2 * Copyright (C) <1999> Erik Walthinsen <omega@cse.ogi.edu>
3 * Copyright (c) 2004 Ronald S. Bultje <rbultje@ronald.bitfreak.net>
5 * This library is free software; you can redistribute it and/or
6 * modify it under the terms of the GNU Library General Public
7 * License as published by the Free Software Foundation; either
8 * version 2 of the License, or (at your option) any later version.
10 * This library is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * Library General Public License for more details.
15 * You should have received a copy of the GNU Library General Public
16 * License along with this library; if not, write to the
17 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
18 * Boston, MA 02111-1307, USA.
27 #include <sys/types.h>
30 #include "gstsubparse.h"
31 #include "gstssaparse.h"
32 #include "samiparse.h"
34 GST_DEBUG_CATEGORY_STATIC (sub_parse_debug);
35 #define GST_CAT_DEFAULT sub_parse_debug
37 #define DEFAULT_ENCODING NULL
46 gst_sub_parse_set_property (GObject * object, guint prop_id,
47 const GValue * value, GParamSpec * pspec);
49 gst_sub_parse_get_property (GObject * object, guint prop_id,
50 GValue * value, GParamSpec * pspec);
53 static const GstElementDetails sub_parse_details =
54 GST_ELEMENT_DETAILS ("Subtitle parser",
55 "Codec/Parser/Subtitle",
56 "Parses subtitle (.sub) files into text streams",
57 "Gustavo J. A. M. Carneiro <gjc@inescporto.pt>\n"
58 "Ronald S. Bultje <rbultje@ronald.bitfreak.net>");
60 #ifndef GST_DISABLE_LOADSAVE_REGISTRY
61 static GstStaticPadTemplate sink_templ = GST_STATIC_PAD_TEMPLATE ("sink",
64 GST_STATIC_CAPS ("application/x-subtitle; application/x-subtitle-sami")
67 static GstStaticPadTemplate sink_templ = GST_STATIC_PAD_TEMPLATE ("sink",
70 GST_STATIC_CAPS ("application/x-subtitle")
74 static GstStaticPadTemplate src_templ = GST_STATIC_PAD_TEMPLATE ("src",
77 GST_STATIC_CAPS ("text/plain; text/x-pango-markup")
80 static void gst_sub_parse_base_init (GstSubParseClass * klass);
81 static void gst_sub_parse_class_init (GstSubParseClass * klass);
82 static void gst_sub_parse_init (GstSubParse * subparse);
84 static gboolean gst_sub_parse_src_event (GstPad * pad, GstEvent * event);
85 static gboolean gst_sub_parse_sink_event (GstPad * pad, GstEvent * event);
87 static GstStateChangeReturn gst_sub_parse_change_state (GstElement * element,
88 GstStateChange transition);
90 static GstFlowReturn gst_sub_parse_chain (GstPad * sinkpad, GstBuffer * buf);
92 static GstElementClass *parent_class = NULL;
95 gst_sub_parse_get_type (void)
97 static GType sub_parse_type = 0;
99 if (!sub_parse_type) {
100 static const GTypeInfo sub_parse_info = {
101 sizeof (GstSubParseClass),
102 (GBaseInitFunc) gst_sub_parse_base_init,
104 (GClassInitFunc) gst_sub_parse_class_init,
107 sizeof (GstSubParse),
109 (GInstanceInitFunc) gst_sub_parse_init,
112 sub_parse_type = g_type_register_static (GST_TYPE_ELEMENT,
113 "GstSubParse", &sub_parse_info, 0);
116 return sub_parse_type;
120 gst_sub_parse_base_init (GstSubParseClass * klass)
122 GstElementClass *element_class = GST_ELEMENT_CLASS (klass);
124 gst_element_class_add_pad_template (element_class,
125 gst_static_pad_template_get (&sink_templ));
126 gst_element_class_add_pad_template (element_class,
127 gst_static_pad_template_get (&src_templ));
128 gst_element_class_set_details (element_class, &sub_parse_details);
132 gst_sub_parse_dispose (GObject * object)
134 GstSubParse *subparse = GST_SUBPARSE (object);
136 GST_DEBUG_OBJECT (subparse, "cleaning up subtitle parser");
138 if (subparse->segment) {
139 gst_segment_free (subparse->segment);
140 subparse->segment = NULL;
142 if (subparse->encoding) {
143 g_free (subparse->encoding);
144 subparse->encoding = NULL;
146 if (subparse->textbuf) {
147 g_string_free (subparse->textbuf, TRUE);
148 subparse->textbuf = NULL;
150 sami_context_deinit (&subparse->state);
152 GST_CALL_PARENT (G_OBJECT_CLASS, dispose, (object));
156 gst_sub_parse_class_init (GstSubParseClass * klass)
158 GObjectClass *object_class = G_OBJECT_CLASS (klass);
159 GstElementClass *element_class = GST_ELEMENT_CLASS (klass);
161 parent_class = g_type_class_peek_parent (klass);
163 object_class->dispose = gst_sub_parse_dispose;
164 object_class->set_property = gst_sub_parse_set_property;
165 object_class->get_property = gst_sub_parse_get_property;
167 element_class->change_state = gst_sub_parse_change_state;
169 g_object_class_install_property (object_class, PROP_ENCODING,
170 g_param_spec_string ("subtitle-encoding", "subtitle charset encoding",
171 "Encoding to assume if input subtitles are not in UTF-8 encoding. "
172 "If not set, the GST_SUBTITLE_ENCODING environment variable will "
173 "be checked for an encoding to use. If that is not set either, "
174 "ISO-8859-15 will be assumed.", DEFAULT_ENCODING, G_PARAM_READWRITE));
178 gst_sub_parse_init (GstSubParse * subparse)
180 subparse->sinkpad = gst_pad_new_from_static_template (&sink_templ, "sink");
181 gst_pad_set_chain_function (subparse->sinkpad,
182 GST_DEBUG_FUNCPTR (gst_sub_parse_chain));
183 gst_pad_set_event_function (subparse->sinkpad,
184 GST_DEBUG_FUNCPTR (gst_sub_parse_sink_event));
185 gst_element_add_pad (GST_ELEMENT (subparse), subparse->sinkpad);
187 subparse->srcpad = gst_pad_new_from_static_template (&src_templ, "src");
188 gst_pad_set_event_function (subparse->srcpad,
189 GST_DEBUG_FUNCPTR (gst_sub_parse_src_event));
190 gst_element_add_pad (GST_ELEMENT (subparse), subparse->srcpad);
192 subparse->textbuf = g_string_new (NULL);
193 subparse->parser_type = GST_SUB_PARSE_FORMAT_UNKNOWN;
194 subparse->flushing = FALSE;
195 subparse->segment = gst_segment_new ();
196 if (subparse->segment) {
197 gst_segment_init (subparse->segment, GST_FORMAT_TIME);
198 subparse->need_segment = TRUE;
200 GST_WARNING_OBJECT (subparse, "segment creation failed");
201 g_assert_not_reached ();
203 subparse->encoding = g_strdup (DEFAULT_ENCODING);
207 * Source pad functions.
211 gst_sub_parse_src_event (GstPad * pad, GstEvent * event)
213 GstSubParse *self = GST_SUBPARSE (gst_pad_get_parent (pad));
214 gboolean ret = FALSE;
216 GST_DEBUG ("Handling %s event", GST_EVENT_TYPE_NAME (event));
218 switch (GST_EVENT_TYPE (event)) {
222 GstSeekType start_type, stop_type;
227 gst_event_parse_seek (event, &rate, &format, &self->segment_flags,
228 &start_type, &start, &stop_type, &stop);
230 if (format != GST_FORMAT_TIME) {
231 GST_WARNING_OBJECT (self, "we only support seeking in TIME format");
232 gst_event_unref (event);
236 /* Convert that seek to a seeking in bytes at position 0,
237 FIXME: could use an index */
238 ret = gst_pad_push_event (self->sinkpad,
239 gst_event_new_seek (rate, GST_FORMAT_BYTES, self->segment_flags,
240 GST_SEEK_TYPE_SET, 0, GST_SEEK_TYPE_NONE, 0));
243 /* Apply the seek to our segment */
244 gst_segment_set_seek (self->segment, rate, format, self->segment_flags,
245 start_type, start, stop_type, stop, &update);
247 GST_DEBUG_OBJECT (self, "segment configured from %" GST_TIME_FORMAT
248 " to %" GST_TIME_FORMAT ", position %" GST_TIME_FORMAT,
249 GST_TIME_ARGS (self->segment->start),
250 GST_TIME_ARGS (self->segment->stop),
251 GST_TIME_ARGS (self->segment->last_stop));
253 self->next_offset = 0;
255 self->need_segment = TRUE;
257 GST_WARNING_OBJECT (self, "seek to 0 bytes failed");
260 gst_event_unref (event);
264 ret = gst_pad_event_default (pad, event);
269 gst_object_unref (self);
275 gst_sub_parse_set_property (GObject * object, guint prop_id,
276 const GValue * value, GParamSpec * pspec)
278 GstSubParse *subparse = GST_SUBPARSE (object);
280 GST_OBJECT_LOCK (subparse);
283 g_free (subparse->encoding);
284 subparse->encoding = g_value_dup_string (value);
285 GST_LOG_OBJECT (object, "subtitle encoding set to %s",
286 GST_STR_NULL (subparse->encoding));
289 G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
292 GST_OBJECT_UNLOCK (subparse);
296 gst_sub_parse_get_property (GObject * object, guint prop_id,
297 GValue * value, GParamSpec * pspec)
299 GstSubParse *subparse = GST_SUBPARSE (object);
301 GST_OBJECT_LOCK (subparse);
304 g_value_set_string (value, subparse->encoding);
307 G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
310 GST_OBJECT_UNLOCK (subparse);
314 convert_encoding (GstSubParse * self, const gchar * str, gsize len)
316 const gchar *encoding;
320 if (self->valid_utf8) {
321 if (g_utf8_validate (str, len, NULL)) {
322 GST_LOG_OBJECT (self, "valid UTF-8, no conversion needed");
323 return g_strndup (str, len);
325 GST_INFO_OBJECT (self, "invalid UTF-8!");
326 self->valid_utf8 = FALSE;
329 encoding = self->encoding;
330 if (encoding == NULL || *encoding == '\0') {
331 encoding = g_getenv ("GST_SUBTITLE_ENCODING");
333 if (encoding == NULL || *encoding == '\0') {
334 /* if local encoding is UTF-8 and no encoding specified
335 * via the environment variable, assume ISO-8859-15 */
336 if (g_get_charset (&encoding)) {
337 encoding = "ISO-8859-15";
341 ret = g_convert_with_fallback (str, len, "UTF-8", encoding, "*", NULL,
345 GST_WARNING_OBJECT (self, "could not convert string from '%s' to UTF-8: %s",
346 encoding, err->message);
349 /* invalid input encoding, fall back to ISO-8859-15 (always succeeds) */
350 ret = g_convert_with_fallback (str, len, "UTF-8", "ISO-8859-15", "*",
354 GST_LOG_OBJECT (self, "successfully converted %d characters from %s to UTF-8"
355 "%s", len, encoding, (err) ? " , using ISO-8859-15 as fallback" : "");
361 get_next_line (GstSubParse * self)
364 const char *line_end;
366 gboolean have_r = FALSE;
368 line_end = strchr (self->textbuf->str, '\n');
371 /* end-of-line not found; return for more data */
375 /* get rid of '\r' */
376 if (line_end != self->textbuf->str && *(line_end - 1) == '\r') {
381 line_len = line_end - self->textbuf->str;
382 line = convert_encoding (self, self->textbuf->str, line_len);
383 self->textbuf = g_string_erase (self->textbuf, 0,
384 line_len + (have_r ? 2 : 1));
389 parse_mdvdsub (ParserState * state, const gchar * line)
391 const gchar *line_split;
393 guint start_frame, end_frame;
394 gint64 clip_start = 0, clip_stop = 0;
395 gboolean in_seg = FALSE;
397 /* FIXME: hardcoded for now, but detecting the correct value is
398 * not going to be easy, I suspect... */
399 const double frames_per_sec = 24000 / 1001.;
403 /* style variables */
408 if (sscanf (line, "{%u}{%u}", &start_frame, &end_frame) != 2) {
409 g_warning ("Parse of the following line, assumed to be in microdvd .sub"
410 " format, failed:\n%s", line);
414 state->start_time = (start_frame - 1000) / frames_per_sec * GST_SECOND;
415 state->duration = (end_frame - start_frame) / frames_per_sec * GST_SECOND;
417 /* Check our segment start/stop */
418 in_seg = gst_segment_clip (state->segment, GST_FORMAT_TIME,
419 state->start_time, state->start_time + state->duration, &clip_start,
422 /* No need to parse that text if it's out of segment */
424 state->start_time = clip_start;
425 state->duration = clip_stop - clip_start;
430 /* skip the {%u}{%u} part */
431 line = strchr (line, '}') + 1;
432 line = strchr (line, '}') + 1;
434 markup = g_string_new (NULL);
439 /* parse style markup */
440 if (strncmp (line, "{y:i}", 5) == 0) {
442 line = strchr (line, '}') + 1;
444 if (strncmp (line, "{y:b}", 5) == 0) {
446 line = strchr (line, '}') + 1;
448 if (sscanf (line, "{s:%u}", &fontsize) == 1) {
449 line = strchr (line, '}') + 1;
451 if ((line_split = strchr (line, '|')))
452 line_chunk = g_markup_escape_text (line, line_split - line);
454 line_chunk = g_markup_escape_text (line, strlen (line));
455 markup = g_string_append (markup, "<span");
457 g_string_append (markup, " style=\"italic\"");
459 g_string_append (markup, " weight=\"bold\"");
461 g_string_append_printf (markup, " size=\"%u\"", fontsize * 1000);
462 g_string_append_printf (markup, ">%s</span>", line_chunk);
465 g_string_append (markup, "\n");
466 line = line_split + 1;
472 g_string_free (markup, FALSE);
473 GST_DEBUG ("parse_mdvdsub returning (%f+%f): %s",
474 state->start_time / (double) GST_SECOND,
475 state->duration / (double) GST_SECOND, ret);
480 strip_trailing_newlines (gchar * txt)
486 while (len > 1 && txt[len - 1] == '\n') {
493 /* we want to escape text in general, but retain basic markup like
494 * <i></i>, <u></u>, and <b></b>. The easiest and safest way is to
495 * just unescape a white list of allowed markups again after
496 * escaping everything (the text between these simple markers isn't
497 * necessarily escaped, so it seems best to do it like this) */
499 subrip_unescape_formatting (gchar * txt)
503 for (pos = txt; pos != NULL && *pos != '\0'; ++pos) {
504 if (g_ascii_strncasecmp (pos, "<u>", 9) == 0 ||
505 g_ascii_strncasecmp (pos, "<i>", 9) == 0 ||
506 g_ascii_strncasecmp (pos, "<b>", 9) == 0) {
508 pos[1] = g_ascii_tolower (pos[4]);
510 /* move NUL terminator as well */
511 g_memmove (pos + 3, pos + 9, strlen (pos + 9) + 1);
516 for (pos = txt; pos != NULL && *pos != '\0'; ++pos) {
517 if (g_ascii_strncasecmp (pos, "</u>", 10) == 0 ||
518 g_ascii_strncasecmp (pos, "</i>", 10) == 0 ||
519 g_ascii_strncasecmp (pos, "</b>", 10) == 0) {
522 pos[2] = g_ascii_tolower (pos[5]);
524 /* move NUL terminator as well */
525 g_memmove (pos + 4, pos + 10, strlen (pos + 10) + 1);
532 parse_subrip (ParserState * state, const gchar * line)
534 guint h1, m1, s1, ms1;
535 guint h2, m2, s2, ms2;
539 switch (state->state) {
541 /* looking for a single integer */
542 if (sscanf (line, "%u", &subnum) == 1)
546 /* looking for start_time --> end_time */
547 if (sscanf (line, "%u:%u:%u,%u --> %u:%u:%u,%u",
548 &h1, &m1, &s1, &ms1, &h2, &m2, &s2, &ms2) == 8) {
551 (((guint64) h1) * 3600 + m1 * 60 + s1) * GST_SECOND +
554 (((guint64) h2) * 3600 + m2 * 60 + s2) * GST_SECOND +
555 ms2 * GST_MSECOND - state->start_time;
557 GST_DEBUG ("error parsing subrip time line");
562 { /* No need to parse that text if it's out of segment */
563 gint64 clip_start = 0, clip_stop = 0;
564 gboolean in_seg = FALSE;
566 /* Check our segment start/stop */
567 in_seg = gst_segment_clip (state->segment, GST_FORMAT_TIME,
568 state->start_time, state->start_time + state->duration,
569 &clip_start, &clip_stop);
572 state->start_time = clip_start;
573 state->duration = clip_stop - clip_start;
579 /* looking for subtitle text; empty line ends this
582 g_string_append_c (state->buf, '\n');
583 g_string_append (state->buf, line);
584 if (strlen (line) == 0) {
585 ret = g_markup_escape_text (state->buf->str, state->buf->len);
586 g_string_truncate (state->buf, 0);
588 subrip_unescape_formatting (ret);
589 strip_trailing_newlines (ret);
594 g_return_val_if_reached (NULL);
599 parse_mpsub (ParserState * state, const gchar * line)
604 switch (state->state) {
606 /* looking for two floats (offset, duration) */
607 if (sscanf (line, "%f %f", &t1, &t2) == 2) {
609 state->start_time += state->duration + GST_SECOND * t1;
610 state->duration = GST_SECOND * t2;
614 { /* No need to parse that text if it's out of segment */
615 gint64 clip_start = 0, clip_stop = 0;
616 gboolean in_seg = FALSE;
618 /* Check our segment start/stop */
619 in_seg = gst_segment_clip (state->segment, GST_FORMAT_TIME,
620 state->start_time, state->start_time + state->duration,
621 &clip_start, &clip_stop);
624 state->start_time = clip_start;
625 state->duration = clip_stop - clip_start;
631 /* looking for subtitle text; empty line ends this
634 g_string_append_c (state->buf, '\n');
635 g_string_append (state->buf, line);
636 if (strlen (line) == 0) {
637 ret = g_strdup (state->buf->str);
638 g_string_truncate (state->buf, 0);
644 g_assert_not_reached ();
650 parser_state_init (ParserState * state)
652 GST_DEBUG ("initialising parser");
655 g_string_truncate (state->buf, 0);
657 state->buf = g_string_new (NULL);
660 state->start_time = 0;
663 state->segment = NULL;
667 parser_state_dispose (ParserState * state)
670 g_string_free (state->buf, TRUE);
673 if (state->user_data) {
674 sami_context_reset (state);
679 * FIXME: maybe we should pass along a second argument, the preceding
680 * text buffer, because that is how this originally worked, even though
681 * I don't really see the use of that.
684 static GstSubParseFormat
685 gst_sub_parse_data_format_autodetect (gchar * match_str)
687 static gboolean need_init_regexps = TRUE;
688 static regex_t mdvd_rx;
689 static regex_t subrip_rx;
691 /* initialize the regexps used the first time around */
692 if (need_init_regexps) {
696 need_init_regexps = FALSE;
697 if ((err = regcomp (&mdvd_rx, "^\\{[0-9]+\\}\\{[0-9]+\\}",
698 REG_EXTENDED | REG_NEWLINE | REG_NOSUB) != 0) ||
699 (err = regcomp (&subrip_rx, "^[1-9]([0-9]){0,3}(\x0d)?\x0a"
700 "[0-9][0-9]:[0-9][0-9]:[0-9][0-9],[0-9]{3}"
701 " --> [0-9][0-9]:[0-9][0-9]:[0-9][0-9],[0-9]{3}",
702 REG_EXTENDED | REG_NEWLINE | REG_NOSUB)) != 0) {
703 regerror (err, &subrip_rx, errstr, 127);
704 GST_WARNING ("Compilation of subrip regex failed: %s", errstr);
708 if (regexec (&mdvd_rx, match_str, 0, NULL, 0) == 0) {
709 GST_LOG ("MicroDVD (frame based) format detected");
710 return GST_SUB_PARSE_FORMAT_MDVDSUB;
712 if (regexec (&subrip_rx, match_str, 0, NULL, 0) == 0) {
713 GST_LOG ("SubRip (time based) format detected");
714 return GST_SUB_PARSE_FORMAT_SUBRIP;
716 if (!strncmp (match_str, "FORMAT=TIME", 11)) {
717 GST_LOG ("MPSub (time based) format detected");
718 return GST_SUB_PARSE_FORMAT_MPSUB;
720 if (strstr (match_str, "<SAMI>") != NULL ||
721 strstr (match_str, "<sami>") != NULL) {
722 GST_LOG ("SAMI (time based) format detected");
723 return GST_SUB_PARSE_FORMAT_SAMI;
726 GST_DEBUG ("no subtitle format detected");
727 return GST_SUB_PARSE_FORMAT_UNKNOWN;
731 gst_sub_parse_format_autodetect (GstSubParse * self)
734 GstSubParseFormat format;
736 if (strlen (self->textbuf->str) < 35) {
737 GST_DEBUG ("File too small to be a subtitles file");
741 data = g_strndup (self->textbuf->str, 35);
742 format = gst_sub_parse_data_format_autodetect (data);
745 self->parser_type = format;
746 parser_state_init (&self->state);
749 case GST_SUB_PARSE_FORMAT_MDVDSUB:
750 self->parse_line = parse_mdvdsub;
751 return gst_caps_new_simple ("text/x-pango-markup", NULL);
752 case GST_SUB_PARSE_FORMAT_SUBRIP:
753 self->parse_line = parse_subrip;
754 return gst_caps_new_simple ("text/x-pango-markup", NULL);
755 case GST_SUB_PARSE_FORMAT_MPSUB:
756 self->parse_line = parse_mpsub;
757 return gst_caps_new_simple ("text/plain", NULL);
758 case GST_SUB_PARSE_FORMAT_SAMI:
759 self->parse_line = parse_sami;
760 sami_context_init (&self->state);
761 return gst_caps_new_simple ("text/x-pango-markup", NULL);
762 case GST_SUB_PARSE_FORMAT_UNKNOWN:
764 GST_DEBUG ("no subtitle format detected");
765 GST_ELEMENT_ERROR (self, STREAM, WRONG_TYPE,
766 ("The input is not a valid/supported subtitle file"), (NULL));
772 feed_textbuf (GstSubParse * self, GstBuffer * buf)
774 if (GST_BUFFER_OFFSET (buf) != self->offset) {
775 /* flush the parser state */
776 parser_state_init (&self->state);
777 g_string_truncate (self->textbuf, 0);
778 sami_context_reset (&self->state);
781 self->textbuf = g_string_append_len (self->textbuf,
782 (gchar *) GST_BUFFER_DATA (buf), GST_BUFFER_SIZE (buf));
783 self->offset = GST_BUFFER_OFFSET (buf) + GST_BUFFER_SIZE (buf);
784 self->next_offset = self->offset;
786 gst_buffer_unref (buf);
790 handle_buffer (GstSubParse * self, GstBuffer * buf)
792 GstFlowReturn ret = GST_FLOW_OK;
793 GstCaps *caps = NULL;
794 gchar *line, *subtitle;
796 feed_textbuf (self, buf);
798 /* make sure we know the format */
799 if (G_UNLIKELY (self->parser_type == GST_SUB_PARSE_FORMAT_UNKNOWN)) {
800 if (!(caps = gst_sub_parse_format_autodetect (self))) {
801 return GST_FLOW_UNEXPECTED;
803 if (!gst_pad_set_caps (self->srcpad, caps)) {
804 gst_caps_unref (caps);
805 return GST_FLOW_UNEXPECTED;
807 gst_caps_unref (caps);
810 while ((line = get_next_line (self)) && !self->flushing) {
811 /* Set segment on our parser state machine */
812 self->state.segment = self->segment;
813 /* Now parse the line, out of segment lines will just return NULL */
814 GST_DEBUG ("Parsing line '%s'", line);
815 subtitle = self->parse_line (&self->state, line);
819 guint subtitle_len = strlen (subtitle);
821 ret = gst_pad_alloc_buffer_and_set_caps (self->srcpad,
822 GST_BUFFER_OFFSET_NONE, subtitle_len,
823 GST_PAD_CAPS (self->srcpad), &buf);
825 if (ret == GST_FLOW_OK) {
826 memcpy (GST_BUFFER_DATA (buf), subtitle, subtitle_len);
827 GST_BUFFER_TIMESTAMP (buf) = self->state.start_time;
828 GST_BUFFER_DURATION (buf) = self->state.duration;
830 gst_segment_set_last_stop (self->segment, GST_FORMAT_TIME,
831 self->state.start_time);
833 GST_DEBUG ("Sending text '%s', %" GST_TIME_FORMAT " + %"
834 GST_TIME_FORMAT, subtitle, GST_TIME_ARGS (self->state.start_time),
835 GST_TIME_ARGS (self->state.duration));
837 ret = gst_pad_push (self->srcpad, buf);
843 if (GST_FLOW_IS_FATAL (ret))
852 gst_sub_parse_chain (GstPad * sinkpad, GstBuffer * buf)
857 GST_DEBUG ("gst_sub_parse_chain");
858 self = GST_SUBPARSE (gst_pad_get_parent (sinkpad));
860 /* Push newsegment if needed */
861 if (self->need_segment) {
862 gst_pad_push_event (self->srcpad, gst_event_new_new_segment (FALSE,
863 self->segment->rate, self->segment->format,
864 self->segment->last_stop, self->segment->stop,
865 self->segment->time));
866 self->need_segment = FALSE;
869 ret = handle_buffer (self, buf);
871 gst_object_unref (self);
877 gst_sub_parse_sink_event (GstPad * pad, GstEvent * event)
879 GstSubParse *self = GST_SUBPARSE (gst_pad_get_parent (pad));
880 gboolean ret = FALSE;
882 GST_DEBUG ("Handling %s event", GST_EVENT_TYPE_NAME (event));
884 switch (GST_EVENT_TYPE (event)) {
886 /* Make sure the last subrip chunk is pushed out even
887 * if the file does not have an empty line at the end */
888 if (self->parser_type == GST_SUB_PARSE_FORMAT_SUBRIP) {
889 GstBuffer *buf = gst_buffer_new_and_alloc (1 + 1);
891 GST_DEBUG ("EOS. Pushing remaining text (if any)");
892 GST_BUFFER_DATA (buf)[0] = '\n';
893 GST_BUFFER_DATA (buf)[1] = '\0'; /* play it safe */
894 GST_BUFFER_SIZE (buf) = 1;
895 GST_BUFFER_OFFSET (buf) = self->offset;
896 gst_sub_parse_chain (pad, buf);
898 ret = gst_pad_event_default (pad, event);
901 case GST_EVENT_NEWSEGMENT:
905 gint64 start, stop, time;
908 GST_DEBUG_OBJECT (self, "received new segment");
910 gst_event_parse_new_segment (event, &update, &rate, &format, &start,
913 /* now copy over the values */
914 gst_segment_set_newsegment (self->segment, update, rate, format,
918 gst_event_unref (event);
921 case GST_EVENT_FLUSH_START:
923 self->flushing = TRUE;
925 ret = gst_pad_event_default (pad, event);
928 case GST_EVENT_FLUSH_STOP:
930 self->flushing = FALSE;
932 ret = gst_pad_event_default (pad, event);
936 ret = gst_pad_event_default (pad, event);
940 gst_object_unref (self);
946 static GstStateChangeReturn
947 gst_sub_parse_change_state (GstElement * element, GstStateChange transition)
949 GstStateChangeReturn ret = GST_STATE_CHANGE_SUCCESS;
950 GstSubParse *self = GST_SUBPARSE (element);
952 switch (transition) {
953 case GST_STATE_CHANGE_READY_TO_PAUSED:
954 /* format detection will init the parser state */
956 self->next_offset = 0;
957 self->parser_type = GST_SUB_PARSE_FORMAT_UNKNOWN;
958 self->valid_utf8 = TRUE;
959 g_string_truncate (self->textbuf, 0);
965 ret = parent_class->change_state (element, transition);
966 if (ret == GST_STATE_CHANGE_FAILURE)
969 switch (transition) {
970 case GST_STATE_CHANGE_PAUSED_TO_READY:
971 parser_state_dispose (&self->state);
972 self->parser_type = GST_SUB_PARSE_FORMAT_UNKNOWN;
985 static GstStaticCaps smi_caps = GST_STATIC_CAPS ("application/x-subtitle-sami");
986 static GstStaticCaps sub_caps = GST_STATIC_CAPS ("application/x-subtitle");
988 #define SUB_CAPS (gst_static_caps_get (&sub_caps))
989 #define SAMI_CAPS (gst_static_caps_get (&smi_caps))
992 gst_subparse_type_find (GstTypeFind * tf, gpointer private)
994 GstSubParseFormat format;
999 if (!(data = gst_type_find_peek (tf, 0, 36)))
1002 /* make sure string passed to _autodetect() is NUL-terminated */
1003 str = g_strndup ((gchar *) data, 35);
1004 format = gst_sub_parse_data_format_autodetect (str);
1008 case GST_SUB_PARSE_FORMAT_MDVDSUB:
1009 GST_DEBUG ("MicroDVD format detected");
1012 case GST_SUB_PARSE_FORMAT_SUBRIP:
1013 GST_DEBUG ("SubRip format detected");
1016 case GST_SUB_PARSE_FORMAT_MPSUB:
1017 GST_DEBUG ("MPSub format detected");
1020 case GST_SUB_PARSE_FORMAT_SAMI:
1021 GST_DEBUG ("SAMI (time-based) format detected");
1025 case GST_SUB_PARSE_FORMAT_UNKNOWN:
1026 GST_DEBUG ("no subtitle format detected");
1030 /* if we're here, it's ok */
1031 gst_type_find_suggest (tf, GST_TYPE_FIND_MAXIMUM, caps);
1035 plugin_init (GstPlugin * plugin)
1037 static gchar *sub_exts[] = { "srt", "sub", "mpsub", "mdvd", "smi", NULL };
1039 GST_DEBUG_CATEGORY_INIT (sub_parse_debug, "subparse", 0, ".sub parser");
1041 if (!gst_type_find_register (plugin, "subparse_typefind", GST_RANK_MARGINAL,
1042 gst_subparse_type_find, sub_exts, SUB_CAPS, NULL, NULL))
1045 if (!gst_element_register (plugin, "subparse",
1046 GST_RANK_PRIMARY, GST_TYPE_SUBPARSE) ||
1047 !gst_element_register (plugin, "ssaparse",
1048 GST_RANK_PRIMARY, GST_TYPE_SSA_PARSE)) {
1055 GST_PLUGIN_DEFINE (GST_VERSION_MAJOR,
1059 plugin_init, VERSION, "LGPL", GST_PACKAGE_NAME, GST_PACKAGE_ORIGIN)