1 /* GStreamer tmplayer format subtitle parser
2 * Copyright (C) 2006-2008 Tim-Philipp Müller <tim centricular net>
3 *
4 * This library is free software; you can redistribute it and/or
5 * modify it under the terms of the GNU Library General Public
6 * License as published by the Free Software Foundation; either
7 * version 2 of the License, or (at your option) any later version.
8 *
9 * This library is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
11 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
12 * Library General Public License for more details.
13 *
14 * You should have received a copy of the GNU Library General Public
15 * License along with this library; if not, write to the
16 * Free Software Foundation, Inc., 51 Franklin St, Fifth Floor,
17 * Boston, MA 02110-1301, USA.
18 */
19
20 #include "tmplayerparse.h"
21
22 #include <stdio.h>
23 #include <string.h>
24
25 /* From http://forum.doom9.org/archive/index.php/t-81059.html:
26 *
27 * TMPlayer format, which comes in five varieties:
28 *
29 * time-base 00:00:00:
30 * 00:00:50:This is the Earth at a time|when the dinosaurs roamed...
31 * 00:00:53:
32 * 00:00:54:a lush and fertile planet.
33 * 00:00:56:
34 *
35 * time-base 0:00:00:
36 * 0:00:50:This is the Earth at a time|when the dinosaurs roamed...
37 * 0:00:53:
38 * 0:00:54:a lush and fertile planet.
39 * 0:00:56:
40 *
41 * time-base 00:00:00=
42 * 00:00:50=This is the Earth at a time|when the dinosaurs roamed...
43 * 00:00:53=
44 * 00:00:54=a lush and fertile planet.
45 * 00:00:56=
46 *
47 * time-base 0:00:00=
48 * 0:00:50=This is the Earth at a time|when the dinosaurs roamed...
49 * 0:00:53=
50 * 0:00:54=a lush and fertile planet.
51 * 0:00:56=
52 *
53 * and multiline time-base 00:00:00,1=
54 * 00:00:50,1=This is the Earth at a time
55 * 00:00:50,2=when the dinosaurs roamed...
56 * 00:00:53,1=
57 * 00:00:54,1=a lush and fertile planet.
58 * 00:00:56,1=
59 *
60 * --------------------------------------------------------------------------
61 *
62 * And another variety (which is 'time-base 0:00:00:' but without empty lines):
63 *
64 * 00:00:01:This is the Earth at a time|when the dinosaurs roamed...
65 * 00:00:03:a lush and fertile planet.
66 * 00:00:06:More text here
67 * 00:00:12:Yet another line
68 *
69 */
70
71 static gchar *
tmplayer_process_buffer(ParserState * state)72 tmplayer_process_buffer (ParserState * state)
73 {
74 gchar *ret;
75
76 ret = g_strndup (state->buf->str, state->buf->len);
77 g_strdelimit (ret, "|", '\n');
78 g_string_truncate (state->buf, 0);
79 return ret;
80 }
81
82 static gchar *
tmplayer_parse_line(ParserState * state,const gchar * line,guint line_num)83 tmplayer_parse_line (ParserState * state, const gchar * line, guint line_num)
84 {
85 GstClockTime ts = GST_CLOCK_TIME_NONE;
86 const gchar *text_start = NULL;
87 gchar *ret = NULL;
88 gchar divc = '\0';
89 guint h, m, s, l = 1;
90
91 if (sscanf (line, "%u:%02u:%02u,%u%c", &h, &m, &s, &l, &divc) == 5 &&
92 (divc == '=')) {
93 GST_LOG ("multiline format %u %u %u %u", h, m, s, l);
94 ts = GST_SECOND * ((((h * 60) + m) * 60) + s);
95 text_start = strchr (line, '=');
96 } else if (sscanf (line, "%u:%02u:%02u%c", &h, &m, &s, &divc) == 4 &&
97 (divc == '=' || divc == ':')) {
98 GST_LOG ("single line format %u %u %u %u %c", h, m, s, l, divc);
99 ts = GST_SECOND * ((((h * 60) + m) * 60) + s);
100 text_start = strchr (line + 6, divc);
101 } else if (line[0] == '\0' && state->buf->len > 0 &&
102 GST_CLOCK_TIME_IS_VALID (state->start_time)) {
103 /* if we get an empty line (could be the end of the file, but doesn't have
104 * to be), just push whatever is still in the buffer without a duration */
105 GST_LOG ("empty line, and there's still text in the buffer");
106 ret = tmplayer_process_buffer (state);
107 state->duration = GST_CLOCK_TIME_NONE;
108 return ret;
109 } else {
110 GST_WARNING ("failed to parse line: '%s'", line);
111 return NULL;
112 }
113
114 /* if this is a line without text, or the first line in a multiline file,
115 * process and return the data in the buffer, which is the previous line(s) */
116 if (text_start == NULL || text_start[1] == '\0' ||
117 (l == 1 && state->buf->len > 0)) {
118
119 if (GST_CLOCK_TIME_IS_VALID (state->start_time) &&
120 state->start_time < ts && line_num > 0) {
121 ret = tmplayer_process_buffer (state);
122 state->duration = ts - state->start_time;
123 /* ..and append current line's text (if there is any) for the next round.
124 * We don't have to store ts as pending_start_time, since we deduce the
125 * durations from the start times anyway, so as long as the parser just
126 * forwards state->start_time by duration after it pushes the line we
127 * are about to return it will all be good. */
128 g_string_append (state->buf, text_start + 1);
129 } else if (line_num > 0) {
130 GST_WARNING ("end of subtitle unit but no valid start time?!");
131 }
132 } else {
133 if (l > 1)
134 g_string_append_c (state->buf, '\n');
135 g_string_append (state->buf, text_start + 1);
136 state->start_time = ts;
137 }
138
139 GST_LOG ("returning: '%s'", GST_STR_NULL (ret));
140 return ret;
141 }
142
143 gchar *
parse_tmplayer(ParserState * state,const gchar * line)144 parse_tmplayer (ParserState * state, const gchar * line)
145 {
146 gchar *ret;
147
148 /* GST_LOG ("Parsing: %s", line); */
149
150 ret = tmplayer_parse_line (state, line, state->state);
151 ++state->state;
152
153 return ret;
154 }
155