\xEF\xBB\xBFWEBVTT", 9) ||
94            !strncmp(p, "WEBVTT", 6) ||
95            !strncmp(p, "NOTE", 4))
96            continue;
97
98        /* optional cue identifier (can be a number like in SRT or some kind of
99         * chaptering id) */
100        for (i = 0; p[i] && p[i] != '\n' && p[i] != '\r'; i++) {
101            if (!strncmp(p + i, "-->", 3)) {
102                identifier = NULL;
103                break;
104            }
105        }
106        if (!identifier)
107            identifier_len = 0;
108        else {
109            identifier_len = strcspn(p, "\r\n");
110            p += identifier_len;
111            if (*p == '\r')
112                p++;
113            if (*p == '\n')
114                p++;
115        }
116
117        /* cue timestamps */
118        if ((ts_start = read_ts(p)) == AV_NOPTS_VALUE)
119            break;
120        if (!(p = strstr(p, "-->")))
121            break;
122        p += 2;
123        do p++; while (*p == ' ' || *p == '\t');
124        if ((ts_end = read_ts(p)) == AV_NOPTS_VALUE)
125            break;
126
127        /* optional cue settings */
128        p += strcspn(p, "\n\r\t ");
129        while (*p == '\t' || *p == ' ')
130            p++;
131        settings = p;
132        settings_len = strcspn(p, "\r\n");
133        p += settings_len;
134        if (*p == '\r')
135            p++;
136        if (*p == '\n')
137            p++;
138
139        /* create packet */
140        sub = ff_subtitles_queue_insert(&webvtt->q, p, strlen(p), 0);
141        if (!sub) {
142            res = AVERROR(ENOMEM);
143            goto end;
144        }
145        sub->pos = pos;
146        sub->pts = ts_start;
147        sub->duration = ts_end - ts_start;
148
149#define SET_SIDE_DATA(name, type) do {                                  \
150    if (name##_len) {                                                   \
151        uint8_t *buf = av_packet_new_side_data(sub, type, name##_len);  \
152        if (!buf) {                                                     \
153            res = AVERROR(ENOMEM);                                      \
154            goto end;                                                   \
155        }                                                               \
156        memcpy(buf, name, name##_len);                                  \
157    }                                                                   \
158} while (0)
159
160        SET_SIDE_DATA(identifier, AV_PKT_DATA_WEBVTT_IDENTIFIER);
161        SET_SIDE_DATA(settings,   AV_PKT_DATA_WEBVTT_SETTINGS);
162    }
163
164    ff_subtitles_queue_finalize(s, &webvtt->q);
165
166end:
167    av_bprint_finalize(&cue,    NULL);
168    return res;
169}
170
171static int webvtt_read_packet(AVFormatContext *s, AVPacket *pkt)
172{
173    WebVTTContext *webvtt = s->priv_data;
174    return ff_subtitles_queue_read_packet(&webvtt->q, pkt);
175}
176
177static int webvtt_read_seek(AVFormatContext *s, int stream_index,
178                            int64_t min_ts, int64_t ts, int64_t max_ts, int flags)
179{
180    WebVTTContext *webvtt = s->priv_data;
181    return ff_subtitles_queue_seek(&webvtt->q, s, stream_index,
182                                   min_ts, ts, max_ts, flags);
183}
184
185static int webvtt_read_close(AVFormatContext *s)
186{
187    WebVTTContext *webvtt = s->priv_data;
188    ff_subtitles_queue_clean(&webvtt->q);
189    return 0;
190}
191
192#define OFFSET(x) offsetof(WebVTTContext, x)
193#define KIND_FLAGS AV_OPT_FLAG_SUBTITLE_PARAM|AV_OPT_FLAG_DECODING_PARAM
194
195static const AVOption options[] = {
196    { "kind", "Set kind of WebVTT track", OFFSET(kind), AV_OPT_TYPE_INT, { .i64 = 0 }, 0, INT_MAX, KIND_FLAGS, "webvtt_kind" },
197        { "subtitles",    "WebVTT subtitles kind",    0, AV_OPT_TYPE_CONST, { .i64 = 0 },                           INT_MIN, INT_MAX, KIND_FLAGS, "webvtt_kind" },
198        { "captions",     "WebVTT captions kind",     0, AV_OPT_TYPE_CONST, { .i64 = AV_DISPOSITION_CAPTIONS },     INT_MIN, INT_MAX, KIND_FLAGS, "webvtt_kind" },
199        { "descriptions", "WebVTT descriptions kind", 0, AV_OPT_TYPE_CONST, { .i64 = AV_DISPOSITION_DESCRIPTIONS }, INT_MIN, INT_MAX, KIND_FLAGS, "webvtt_kind" },
200        { "metadata",     "WebVTT metadata kind",     0, AV_OPT_TYPE_CONST, { .i64 = AV_DISPOSITION_METADATA },     INT_MIN, INT_MAX, KIND_FLAGS, "webvtt_kind" },
201    { NULL }
202};
203
204static const AVClass webvtt_demuxer_class = {
205    .class_name  = "WebVTT demuxer",
206    .item_name   = av_default_item_name,
207    .option      = options,
208    .version     = LIBAVUTIL_VERSION_INT,
209};
210
211const AVInputFormat ff_webvtt_demuxer = {
212    .name           = "webvtt",
213    .long_name      = NULL_IF_CONFIG_SMALL("WebVTT subtitle"),
214    .priv_data_size = sizeof(WebVTTContext),
215    .flags_internal = FF_FMT_INIT_CLEANUP,
216    .read_probe     = webvtt_probe,
217    .read_header    = webvtt_read_header,
218    .read_packet    = webvtt_read_packet,
219    .read_seek2     = webvtt_read_seek,
220    .read_close     = webvtt_read_close,
221    .extensions     = "vtt",
222    .priv_class     = &webvtt_demuxer_class,
223};
224