/src/ffmpeg/libavformat/srtdec.c
Line | Count | Source |
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 "demux.h" |
25 | | #include "internal.h" |
26 | | #include "subtitles.h" |
27 | | #include "libavutil/bprint.h" |
28 | | #include "libavutil/intreadwrite.h" |
29 | | |
30 | | typedef struct { |
31 | | FFDemuxSubtitlesQueue q; |
32 | | } SRTContext; |
33 | | |
34 | | static int srt_probe(const AVProbeData *p) |
35 | 936k | { |
36 | 936k | int v; |
37 | 936k | char buf[64], *pbuf; |
38 | 936k | FFTextReader tr; |
39 | | |
40 | 936k | ff_text_init_buf(&tr, p->buf, p->buf_size); |
41 | | |
42 | 2.89M | while (ff_text_peek_r8(&tr) == '\r' || ff_text_peek_r8(&tr) == '\n') |
43 | 1.96M | ff_text_r8(&tr); |
44 | | |
45 | | /* Check if the first non-empty line is a number. We do not check what the |
46 | | * number is because in practice it can be anything. |
47 | | * Also, that number can be followed by random garbage, so we can not |
48 | | * unfortunately check that we only have a number. */ |
49 | 936k | if (ff_subtitles_read_line(&tr, buf, sizeof(buf)) < 0 || |
50 | 233k | strtol(buf, &pbuf, 10) < 0 || pbuf == buf) |
51 | 917k | return 0; |
52 | | |
53 | | /* Check if the next line matches a SRT timestamp */ |
54 | 19.3k | if (ff_subtitles_read_line(&tr, buf, sizeof(buf)) < 0) |
55 | 3.44k | return 0; |
56 | 15.8k | pbuf = buf; |
57 | 15.8k | if (buf[0] == '-') |
58 | 787 | pbuf++; |
59 | 15.8k | if (pbuf[0] >= '0' && pbuf[0] <= '9' && strstr(buf, " --> ") |
60 | 318 | && sscanf(buf, "%*d:%*d:%*d%*1[,.]%*d --> %*d:%*d:%*d%*1[,.]%d", &v) == 1) |
61 | 107 | return AVPROBE_SCORE_MAX; |
62 | | |
63 | 15.7k | return 0; |
64 | 15.8k | } |
65 | | |
66 | | struct event_info { |
67 | | int32_t x1, x2, y1, y2; |
68 | | int duration; |
69 | | int64_t pts; |
70 | | int64_t pos; |
71 | | }; |
72 | | |
73 | | static int get_event_info(const char *line, struct event_info *ei) |
74 | 1.22M | { |
75 | 1.22M | int hh1, mm1, ss1, ms1; |
76 | 1.22M | int hh2, mm2, ss2, ms2; |
77 | | |
78 | 1.22M | ei->x1 = ei->x2 = ei->y1 = ei->y2 = ei->duration = -1; |
79 | 1.22M | ei->pts = AV_NOPTS_VALUE; |
80 | 1.22M | ei->pos = -1; |
81 | 1.22M | if (sscanf(line, "%d:%d:%d%*1[,.]%d --> %d:%d:%d%*1[,.]%d" |
82 | 1.22M | "%*[ ]X1:%"PRId32" X2:%"PRId32" Y1:%"PRId32" Y2:%"PRId32, |
83 | 1.22M | &hh1, &mm1, &ss1, &ms1, |
84 | 1.22M | &hh2, &mm2, &ss2, &ms2, |
85 | 1.22M | &ei->x1, &ei->x2, &ei->y1, &ei->y2) >= 8) { |
86 | 549k | const int64_t start = (hh1*3600LL + mm1*60LL + ss1) * 1000LL + ms1; |
87 | 549k | const int64_t end = (hh2*3600LL + mm2*60LL + ss2) * 1000LL + ms2; |
88 | 549k | ei->duration = end - start; |
89 | 549k | ei->pts = start; |
90 | 549k | return 0; |
91 | 549k | } |
92 | 673k | return -1; |
93 | 1.22M | } |
94 | | |
95 | | static int add_event(FFDemuxSubtitlesQueue *q, AVBPrint *buf, char *line_cache, |
96 | | const struct event_info *ei, int append_cache) |
97 | 549k | { |
98 | 549k | if (append_cache && line_cache[0]) |
99 | 984 | av_bprintf(buf, "%s\n", line_cache); |
100 | 549k | line_cache[0] = 0; |
101 | 549k | if (!av_bprint_is_complete(buf)) |
102 | 0 | return AVERROR(ENOMEM); |
103 | | |
104 | 1.09M | while (buf->len > 0 && buf->str[buf->len - 1] == '\n') |
105 | 548k | buf->str[--buf->len] = 0; |
106 | | |
107 | 549k | if (buf->len) { |
108 | 548k | AVPacket *sub = ff_subtitles_queue_insert_bprint(q, buf, 0); |
109 | 548k | if (!sub) |
110 | 0 | return AVERROR(ENOMEM); |
111 | 548k | av_bprint_clear(buf); |
112 | 548k | sub->pos = ei->pos; |
113 | 548k | sub->pts = ei->pts; |
114 | 548k | sub->duration = ei->duration; |
115 | 548k | if (ei->x1 != -1) { |
116 | 677 | uint8_t *p = av_packet_new_side_data(sub, AV_PKT_DATA_SUBTITLE_POSITION, 16); |
117 | 677 | if (p) { |
118 | 677 | AV_WL32(p, ei->x1); |
119 | 677 | AV_WL32(p + 4, ei->y1); |
120 | 677 | AV_WL32(p + 8, ei->x2); |
121 | 677 | AV_WL32(p + 12, ei->y2); |
122 | 677 | } |
123 | 677 | } |
124 | 548k | } |
125 | | |
126 | 549k | return 0; |
127 | 549k | } |
128 | | |
129 | | static int srt_read_header(AVFormatContext *s) |
130 | 2.07k | { |
131 | 2.07k | SRTContext *srt = s->priv_data; |
132 | 2.07k | AVBPrint buf; |
133 | 2.07k | AVStream *st = avformat_new_stream(s, NULL); |
134 | 2.07k | int res = 0; |
135 | 2.07k | char line[4096], line_cache[4096]; |
136 | 2.07k | int has_event_info = 0; |
137 | 2.07k | struct event_info ei; |
138 | 2.07k | FFTextReader tr; |
139 | 2.07k | ff_text_init_avio(s, &tr, s->pb); |
140 | | |
141 | 2.07k | if (!st) |
142 | 0 | return AVERROR(ENOMEM); |
143 | 2.07k | avpriv_set_pts_info(st, 64, 1, 1000); |
144 | 2.07k | st->codecpar->codec_type = AVMEDIA_TYPE_SUBTITLE; |
145 | 2.07k | st->codecpar->codec_id = AV_CODEC_ID_SUBRIP; |
146 | | |
147 | 2.07k | av_bprint_init(&buf, 0, AV_BPRINT_SIZE_UNLIMITED); |
148 | | |
149 | 2.07k | line_cache[0] = 0; |
150 | | |
151 | 1.22M | while (!ff_text_eof(&tr)) { |
152 | 1.22M | struct event_info tmp_ei; |
153 | 1.22M | const int64_t pos = ff_text_pos(&tr); |
154 | 1.22M | ptrdiff_t len = ff_subtitles_read_line(&tr, line, sizeof(line)); |
155 | | |
156 | 1.22M | if (len < 0) |
157 | 298 | break; |
158 | | |
159 | 1.22M | if (!len || !line[0]) |
160 | 3.37k | continue; |
161 | | |
162 | 1.22M | if (get_event_info(line, &tmp_ei) < 0) { |
163 | 673k | char *pline; |
164 | | |
165 | 673k | if (!has_event_info) |
166 | 1.95k | continue; |
167 | | |
168 | 671k | if (line_cache[0]) { |
169 | | /* We got some cache and a new line so we assume the cached |
170 | | * line was actually part of the payload */ |
171 | 23.0k | av_bprintf(&buf, "%s\n", line_cache); |
172 | 23.0k | line_cache[0] = 0; |
173 | 23.0k | } |
174 | | |
175 | | /* If the line doesn't start with a number, we assume it's part of |
176 | | * the payload, otherwise is likely an event number preceding the |
177 | | * timing information... but we can't be sure of this yet, so we |
178 | | * cache it */ |
179 | 671k | if (strtol(line, &pline, 10) < 0 || line == pline) |
180 | 645k | av_bprintf(&buf, "%s\n", line); |
181 | 25.6k | else |
182 | 25.6k | strcpy(line_cache, line); |
183 | 671k | } else { |
184 | 549k | if (has_event_info) { |
185 | | /* We have the information of previous event, append it to the |
186 | | * queue. We insert the cached line if and only if the payload |
187 | | * is empty and the cached line is not a standalone number. */ |
188 | 548k | char *pline = NULL; |
189 | 548k | const int standalone_number = strtol(line_cache, &pline, 10) >= 0 && pline && !*pline; |
190 | 548k | res = add_event(&srt->q, &buf, line_cache, &ei, !buf.len && !standalone_number); |
191 | 548k | if (res < 0) |
192 | 0 | goto end; |
193 | 548k | } else { |
194 | 1.45k | has_event_info = 1; |
195 | 1.45k | } |
196 | 549k | tmp_ei.pos = pos; |
197 | 549k | ei = tmp_ei; |
198 | 549k | } |
199 | 1.22M | } |
200 | | |
201 | | /* Append the last event. Here we force the cache to be flushed, because a |
202 | | * trailing number is more likely to be genuine (for example a copyright |
203 | | * date) and not the event index of an inexistent event */ |
204 | 2.07k | if (has_event_info) { |
205 | 1.45k | res = add_event(&srt->q, &buf, line_cache, &ei, 1); |
206 | 1.45k | if (res < 0) |
207 | 0 | goto end; |
208 | 1.45k | } |
209 | | |
210 | 2.07k | ff_subtitles_queue_finalize(s, &srt->q); |
211 | | |
212 | 2.07k | end: |
213 | | av_bprint_finalize(&buf, NULL); |
214 | 2.07k | return res; |
215 | 2.07k | } |
216 | | |
217 | | const FFInputFormat ff_srt_demuxer = { |
218 | | .p.name = "srt", |
219 | | .p.long_name = NULL_IF_CONFIG_SMALL("SubRip subtitle"), |
220 | | .priv_data_size = sizeof(SRTContext), |
221 | | .flags_internal = FF_INFMT_FLAG_INIT_CLEANUP, |
222 | | .read_probe = srt_probe, |
223 | | .read_header = srt_read_header, |
224 | | .read_packet = ff_subtitles_read_packet, |
225 | | .read_seek2 = ff_subtitles_read_seek, |
226 | | .read_close = ff_subtitles_read_close, |
227 | | }; |