blob: 56bd0c43f69020502b88e859a9d47b86c9791cd8
1 | /* |
2 | * SubRip subtitle demuxer |
3 | * Copyright (c) 2010 Aurelien Jacobs <aurel@gnuage.org> |
4 | * Copyright (c) 2015 Clément Bœsch <u pkh me> |
5 | * |
6 | * This file is part of FFmpeg. |
7 | * |
8 | * FFmpeg is free software; you can redistribute it and/or |
9 | * modify it under the terms of the GNU Lesser General Public |
10 | * License as published by the Free Software Foundation; either |
11 | * version 2.1 of the License, or (at your option) any later version. |
12 | * |
13 | * FFmpeg is distributed in the hope that it will be useful, |
14 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
15 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
16 | * Lesser General Public License for more details. |
17 | * |
18 | * You should have received a copy of the GNU Lesser General Public |
19 | * License along with FFmpeg; if not, write to the Free Software |
20 | * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA |
21 | */ |
22 | |
23 | #include "avformat.h" |
24 | #include "internal.h" |
25 | #include "subtitles.h" |
26 | #include "libavutil/bprint.h" |
27 | #include "libavutil/intreadwrite.h" |
28 | |
29 | typedef struct { |
30 | FFDemuxSubtitlesQueue q; |
31 | } SRTContext; |
32 | |
33 | static int srt_probe(AVProbeData *p) |
34 | { |
35 | int v; |
36 | char buf[64], *pbuf; |
37 | FFTextReader tr; |
38 | |
39 | ff_text_init_buf(&tr, p->buf, p->buf_size); |
40 | |
41 | while (ff_text_peek_r8(&tr) == '\r' || ff_text_peek_r8(&tr) == '\n') |
42 | ff_text_r8(&tr); |
43 | |
44 | /* Check if the first non-empty line is a number. We do not check what the |
45 | * number is because in practice it can be anything. |
46 | * Also, that number can be followed by random garbage, so we can not |
47 | * unfortunately check that we only have a number. */ |
48 | if (ff_subtitles_read_line(&tr, buf, sizeof(buf)) < 0 || |
49 | strtol(buf, &pbuf, 10) < 0 || pbuf == buf) |
50 | return 0; |
51 | |
52 | /* Check if the next line matches a SRT timestamp */ |
53 | if (ff_subtitles_read_line(&tr, buf, sizeof(buf)) < 0) |
54 | return 0; |
55 | pbuf = buf; |
56 | if (buf[0] == '-') |
57 | pbuf++; |
58 | if (pbuf[0] >= '0' && pbuf[0] <= '9' && strstr(buf, " --> ") |
59 | && sscanf(buf, "%*d:%*d:%*d%*1[,.]%*d --> %*d:%*d:%*d%*1[,.]%d", &v) == 1) |
60 | return AVPROBE_SCORE_MAX; |
61 | |
62 | return 0; |
63 | } |
64 | |
65 | struct event_info { |
66 | int32_t x1, x2, y1, y2; |
67 | int duration; |
68 | int64_t pts; |
69 | int64_t pos; |
70 | }; |
71 | |
72 | static int get_event_info(const char *line, struct event_info *ei) |
73 | { |
74 | int hh1, mm1, ss1, ms1; |
75 | int hh2, mm2, ss2, ms2; |
76 | |
77 | ei->x1 = ei->x2 = ei->y1 = ei->y2 = ei->duration = -1; |
78 | ei->pts = AV_NOPTS_VALUE; |
79 | ei->pos = -1; |
80 | if (sscanf(line, "%d:%d:%d%*1[,.]%d --> %d:%d:%d%*1[,.]%d" |
81 | "%*[ ]X1:%"PRId32" X2:%"PRId32" Y1:%"PRId32" Y2:%"PRId32, |
82 | &hh1, &mm1, &ss1, &ms1, |
83 | &hh2, &mm2, &ss2, &ms2, |
84 | &ei->x1, &ei->x2, &ei->y1, &ei->y2) >= 8) { |
85 | const int64_t start = (hh1*3600LL + mm1*60LL + ss1) * 1000LL + ms1; |
86 | const int64_t end = (hh2*3600LL + mm2*60LL + ss2) * 1000LL + ms2; |
87 | ei->duration = end - start; |
88 | ei->pts = start; |
89 | return 0; |
90 | } |
91 | return -1; |
92 | } |
93 | |
94 | static int add_event(FFDemuxSubtitlesQueue *q, AVBPrint *buf, char *line_cache, |
95 | const struct event_info *ei, int append_cache) |
96 | { |
97 | if (append_cache && line_cache[0]) |
98 | av_bprintf(buf, "%s\n", line_cache); |
99 | line_cache[0] = 0; |
100 | |
101 | while (buf->len > 0 && buf->str[buf->len - 1] == '\n') |
102 | buf->str[--buf->len] = 0; |
103 | |
104 | if (buf->len) { |
105 | AVPacket *sub = ff_subtitles_queue_insert(q, buf->str, buf->len, 0); |
106 | if (!sub) |
107 | return AVERROR(ENOMEM); |
108 | av_bprint_clear(buf); |
109 | sub->pos = ei->pos; |
110 | sub->pts = ei->pts; |
111 | sub->duration = ei->duration; |
112 | if (ei->x1 != -1) { |
113 | uint8_t *p = av_packet_new_side_data(sub, AV_PKT_DATA_SUBTITLE_POSITION, 16); |
114 | if (p) { |
115 | AV_WL32(p, ei->x1); |
116 | AV_WL32(p + 4, ei->y1); |
117 | AV_WL32(p + 8, ei->x2); |
118 | AV_WL32(p + 12, ei->y2); |
119 | } |
120 | } |
121 | } |
122 | |
123 | return 0; |
124 | } |
125 | |
126 | static int srt_read_header(AVFormatContext *s) |
127 | { |
128 | SRTContext *srt = s->priv_data; |
129 | AVBPrint buf; |
130 | AVStream *st = avformat_new_stream(s, NULL); |
131 | int res = 0; |
132 | char line[4096], line_cache[4096]; |
133 | int has_event_info = 0; |
134 | struct event_info ei; |
135 | FFTextReader tr; |
136 | ff_text_init_avio(s, &tr, s->pb); |
137 | |
138 | if (!st) |
139 | return AVERROR(ENOMEM); |
140 | avpriv_set_pts_info(st, 64, 1, 1000); |
141 | st->codecpar->codec_type = AVMEDIA_TYPE_SUBTITLE; |
142 | st->codecpar->codec_id = AV_CODEC_ID_SUBRIP; |
143 | |
144 | av_bprint_init(&buf, 0, AV_BPRINT_SIZE_UNLIMITED); |
145 | |
146 | line_cache[0] = 0; |
147 | |
148 | while (!ff_text_eof(&tr)) { |
149 | struct event_info tmp_ei; |
150 | const int64_t pos = ff_text_pos(&tr); |
151 | ptrdiff_t len = ff_subtitles_read_line(&tr, line, sizeof(line)); |
152 | |
153 | if (len < 0) |
154 | break; |
155 | |
156 | if (!len || !line[0]) |
157 | continue; |
158 | |
159 | if (get_event_info(line, &tmp_ei) < 0) { |
160 | char *pline; |
161 | |
162 | if (!has_event_info) |
163 | continue; |
164 | |
165 | if (line_cache[0]) { |
166 | /* We got some cache and a new line so we assume the cached |
167 | * line was actually part of the payload */ |
168 | av_bprintf(&buf, "%s\n", line_cache); |
169 | line_cache[0] = 0; |
170 | } |
171 | |
172 | /* If the line doesn't start with a number, we assume it's part of |
173 | * the payload, otherwise is likely an event number preceding the |
174 | * timing information... but we can't be sure of this yet, so we |
175 | * cache it */ |
176 | if (strtol(line, &pline, 10) < 0 || line == pline) |
177 | av_bprintf(&buf, "%s\n", line); |
178 | else |
179 | strcpy(line_cache, line); |
180 | } else { |
181 | if (has_event_info) { |
182 | /* We have the information of previous event, append it to the |
183 | * queue. We insert the cached line if and only if the payload |
184 | * is empty and the cached line is not a standalone number. */ |
185 | char *pline = NULL; |
186 | const int standalone_number = strtol(line_cache, &pline, 10) >= 0 && pline && !*pline; |
187 | res = add_event(&srt->q, &buf, line_cache, &ei, !buf.len && !standalone_number); |
188 | if (res < 0) |
189 | goto end; |
190 | } else { |
191 | has_event_info = 1; |
192 | } |
193 | tmp_ei.pos = pos; |
194 | ei = tmp_ei; |
195 | } |
196 | } |
197 | |
198 | /* Append the last event. Here we force the cache to be flushed, because a |
199 | * trailing number is more likely to be geniune (for example a copyright |
200 | * date) and not the event index of an inexistant event */ |
201 | if (has_event_info) { |
202 | res = add_event(&srt->q, &buf, line_cache, &ei, 1); |
203 | if (res < 0) |
204 | goto end; |
205 | } |
206 | |
207 | ff_subtitles_queue_finalize(s, &srt->q); |
208 | |
209 | end: |
210 | av_bprint_finalize(&buf, NULL); |
211 | return res; |
212 | } |
213 | |
214 | static int srt_read_packet(AVFormatContext *s, AVPacket *pkt) |
215 | { |
216 | SRTContext *srt = s->priv_data; |
217 | return ff_subtitles_queue_read_packet(&srt->q, pkt); |
218 | } |
219 | |
220 | static int srt_read_seek(AVFormatContext *s, int stream_index, |
221 | int64_t min_ts, int64_t ts, int64_t max_ts, int flags) |
222 | { |
223 | SRTContext *srt = s->priv_data; |
224 | return ff_subtitles_queue_seek(&srt->q, s, stream_index, |
225 | min_ts, ts, max_ts, flags); |
226 | } |
227 | |
228 | static int srt_read_close(AVFormatContext *s) |
229 | { |
230 | SRTContext *srt = s->priv_data; |
231 | ff_subtitles_queue_clean(&srt->q); |
232 | return 0; |
233 | } |
234 | |
235 | AVInputFormat ff_srt_demuxer = { |
236 | .name = "srt", |
237 | .long_name = NULL_IF_CONFIG_SMALL("SubRip subtitle"), |
238 | .priv_data_size = sizeof(SRTContext), |
239 | .read_probe = srt_probe, |
240 | .read_header = srt_read_header, |
241 | .read_packet = srt_read_packet, |
242 | .read_seek2 = srt_read_seek, |
243 | .read_close = srt_read_close, |
244 | }; |
245 |