#include "config.h"
#endif
-#include <string.h>
+#include <stdio.h>
#include <stdlib.h>
+#include <string.h>
#include <sys/types.h>
-#include <regex.h>
+#include <glib.h>
#include "gstsubparse.h"
#include "gstssaparse.h"
#include "samiparse.h"
#include "tmplayerparse.h"
#include "mpl2parse.h"
+#include "qttextparse.h"
GST_DEBUG_CATEGORY (sub_parse_debug);
enum
{
PROP_0,
- PROP_ENCODING
+ PROP_ENCODING,
+ PROP_VIDEOFPS
};
static void
GValue * value, GParamSpec * pspec);
-static const GstElementDetails sub_parse_details =
-GST_ELEMENT_DETAILS ("Subtitle parser",
- "Codec/Parser/Subtitle",
- "Parses subtitle (.sub) files into text streams",
- "Gustavo J. A. M. Carneiro <gjc@inescporto.pt>\n"
- "Ronald S. Bultje <rbultje@ronald.bitfreak.net>");
-
#ifndef GST_DISABLE_XML
static GstStaticPadTemplate sink_templ = GST_STATIC_PAD_TEMPLATE ("sink",
GST_PAD_SINK,
GST_PAD_ALWAYS,
GST_STATIC_CAPS ("application/x-subtitle; application/x-subtitle-sami; "
- "application/x-subtitle-tmplayer; application/x-subtitle-mpl2")
+ "application/x-subtitle-tmplayer; application/x-subtitle-mpl2; "
+ "application/x-subtitle-dks; application/x-subtitle-qttext")
);
#else
static GstStaticPadTemplate sink_templ = GST_STATIC_PAD_TEMPLATE ("sink",
GST_PAD_SINK,
GST_PAD_ALWAYS,
- GST_STATIC_CAPS ("application/x-subtitle")
+ GST_STATIC_CAPS ("application/x-subtitle; application/x-subtitle-dks; "
+ "application/x-subtitle-tmplayer; application/x-subtitle-mpl2; "
+ "application/x-subtitle-qttext")
);
#endif
GST_STATIC_CAPS ("text/plain; text/x-pango-markup")
);
-static void gst_sub_parse_base_init (GstSubParseClass * klass);
-static void gst_sub_parse_class_init (GstSubParseClass * klass);
-static void gst_sub_parse_init (GstSubParse * subparse);
static gboolean gst_sub_parse_src_event (GstPad * pad, GstEvent * event);
+static gboolean gst_sub_parse_src_query (GstPad * pad, GstQuery * query);
static gboolean gst_sub_parse_sink_event (GstPad * pad, GstEvent * event);
static GstStateChangeReturn gst_sub_parse_change_state (GstElement * element,
static GstFlowReturn gst_sub_parse_chain (GstPad * sinkpad, GstBuffer * buf);
-static GstElementClass *parent_class = NULL;
-
-GType
-gst_sub_parse_get_type (void)
-{
- static GType sub_parse_type = 0;
-
- if (!sub_parse_type) {
- static const GTypeInfo sub_parse_info = {
- sizeof (GstSubParseClass),
- (GBaseInitFunc) gst_sub_parse_base_init,
- NULL,
- (GClassInitFunc) gst_sub_parse_class_init,
- NULL,
- NULL,
- sizeof (GstSubParse),
- 0,
- (GInstanceInitFunc) gst_sub_parse_init,
- };
-
- sub_parse_type = g_type_register_static (GST_TYPE_ELEMENT,
- "GstSubParse", &sub_parse_info, 0);
- }
-
- return sub_parse_type;
-}
-
-static void
-gst_sub_parse_base_init (GstSubParseClass * klass)
-{
- GstElementClass *element_class = GST_ELEMENT_CLASS (klass);
-
- gst_element_class_add_pad_template (element_class,
- gst_static_pad_template_get (&sink_templ));
- gst_element_class_add_pad_template (element_class,
- gst_static_pad_template_get (&src_templ));
- gst_element_class_set_details (element_class, &sub_parse_details);
-}
+#define gst_sub_parse_parent_class parent_class
+G_DEFINE_TYPE (GstSubParse, gst_sub_parse, GST_TYPE_ELEMENT);
static void
gst_sub_parse_dispose (GObject * object)
GST_DEBUG_OBJECT (subparse, "cleaning up subtitle parser");
+ switch (subparse->parser_type) {
+ case GST_SUB_PARSE_FORMAT_QTTEXT:
+ qttext_context_deinit (&subparse->state);
+ break;
+#ifndef GST_DISABLE_XML
+ case GST_SUB_PARSE_FORMAT_SAMI:
+ sami_context_deinit (&subparse->state);
+ break;
+#endif
+ default:
+ break;
+ }
+
if (subparse->encoding) {
g_free (subparse->encoding);
subparse->encoding = NULL;
}
+
+ if (subparse->detected_encoding) {
+ g_free (subparse->detected_encoding);
+ subparse->detected_encoding = NULL;
+ }
+
+ if (subparse->adapter) {
+ g_object_unref (subparse->adapter);
+ subparse->adapter = NULL;
+ }
+
if (subparse->textbuf) {
g_string_free (subparse->textbuf, TRUE);
subparse->textbuf = NULL;
}
- sami_context_deinit (&subparse->state);
GST_CALL_PARENT (G_OBJECT_CLASS, dispose, (object));
}
GObjectClass *object_class = G_OBJECT_CLASS (klass);
GstElementClass *element_class = GST_ELEMENT_CLASS (klass);
- parent_class = g_type_class_peek_parent (klass);
-
object_class->dispose = gst_sub_parse_dispose;
object_class->set_property = gst_sub_parse_set_property;
object_class->get_property = gst_sub_parse_get_property;
+ gst_element_class_add_pad_template (element_class,
+ gst_static_pad_template_get (&sink_templ));
+ gst_element_class_add_pad_template (element_class,
+ gst_static_pad_template_get (&src_templ));
+ gst_element_class_set_details_simple (element_class,
+ "Subtitle parser", "Codec/Parser/Subtitle",
+ "Parses subtitle (.sub) files into text streams",
+ "Gustavo J. A. M. Carneiro <gjc@inescporto.pt>, "
+ "GStreamer maintainers <gstreamer-devel@lists.sourceforge.net>");
+
element_class->change_state = gst_sub_parse_change_state;
g_object_class_install_property (object_class, PROP_ENCODING,
g_param_spec_string ("subtitle-encoding", "subtitle charset encoding",
- "Encoding to assume if input subtitles are not in UTF-8 encoding. "
- "If not set, the GST_SUBTITLE_ENCODING environment variable will "
- "be checked for an encoding to use. If that is not set either, "
- "ISO-8859-15 will be assumed.", DEFAULT_ENCODING, G_PARAM_READWRITE));
+ "Encoding to assume if input subtitles are not in UTF-8 or any other "
+ "Unicode encoding. If not set, the GST_SUBTITLE_ENCODING environment "
+ "variable will be checked for an encoding to use. If that is not set "
+ "either, ISO-8859-15 will be assumed.", DEFAULT_ENCODING,
+ G_PARAM_READWRITE | G_PARAM_STATIC_STRINGS));
+
+ g_object_class_install_property (object_class, PROP_VIDEOFPS,
+ gst_param_spec_fraction ("video-fps", "Video framerate",
+ "Framerate of the video stream. This is needed by some subtitle "
+ "formats to synchronize subtitles and video properly. If not set "
+ "and the subtitle format requires it subtitles may be out of sync.",
+ 0, 1, G_MAXINT, 1, 24000, 1001,
+ G_PARAM_READWRITE | G_PARAM_STATIC_STRINGS));
}
static void
subparse->srcpad = gst_pad_new_from_static_template (&src_templ, "src");
gst_pad_set_event_function (subparse->srcpad,
GST_DEBUG_FUNCPTR (gst_sub_parse_src_event));
+ gst_pad_set_query_function (subparse->srcpad,
+ GST_DEBUG_FUNCPTR (gst_sub_parse_src_query));
gst_element_add_pad (GST_ELEMENT (subparse), subparse->srcpad);
subparse->textbuf = g_string_new (NULL);
gst_segment_init (&subparse->segment, GST_FORMAT_TIME);
subparse->need_segment = TRUE;
subparse->encoding = g_strdup (DEFAULT_ENCODING);
+ subparse->detected_encoding = NULL;
+ subparse->adapter = gst_adapter_new ();
+
+ subparse->fps_n = 24000;
+ subparse->fps_d = 1001;
}
/*
*/
static gboolean
+gst_sub_parse_src_query (GstPad * pad, GstQuery * query)
+{
+ GstSubParse *self = GST_SUBPARSE (gst_pad_get_parent (pad));
+ gboolean ret = FALSE;
+
+ GST_DEBUG ("Handling %s query", GST_QUERY_TYPE_NAME (query));
+
+ switch (GST_QUERY_TYPE (query)) {
+ case GST_QUERY_POSITION:{
+ GstFormat fmt;
+
+ gst_query_parse_position (query, &fmt, NULL);
+ if (fmt != GST_FORMAT_TIME) {
+ ret = gst_pad_peer_query (self->sinkpad, query);
+ } else {
+ ret = TRUE;
+ gst_query_set_position (query, GST_FORMAT_TIME, self->segment.position);
+ }
+ break;
+ }
+ case GST_QUERY_SEEKING:
+ {
+ GstFormat fmt;
+ gboolean seekable = FALSE;
+
+ ret = TRUE;
+
+ gst_query_parse_seeking (query, &fmt, NULL, NULL, NULL);
+ if (fmt == GST_FORMAT_TIME) {
+ GstQuery *peerquery = gst_query_new_seeking (GST_FORMAT_BYTES);
+
+ seekable = gst_pad_peer_query (self->sinkpad, peerquery);
+ if (seekable)
+ gst_query_parse_seeking (peerquery, NULL, &seekable, NULL, NULL);
+ gst_query_unref (peerquery);
+ }
+
+ gst_query_set_seeking (query, fmt, seekable, seekable ? 0 : -1, -1);
+ break;
+ }
+ default:
+ ret = gst_pad_peer_query (self->sinkpad, query);
+ break;
+ }
+
+ gst_object_unref (self);
+
+ return ret;
+}
+
+static gboolean
gst_sub_parse_src_event (GstPad * pad, GstEvent * event)
{
GstSubParse *self = GST_SUBPARSE (gst_pad_get_parent (pad));
case GST_EVENT_SEEK:
{
GstFormat format;
+ GstSeekFlags flags;
GstSeekType start_type, stop_type;
gint64 start, stop;
gdouble rate;
gboolean update;
- gst_event_parse_seek (event, &rate, &format, &self->segment_flags,
+ gst_event_parse_seek (event, &rate, &format, &flags,
&start_type, &start, &stop_type, &stop);
if (format != GST_FORMAT_TIME) {
/* Convert that seek to a seeking in bytes at position 0,
FIXME: could use an index */
ret = gst_pad_push_event (self->sinkpad,
- gst_event_new_seek (rate, GST_FORMAT_BYTES, self->segment_flags,
+ gst_event_new_seek (rate, GST_FORMAT_BYTES, flags,
GST_SEEK_TYPE_SET, 0, GST_SEEK_TYPE_NONE, 0));
if (ret) {
/* Apply the seek to our segment */
- gst_segment_set_seek (&self->segment, rate, format, self->segment_flags,
+ gst_segment_do_seek (&self->segment, rate, format, flags,
start_type, start, stop_type, stop, &update);
GST_DEBUG_OBJECT (self, "segment after seek: %" GST_SEGMENT_FORMAT,
GST_LOG_OBJECT (object, "subtitle encoding set to %s",
GST_STR_NULL (subparse->encoding));
break;
+ case PROP_VIDEOFPS:
+ {
+ subparse->fps_n = gst_value_get_fraction_numerator (value);
+ subparse->fps_d = gst_value_get_fraction_denominator (value);
+ GST_DEBUG_OBJECT (object, "video framerate set to %d/%d", subparse->fps_n,
+ subparse->fps_d);
+
+ if (!subparse->state.have_internal_fps) {
+ subparse->state.fps_n = subparse->fps_n;
+ subparse->state.fps_d = subparse->fps_d;
+ }
+ break;
+ }
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
case PROP_ENCODING:
g_value_set_string (value, subparse->encoding);
break;
+ case PROP_VIDEOFPS:
+ gst_value_set_fraction (value, subparse->fps_n, subparse->fps_d);
+ break;
default:
G_OBJECT_WARN_INVALID_PROPERTY_ID (object, prop_id, pspec);
break;
GST_OBJECT_UNLOCK (subparse);
}
+static const gchar *
+gst_sub_parse_get_format_description (GstSubParseFormat format)
+{
+ switch (format) {
+ case GST_SUB_PARSE_FORMAT_MDVDSUB:
+ return "MicroDVD";
+ case GST_SUB_PARSE_FORMAT_SUBRIP:
+ return "SubRip";
+ case GST_SUB_PARSE_FORMAT_MPSUB:
+ return "MPSub";
+ case GST_SUB_PARSE_FORMAT_SAMI:
+ return "SAMI";
+ case GST_SUB_PARSE_FORMAT_TMPLAYER:
+ return "TMPlayer";
+ case GST_SUB_PARSE_FORMAT_MPL2:
+ return "MPL2";
+ case GST_SUB_PARSE_FORMAT_SUBVIEWER:
+ return "SubViewer";
+ case GST_SUB_PARSE_FORMAT_DKS:
+ return "DKS";
+ case GST_SUB_PARSE_FORMAT_QTTEXT:
+ return "QTtext";
+ default:
+ case GST_SUB_PARSE_FORMAT_UNKNOWN:
+ break;
+ }
+ return NULL;
+}
+
+static gchar *
+gst_convert_to_utf8 (const gchar * str, gsize len, const gchar * encoding,
+ gsize * consumed, GError ** err)
+{
+ gchar *ret = NULL;
+
+ *consumed = 0;
+ /* The char cast is necessary in glib < 2.24 */
+ ret =
+ g_convert_with_fallback (str, len, "UTF-8", encoding, (char *) "*",
+ consumed, NULL, err);
+ if (ret == NULL)
+ return ret;
+
+ /* + 3 to skip UTF-8 BOM if it was added */
+ len = strlen (ret);
+ if (len >= 3 && (guint8) ret[0] == 0xEF && (guint8) ret[1] == 0xBB
+ && (guint8) ret[2] == 0xBF)
+ g_memmove (ret, ret + 3, len + 1 - 3);
+
+ return ret;
+}
+
+static gchar *
+detect_encoding (const gchar * str, gsize len)
+{
+ if (len >= 3 && (guint8) str[0] == 0xEF && (guint8) str[1] == 0xBB
+ && (guint8) str[2] == 0xBF)
+ return g_strdup ("UTF-8");
+
+ if (len >= 2 && (guint8) str[0] == 0xFE && (guint8) str[1] == 0xFF)
+ return g_strdup ("UTF-16BE");
+
+ if (len >= 2 && (guint8) str[0] == 0xFF && (guint8) str[1] == 0xFE)
+ return g_strdup ("UTF-16LE");
+
+ if (len >= 4 && (guint8) str[0] == 0x00 && (guint8) str[1] == 0x00
+ && (guint8) str[2] == 0xFE && (guint8) str[3] == 0xFF)
+ return g_strdup ("UTF-32BE");
+
+ if (len >= 4 && (guint8) str[0] == 0xFF && (guint8) str[1] == 0xFE
+ && (guint8) str[2] == 0x00 && (guint8) str[3] == 0x00)
+ return g_strdup ("UTF-32LE");
+
+ return NULL;
+}
+
static gchar *
-convert_encoding (GstSubParse * self, const gchar * str, gsize len)
+convert_encoding (GstSubParse * self, const gchar * str, gsize len,
+ gsize * consumed)
{
const gchar *encoding;
GError *err = NULL;
- gchar *ret;
+ gchar *ret = NULL;
+
+ *consumed = 0;
+
+ /* First try any detected encoding */
+ if (self->detected_encoding) {
+ ret =
+ gst_convert_to_utf8 (str, len, self->detected_encoding, consumed, &err);
+ if (!err)
+ return ret;
+
+ GST_WARNING_OBJECT (self, "could not convert string from '%s' to UTF-8: %s",
+ self->detected_encoding, err->message);
+ g_free (self->detected_encoding);
+ self->detected_encoding = NULL;
+ g_error_free (err);
+ }
+
+ /* Otherwise check if it's UTF8 */
if (self->valid_utf8) {
if (g_utf8_validate (str, len, NULL)) {
GST_LOG_OBJECT (self, "valid UTF-8, no conversion needed");
+ *consumed = len;
return g_strndup (str, len);
}
GST_INFO_OBJECT (self, "invalid UTF-8!");
self->valid_utf8 = FALSE;
}
+ /* Else try fallback */
encoding = self->encoding;
if (encoding == NULL || *encoding == '\0') {
encoding = g_getenv ("GST_SUBTITLE_ENCODING");
}
}
- ret = g_convert_with_fallback (str, len, "UTF-8", encoding, "*", NULL,
- NULL, &err);
+ ret = gst_convert_to_utf8 (str, len, encoding, consumed, &err);
if (err) {
GST_WARNING_OBJECT (self, "could not convert string from '%s' to UTF-8: %s",
g_error_free (err);
/* invalid input encoding, fall back to ISO-8859-15 (always succeeds) */
- ret = g_convert_with_fallback (str, len, "UTF-8", "ISO-8859-15", "*",
- NULL, NULL, NULL);
+ ret = gst_convert_to_utf8 (str, len, "ISO-8859-15", consumed, NULL);
}
GST_LOG_OBJECT (self,
}
line_len = line_end - self->textbuf->str;
- line = convert_encoding (self, self->textbuf->str, line_len);
+ line = g_strndup (self->textbuf->str, line_len);
self->textbuf = g_string_erase (self->textbuf, 0,
line_len + (have_r ? 2 : 1));
return line;
const gchar *line_split;
gchar *line_chunk;
guint start_frame, end_frame;
- gint64 clip_start = 0, clip_stop = 0;
+ guint64 clip_start = 0, clip_stop = 0;
gboolean in_seg = FALSE;
GString *markup;
gchar *ret;
gboolean italic;
gboolean bold;
guint fontsize;
+ gdouble fps = 0.0;
if (sscanf (line, "{%u}{%u}", &start_frame, &end_frame) != 2) {
g_warning ("Parse of the following line, assumed to be in microdvd .sub"
line = strchr (line, '}') + 1;
/* see if there's a first line with a framerate */
- if (state->fps == 0.0 && start_frame == 1 && end_frame == 1) {
+ if (start_frame == 1 && end_frame == 1) {
gchar *rest, *end = NULL;
rest = g_strdup (line);
g_strdelimit (rest, ",", '.');
- state->fps = g_ascii_strtod (rest, &end);
- if (end == rest)
- state->fps = 0.0;
- GST_INFO ("framerate from file: %f ('%s')", state->fps, rest);
+ fps = g_ascii_strtod (rest, &end);
+ if (end != rest) {
+ gst_util_double_to_fraction (fps, &state->fps_n, &state->fps_d);
+ GST_INFO ("framerate from file: %d/%d ('%s')", state->fps_n,
+ state->fps_d, rest);
+ }
g_free (rest);
return NULL;
}
- if (state->fps == 0.0) {
- /* FIXME: hardcoded for now, is there a better way/assumption? */
- state->fps = 24000.0 / 1001.0;
- GST_INFO ("no framerate specified, assuming %f", state->fps);
- }
-
- state->start_time = start_frame / state->fps * GST_SECOND;
- state->duration = (end_frame - start_frame) / state->fps * GST_SECOND;
+ state->start_time =
+ gst_util_uint64_scale (start_frame, GST_SECOND * state->fps_d,
+ state->fps_n);
+ state->duration =
+ gst_util_uint64_scale (end_frame - start_frame, GST_SECOND * state->fps_d,
+ state->fps_n);
/* Check our segment start/stop */
in_seg = gst_segment_clip (state->segment, GST_FORMAT_TIME,
if (sscanf (line, "{s:%u}", &fontsize) == 1) {
line = strchr (line, '}') + 1;
}
+ /* forward slashes at beginning/end signify italics too */
+ if (g_str_has_prefix (line, "/")) {
+ italic = TRUE;
+ ++line;
+ }
if ((line_split = strchr (line, '|')))
line_chunk = g_markup_escape_text (line, line_split - line);
else
line_chunk = g_markup_escape_text (line, strlen (line));
+
+ /* Remove italics markers at end of line/stanza (CHECKME: are end slashes
+ * always at the end of a line or can they span multiple lines?) */
+ if (g_str_has_suffix (line_chunk, "/")) {
+ line_chunk[strlen (line_chunk) - 1] = '\0';
+ }
+
markup = g_string_append (markup, "<span");
if (italic)
g_string_append (markup, " style=\"italic\"");
}
}
+static gboolean
+parse_subrip_time (const gchar * ts_string, GstClockTime * t)
+{
+ gchar s[128] = { '\0', };
+ gchar *end, *p;
+ guint hour, min, sec, msec, len;
+
+ while (*ts_string == ' ')
+ ++ts_string;
+
+ g_strlcpy (s, ts_string, sizeof (s));
+ if ((end = strstr (s, "-->")))
+ *end = '\0';
+ g_strchomp (s);
+
+ /* ms may be in these formats:
+ * hh:mm:ss,500 = 500ms
+ * hh:mm:ss, 5 = 5ms
+ * hh:mm:ss, 5 = 50ms
+ * hh:mm:ss, 50 = 50ms
+ * hh:mm:ss,5 = 500ms
+ * and the same with . instead of ,.
+ * sscanf() doesn't differentiate between ' 5' and '5' so munge
+ * the white spaces within the timestamp to '0' (I'm sure there's a
+ * way to make sscanf() do this for us, but how?)
+ */
+ g_strdelimit (s, " ", '0');
+ g_strdelimit (s, ".", ',');
+
+ /* make sure we have exactly three digits after he comma */
+ p = strchr (s, ',');
+ g_assert (p != NULL);
+ ++p;
+ len = strlen (p);
+ if (len > 3) {
+ p[3] = '\0';
+ } else
+ while (len < 3) {
+ g_strlcat (&p[len], "0", 2);
+ ++len;
+ }
+
+ GST_LOG ("parsing timestamp '%s'", s);
+ if (sscanf (s, "%u:%u:%u,%u", &hour, &min, &sec, &msec) != 4) {
+ GST_WARNING ("failed to parse subrip timestamp string '%s'", s);
+ return FALSE;
+ }
+
+ *t = ((hour * 3600) + (min * 60) + sec) * GST_SECOND + msec * GST_MSECOND;
+ return TRUE;
+}
+
static gchar *
parse_subrip (ParserState * state, const gchar * line)
{
- guint h1, m1, s1, ms1;
- guint h2, m2, s2, ms2;
int subnum;
gchar *ret;
state->state = 1;
return NULL;
case 1:
+ {
+ GstClockTime ts_start, ts_end;
+ gchar *end_time;
+
/* looking for start_time --> end_time */
- if (sscanf (line, "%u:%u:%u,%u --> %u:%u:%u,%u",
- &h1, &m1, &s1, &ms1, &h2, &m2, &s2, &ms2) == 8) {
+ if ((end_time = strstr (line, " --> ")) &&
+ parse_subrip_time (line, &ts_start) &&
+ parse_subrip_time (end_time + strlen (" --> "), &ts_end) &&
+ state->start_time <= ts_end) {
state->state = 2;
- state->start_time =
- (((guint64) h1) * 3600 + m1 * 60 + s1) * GST_SECOND +
- ms1 * GST_MSECOND;
- state->duration =
- (((guint64) h2) * 3600 + m2 * 60 + s2) * GST_SECOND +
- ms2 * GST_MSECOND - state->start_time;
+ state->start_time = ts_start;
+ state->duration = ts_end - ts_start;
} else {
- GST_DEBUG ("error parsing subrip time line");
+ GST_DEBUG ("error parsing subrip time line '%s'", line);
state->state = 0;
}
return NULL;
+ }
case 2:
{
/* No need to parse that text if it's out of segment */
- gint64 clip_start = 0, clip_stop = 0;
+ guint64 clip_start = 0, clip_stop = 0;
gboolean in_seg = FALSE;
/* Check our segment start/stop */
}
static void
-subviewer_unescape_newlines (gchar * read)
+unescape_newlines_br (gchar * read)
{
gchar *write = read;
case 1:
{
/* No need to parse that text if it's out of segment */
- gint64 clip_start = 0, clip_stop = 0;
+ guint64 clip_start = 0, clip_stop = 0;
gboolean in_seg = FALSE;
/* Check our segment start/stop */
g_string_append (state->buf, line);
if (strlen (line) == 0) {
ret = g_strdup (state->buf->str);
- subviewer_unescape_newlines (ret);
+ unescape_newlines_br (ret);
strip_trailing_newlines (ret);
g_string_truncate (state->buf, 0);
state->state = 0;
return NULL;
case 1:
{ /* No need to parse that text if it's out of segment */
- gint64 clip_start = 0, clip_stop = 0;
+ guint64 clip_start = 0, clip_stop = 0;
gboolean in_seg = FALSE;
/* Check our segment start/stop */
}
}
+static const gchar *
+dks_skip_timestamp (const gchar * line)
+{
+ while (*line && *line != ']')
+ line++;
+ if (*line == ']')
+ line++;
+ return line;
+}
+
+static gchar *
+parse_dks (ParserState * state, const gchar * line)
+{
+ guint h, m, s;
+
+ switch (state->state) {
+ case 0:
+ /* Looking for the start time and text */
+ if (sscanf (line, "[%u:%u:%u]", &h, &m, &s) == 3) {
+ const gchar *text;
+ state->start_time = (((guint64) h) * 3600 + m * 60 + s) * GST_SECOND;
+ text = dks_skip_timestamp (line);
+ if (*text) {
+ state->state = 1;
+ g_string_append (state->buf, text);
+ }
+ }
+ return NULL;
+ case 1:
+ {
+ guint64 clip_start = 0, clip_stop = 0;
+ gboolean in_seg;
+ gchar *ret;
+
+ /* Looking for the end time */
+ if (sscanf (line, "[%u:%u:%u]", &h, &m, &s) == 3) {
+ state->state = 0;
+ state->duration = (((guint64) h) * 3600 + m * 60 + s) * GST_SECOND -
+ state->start_time;
+ } else {
+ GST_WARNING ("Failed to parse subtitle end time");
+ return NULL;
+ }
+
+ /* Check if this subtitle is out of the current segment */
+ in_seg = gst_segment_clip (state->segment, GST_FORMAT_TIME,
+ state->start_time, state->start_time + state->duration,
+ &clip_start, &clip_stop);
+
+ if (!in_seg) {
+ return NULL;
+ }
+
+ state->start_time = clip_start;
+ state->duration = clip_stop - clip_start;
+
+ ret = g_strdup (state->buf->str);
+ g_string_truncate (state->buf, 0);
+ unescape_newlines_br (ret);
+ return ret;
+ }
+ default:
+ g_assert_not_reached ();
+ return NULL;
+ }
+}
+
static void
parser_state_init (ParserState * state)
{
state->start_time = 0;
state->duration = 0;
+ state->max_duration = 0; /* no limit */
state->state = 0;
state->segment = NULL;
}
static void
-parser_state_dispose (ParserState * state)
+parser_state_dispose (GstSubParse * self, ParserState * state)
{
if (state->buf) {
g_string_free (state->buf, TRUE);
state->buf = NULL;
}
if (state->user_data) {
- sami_context_reset (state);
+ switch (self->parser_type) {
+#ifndef GST_DISABLE_XML
+ case GST_SUB_PARSE_FORMAT_SAMI:
+ sami_context_reset (state);
+ break;
+#endif
+ default:
+ break;
+ }
+ }
+}
+
+/* regex type enum */
+typedef enum
+{
+ GST_SUB_PARSE_REGEX_UNKNOWN = 0,
+ GST_SUB_PARSE_REGEX_MDVDSUB = 1,
+ GST_SUB_PARSE_REGEX_SUBRIP = 2,
+ GST_SUB_PARSE_REGEX_DKS = 3,
+} GstSubParseRegex;
+
+static gpointer
+gst_sub_parse_data_format_autodetect_regex_once (GstSubParseRegex regtype)
+{
+ gpointer result = NULL;
+ GError *gerr = NULL;
+ switch (regtype) {
+ case GST_SUB_PARSE_REGEX_MDVDSUB:
+ result =
+ (gpointer) g_regex_new ("^\\{[0-9]+\\}\\{[0-9]+\\}",
+ G_REGEX_RAW | G_REGEX_OPTIMIZE, 0, &gerr);
+ if (result == NULL) {
+ g_warning ("Compilation of mdvd regex failed: %s", gerr->message);
+ g_error_free (gerr);
+ }
+ break;
+ case GST_SUB_PARSE_REGEX_SUBRIP:
+ result = (gpointer) g_regex_new ("^([ 0-9]){0,3}[0-9]\\s*(\x0d)?\x0a"
+ "[ 0-9][0-9]:[ 0-9][0-9]:[ 0-9][0-9][,.][ 0-9]{0,2}[0-9]"
+ " +--> +([ 0-9])?[0-9]:[ 0-9][0-9]:[ 0-9][0-9][,.][ 0-9]{0,2}[0-9]",
+ G_REGEX_RAW | G_REGEX_OPTIMIZE, 0, &gerr);
+ if (result == NULL) {
+ g_warning ("Compilation of subrip regex failed: %s", gerr->message);
+ g_error_free (gerr);
+ }
+ break;
+ case GST_SUB_PARSE_REGEX_DKS:
+ result = (gpointer) g_regex_new ("^\\[[0-9]+:[0-9]+:[0-9]+\\].*",
+ G_REGEX_RAW | G_REGEX_OPTIMIZE, 0, &gerr);
+ if (result == NULL) {
+ g_warning ("Compilation of dks regex failed: %s", gerr->message);
+ g_error_free (gerr);
+ }
+ break;
+ default:
+ GST_WARNING ("Trying to allocate regex of unknown type %u", regtype);
}
+ return result;
}
/*
static GstSubParseFormat
gst_sub_parse_data_format_autodetect (gchar * match_str)
{
- static gboolean need_init_regexps = TRUE;
- static regex_t mdvd_rx;
- static regex_t subrip_rx;
guint n1, n2, n3;
- /* initialize the regexps used the first time around */
- if (need_init_regexps) {
- int err;
- char errstr[128];
-
- need_init_regexps = FALSE;
- if ((err = regcomp (&mdvd_rx, "^\\{[0-9]+\\}\\{[0-9]+\\}",
- REG_EXTENDED | REG_NEWLINE | REG_NOSUB) != 0) ||
- (err = regcomp (&subrip_rx, "^[0-9]([0-9]){0,3}(\x0d)?\x0a"
- "[0-9][0-9]:[0-9][0-9]:[0-9][0-9],[0-9]{3}"
- " --> [0-9][0-9]:[0-9][0-9]:[0-9][0-9],[0-9]{3}",
- REG_EXTENDED | REG_NEWLINE | REG_NOSUB)) != 0) {
- regerror (err, &subrip_rx, errstr, 127);
- GST_WARNING ("Compilation of subrip regex failed: %s", errstr);
- }
- }
-
- if (regexec (&mdvd_rx, match_str, 0, NULL, 0) == 0) {
+ static GOnce mdvd_rx_once = G_ONCE_INIT;
+ static GOnce subrip_rx_once = G_ONCE_INIT;
+ static GOnce dks_rx_once = G_ONCE_INIT;
+
+ GRegex *mdvd_grx;
+ GRegex *subrip_grx;
+ GRegex *dks_grx;
+
+ g_once (&mdvd_rx_once,
+ (GThreadFunc) gst_sub_parse_data_format_autodetect_regex_once,
+ (gpointer) GST_SUB_PARSE_REGEX_MDVDSUB);
+ g_once (&subrip_rx_once,
+ (GThreadFunc) gst_sub_parse_data_format_autodetect_regex_once,
+ (gpointer) GST_SUB_PARSE_REGEX_SUBRIP);
+ g_once (&dks_rx_once,
+ (GThreadFunc) gst_sub_parse_data_format_autodetect_regex_once,
+ (gpointer) GST_SUB_PARSE_REGEX_DKS);
+
+ mdvd_grx = (GRegex *) mdvd_rx_once.retval;
+ subrip_grx = (GRegex *) subrip_rx_once.retval;
+ dks_grx = (GRegex *) dks_rx_once.retval;
+
+ if (g_regex_match (mdvd_grx, match_str, 0, NULL) == TRUE) {
GST_LOG ("MicroDVD (frame based) format detected");
return GST_SUB_PARSE_FORMAT_MDVDSUB;
}
- if (regexec (&subrip_rx, match_str, 0, NULL, 0) == 0) {
+ if (g_regex_match (subrip_grx, match_str, 0, NULL) == TRUE) {
GST_LOG ("SubRip (time based) format detected");
return GST_SUB_PARSE_FORMAT_SUBRIP;
}
+ if (g_regex_match (dks_grx, match_str, 0, NULL) == TRUE) {
+ GST_LOG ("DKS (time based) format detected");
+ return GST_SUB_PARSE_FORMAT_DKS;
+ }
if (!strncmp (match_str, "FORMAT=TIME", 11)) {
GST_LOG ("MPSub (time based) format detected");
return GST_SUB_PARSE_FORMAT_MPSUB;
}
+#ifndef GST_DISABLE_XML
if (strstr (match_str, "<SAMI>") != NULL ||
strstr (match_str, "<sami>") != NULL) {
GST_LOG ("SAMI (time based) format detected");
return GST_SUB_PARSE_FORMAT_SAMI;
}
+#endif
/* we're boldly assuming the first subtitle appears within the first hour */
if (sscanf (match_str, "0:%02u:%02u:", &n1, &n2) == 2 ||
sscanf (match_str, "0:%02u:%02u=", &n1, &n2) == 2 ||
GST_LOG ("SubViewer (time based) format detected");
return GST_SUB_PARSE_FORMAT_SUBVIEWER;
}
+ if (strstr (match_str, "{QTtext}") != NULL) {
+ GST_LOG ("QTtext (time based) format detected");
+ return GST_SUB_PARSE_FORMAT_QTTEXT;
+ }
GST_DEBUG ("no subtitle format detected");
return GST_SUB_PARSE_FORMAT_UNKNOWN;
gchar *data;
GstSubParseFormat format;
- if (strlen (self->textbuf->str) < 35) {
+ if (strlen (self->textbuf->str) < 30) {
GST_DEBUG ("File too small to be a subtitles file");
return NULL;
}
g_free (data);
self->parser_type = format;
+ self->subtitle_codec = gst_sub_parse_get_format_description (format);
parser_state_init (&self->state);
switch (format) {
case GST_SUB_PARSE_FORMAT_MPSUB:
self->parse_line = parse_mpsub;
return gst_caps_new_simple ("text/plain", NULL);
+#ifndef GST_DISABLE_XML
case GST_SUB_PARSE_FORMAT_SAMI:
self->parse_line = parse_sami;
sami_context_init (&self->state);
return gst_caps_new_simple ("text/x-pango-markup", NULL);
+#endif
case GST_SUB_PARSE_FORMAT_TMPLAYER:
self->parse_line = parse_tmplayer;
+ self->state.max_duration = 5 * GST_SECOND;
return gst_caps_new_simple ("text/plain", NULL);
case GST_SUB_PARSE_FORMAT_MPL2:
self->parse_line = parse_mpl2;
return gst_caps_new_simple ("text/x-pango-markup", NULL);
+ case GST_SUB_PARSE_FORMAT_DKS:
+ self->parse_line = parse_dks;
+ return gst_caps_new_simple ("text/plain", NULL);
case GST_SUB_PARSE_FORMAT_SUBVIEWER:
self->parse_line = parse_subviewer;
return gst_caps_new_simple ("text/plain", NULL);
+ case GST_SUB_PARSE_FORMAT_QTTEXT:
+ self->parse_line = parse_qttext;
+ qttext_context_init (&self->state);
+ return gst_caps_new_simple ("text/x-pango-markup", NULL);
case GST_SUB_PARSE_FORMAT_UNKNOWN:
default:
GST_DEBUG ("no subtitle format detected");
static void
feed_textbuf (GstSubParse * self, GstBuffer * buf)
{
- if (GST_BUFFER_OFFSET (buf) != self->offset) {
+ gboolean discont;
+ gsize consumed;
+ gchar *input = NULL;
+ const guint8 *data;
+ gsize avail;
+
+ discont = GST_BUFFER_IS_DISCONT (buf);
+
+ if (GST_BUFFER_OFFSET_IS_VALID (buf) &&
+ GST_BUFFER_OFFSET (buf) != self->offset) {
+ self->offset = GST_BUFFER_OFFSET (buf);
+ discont = TRUE;
+ }
+
+ if (discont) {
+ GST_INFO ("discontinuity");
/* flush the parser state */
parser_state_init (&self->state);
g_string_truncate (self->textbuf, 0);
- sami_context_reset (&self->state);
+ gst_adapter_clear (self->adapter);
+#ifndef GST_DISABLE_XML
+ if (self->parser_type == GST_SUB_PARSE_FORMAT_SAMI)
+ sami_context_reset (&self->state);
+#endif
+ /* we could set a flag to make sure that the next buffer we push out also
+ * has the DISCONT flag set, but there's no point really given that it's
+ * subtitles which are discontinuous by nature. */
}
- self->textbuf = g_string_append_len (self->textbuf,
- (gchar *) GST_BUFFER_DATA (buf), GST_BUFFER_SIZE (buf));
- self->offset = GST_BUFFER_OFFSET (buf) + GST_BUFFER_SIZE (buf);
+ self->offset = GST_BUFFER_OFFSET (buf) + gst_buffer_get_size (buf);
self->next_offset = self->offset;
- gst_buffer_unref (buf);
+ gst_adapter_push (self->adapter, buf);
+
+ avail = gst_adapter_available (self->adapter);
+ data = gst_adapter_map (self->adapter, avail);
+ input = convert_encoding (self, (const gchar *) data, avail, &consumed);
+
+ if (input && consumed > 0) {
+ self->textbuf = g_string_append (self->textbuf, input);
+ gst_adapter_unmap (self->adapter, consumed);
+ } else {
+ gst_adapter_unmap (self->adapter, 0);
+ }
+
+ g_free (input);
}
static GstFlowReturn
{
GstFlowReturn ret = GST_FLOW_OK;
GstCaps *caps = NULL;
- gchar *line, *subtitle;
+ gchar *line, *subtitle, *data;
+ gsize size;
+
+ if (self->first_buffer) {
+ data = gst_buffer_map (buf, &size, NULL, GST_MAP_READ);
+ self->detected_encoding = detect_encoding (data, size);
+ gst_buffer_unmap (buf, data, size);
+ self->first_buffer = FALSE;
+ self->state.fps_n = self->fps_n;
+ self->state.fps_d = self->fps_d;
+ }
feed_textbuf (self, buf);
return GST_FLOW_UNEXPECTED;
}
gst_caps_unref (caps);
+
+ /* push tags */
+ if (self->subtitle_codec != NULL) {
+ GstTagList *tags;
+
+ tags = gst_tag_list_new ();
+ gst_tag_list_add (tags, GST_TAG_MERGE_APPEND, GST_TAG_SUBTITLE_CODEC,
+ self->subtitle_codec, NULL);
+ gst_element_found_tags_for_pad (GST_ELEMENT (self), self->srcpad, tags);
+ }
}
- while ((line = get_next_line (self)) && !self->flushing) {
+ while (!self->flushing && (line = get_next_line (self))) {
+ guint offset = 0;
+
/* Set segment on our parser state machine */
self->state.segment = &self->segment;
/* Now parse the line, out of segment lines will just return NULL */
- GST_LOG_OBJECT (self, "Parsing line '%s'", line);
- subtitle = self->parse_line (&self->state, line);
+ GST_LOG_OBJECT (self, "Parsing line '%s'", line + offset);
+ subtitle = self->parse_line (&self->state, line + offset);
g_free (line);
if (subtitle) {
guint subtitle_len = strlen (subtitle);
/* +1 for terminating NUL character */
- ret = gst_pad_alloc_buffer_and_set_caps (self->srcpad,
- GST_BUFFER_OFFSET_NONE, subtitle_len + 1,
- GST_PAD_CAPS (self->srcpad), &buf);
+ buf = gst_buffer_new_and_alloc (subtitle_len + 1);
+
+ /* copy terminating NUL character as well */
+ gst_buffer_fill (buf, 0, subtitle, subtitle_len + 1);
+ gst_buffer_set_size (buf, subtitle_len);
+
+ GST_BUFFER_TIMESTAMP (buf) = self->state.start_time;
+ GST_BUFFER_DURATION (buf) = self->state.duration;
+
+ /* in some cases (e.g. tmplayer) we can only determine the duration
+ * of a text chunk from the timestamp of the next text chunk; in those
+ * cases, we probably want to limit the duration to something
+ * reasonable, so we don't end up showing some text for e.g. 40 seconds
+ * just because nothing else is being said during that time */
+ if (self->state.max_duration > 0 && GST_BUFFER_DURATION_IS_VALID (buf)) {
+ if (GST_BUFFER_DURATION (buf) > self->state.max_duration)
+ GST_BUFFER_DURATION (buf) = self->state.max_duration;
+ }
- if (ret == GST_FLOW_OK) {
- /* copy terminating NUL character as well */
- memcpy (GST_BUFFER_DATA (buf), subtitle, subtitle_len + 1);
- GST_BUFFER_SIZE (buf) = subtitle_len;
- GST_BUFFER_TIMESTAMP (buf) = self->state.start_time;
- GST_BUFFER_DURATION (buf) = self->state.duration;
+ self->segment.position = self->state.start_time;
- gst_segment_set_last_stop (&self->segment, GST_FORMAT_TIME,
- self->state.start_time);
+ GST_DEBUG_OBJECT (self, "Sending text '%s', %" GST_TIME_FORMAT " + %"
+ GST_TIME_FORMAT, subtitle, GST_TIME_ARGS (self->state.start_time),
+ GST_TIME_ARGS (self->state.duration));
- GST_DEBUG_OBJECT (self, "Sending text '%s', %" GST_TIME_FORMAT " + %"
- GST_TIME_FORMAT, subtitle, GST_TIME_ARGS (self->state.start_time),
- GST_TIME_ARGS (self->state.duration));
+ ret = gst_pad_push (self->srcpad, buf);
- ret = gst_pad_push (self->srcpad, buf);
- }
+ /* move this forward (the tmplayer parser needs this) */
+ if (self->state.duration != GST_CLOCK_TIME_NONE)
+ self->state.start_time += self->state.duration;
g_free (subtitle);
subtitle = NULL;
GST_LOG_OBJECT (self, "pushing newsegment event with %" GST_SEGMENT_FORMAT,
&self->segment);
- gst_pad_push_event (self->srcpad, gst_event_new_new_segment (FALSE,
- self->segment.rate, self->segment.format,
- self->segment.last_stop, self->segment.stop, self->segment.time));
+ gst_pad_push_event (self->srcpad, gst_event_new_segment (&self->segment));
self->need_segment = FALSE;
}
/* Make sure the last subrip chunk is pushed out even
* if the file does not have an empty line at the end */
if (self->parser_type == GST_SUB_PARSE_FORMAT_SUBRIP ||
- self->parser_type == GST_SUB_PARSE_FORMAT_MPL2) {
- GstBuffer *buf = gst_buffer_new_and_alloc (1 + 1);
+ self->parser_type == GST_SUB_PARSE_FORMAT_TMPLAYER ||
+ self->parser_type == GST_SUB_PARSE_FORMAT_MPL2 ||
+ self->parser_type == GST_SUB_PARSE_FORMAT_QTTEXT) {
+ gchar term_chars[] = { '\n', '\n', '\0' };
+ GstBuffer *buf = gst_buffer_new_and_alloc (2 + 1);
GST_DEBUG ("EOS. Pushing remaining text (if any)");
- GST_BUFFER_DATA (buf)[0] = '\n';
- GST_BUFFER_DATA (buf)[1] = '\0'; /* play it safe */
- GST_BUFFER_SIZE (buf) = 1;
+ gst_buffer_fill (buf, 0, term_chars, 3);
+ gst_buffer_set_size (buf, 2);
+
GST_BUFFER_OFFSET (buf) = self->offset;
gst_sub_parse_chain (pad, buf);
}
ret = gst_pad_event_default (pad, event);
break;
}
- case GST_EVENT_NEWSEGMENT:
+ case GST_EVENT_SEGMENT:
{
- GstFormat format;
- gdouble rate;
- gint64 start, stop, time;
- gboolean update;
-
- gst_event_parse_new_segment (event, &update, &rate, &format, &start,
- &stop, &time);
-
- GST_DEBUG_OBJECT (self, "newsegment (%s)", gst_format_get_name (format));
-
- if (format == GST_FORMAT_TIME) {
- gst_segment_set_newsegment (&self->segment, update, rate, format,
- start, stop, time);
- } else {
- /* if not time format, we'll either start with a 0 timestamp anyway or
- * it's following a seek in which case we'll have saved the requested
- * seek segment and don't want to overwrite it (remember that on a seek
- * we always just seek back to the start in BYTES format and just throw
- * away all text that's before the requested position; if the subtitles
- * come from an upstream demuxer, it won't be able to handle our BYTES
- * seek request and instead send us a newsegment from the seek request
- * it received via its video pads instead, so all is fine then too) */
- }
-
+ gst_event_copy_segment (event, &self->segment);
+ GST_DEBUG_OBJECT (self, "newsegment (%s)",
+ gst_format_get_name (self->segment.format));
+
+ /* if not time format, we'll either start with a 0 timestamp anyway or
+ * it's following a seek in which case we'll have saved the requested
+ * seek segment and don't want to overwrite it (remember that on a seek
+ * we always just seek back to the start in BYTES format and just throw
+ * away all text that's before the requested position; if the subtitles
+ * come from an upstream demuxer, it won't be able to handle our BYTES
+ * seek request and instead send us a newsegment from the seek request
+ * it received via its video pads instead, so all is fine then too) */
ret = TRUE;
gst_event_unref (event);
break;
self->next_offset = 0;
self->parser_type = GST_SUB_PARSE_FORMAT_UNKNOWN;
self->valid_utf8 = TRUE;
+ self->first_buffer = TRUE;
+ g_free (self->detected_encoding);
+ self->detected_encoding = NULL;
g_string_truncate (self->textbuf, 0);
+ gst_adapter_clear (self->adapter);
break;
default:
break;
}
- ret = parent_class->change_state (element, transition);
+ ret = GST_ELEMENT_CLASS (parent_class)->change_state (element, transition);
if (ret == GST_STATE_CHANGE_FAILURE)
return ret;
switch (transition) {
case GST_STATE_CHANGE_PAUSED_TO_READY:
- parser_state_dispose (&self->state);
+ parser_state_dispose (self, &self->state);
self->parser_type = GST_SUB_PARSE_FORMAT_UNKNOWN;
break;
default:
* also, give different subtitle formats really different types */
static GstStaticCaps mpl2_caps =
GST_STATIC_CAPS ("application/x-subtitle-mpl2");
+#define SUB_CAPS (gst_static_caps_get (&sub_caps))
+
static GstStaticCaps tmp_caps =
GST_STATIC_CAPS ("application/x-subtitle-tmplayer");
-static GstStaticCaps smi_caps = GST_STATIC_CAPS ("application/x-subtitle-sami");
+#define TMP_CAPS (gst_static_caps_get (&tmp_caps))
+
static GstStaticCaps sub_caps = GST_STATIC_CAPS ("application/x-subtitle");
+#define MPL2_CAPS (gst_static_caps_get (&mpl2_caps))
-#define SUB_CAPS (gst_static_caps_get (&sub_caps))
+#ifndef GST_DISABLE_XML
+static GstStaticCaps smi_caps = GST_STATIC_CAPS ("application/x-subtitle-sami");
#define SAMI_CAPS (gst_static_caps_get (&smi_caps))
-#define TMP_CAPS (gst_static_caps_get (&tmp_caps))
-#define MPL2_CAPS (gst_static_caps_get (&mpl2_caps))
+#endif
+
+static GstStaticCaps dks_caps = GST_STATIC_CAPS ("application/x-subtitle-dks");
+#define DKS_CAPS (gst_static_caps_get (&dks_caps))
+
+static GstStaticCaps qttext_caps =
+GST_STATIC_CAPS ("application/x-subtitle-qttext");
+#define QTTEXT_CAPS (gst_static_caps_get (&qttext_caps))
static void
gst_subparse_type_find (GstTypeFind * tf, gpointer private)
const guint8 *data;
GstCaps *caps;
gchar *str;
+ gchar *encoding = NULL;
+ const gchar *end;
- if (!(data = gst_type_find_peek (tf, 0, 36)))
+ if (!(data = gst_type_find_peek (tf, 0, 129)))
return;
/* make sure string passed to _autodetect() is NUL-terminated */
- str = g_strndup ((gchar *) data, 35);
+ str = g_malloc0 (129);
+ memcpy (str, data, 128);
+
+ if ((encoding = detect_encoding (str, 128)) != NULL) {
+ gchar *converted_str;
+ GError *err = NULL;
+ gsize tmp;
+
+ converted_str = gst_convert_to_utf8 (str, 128, encoding, &tmp, &err);
+ if (converted_str == NULL) {
+ GST_DEBUG ("Encoding '%s' detected but conversion failed: %s", encoding,
+ err->message);
+ g_error_free (err);
+ g_free (encoding);
+ } else {
+ g_free (str);
+ str = converted_str;
+ g_free (encoding);
+ }
+ }
+
+ /* Check if at least the first 120 chars are valid UTF8,
+ * otherwise convert as always */
+ if (!g_utf8_validate (str, 128, &end) && (end - str) < 120) {
+ gchar *converted_str;
+ GError *err = NULL;
+ gsize tmp;
+ const gchar *enc;
+
+ enc = g_getenv ("GST_SUBTITLE_ENCODING");
+ if (enc == NULL || *enc == '\0') {
+ /* if local encoding is UTF-8 and no encoding specified
+ * via the environment variable, assume ISO-8859-15 */
+ if (g_get_charset (&enc)) {
+ enc = "ISO-8859-15";
+ }
+ }
+ converted_str = gst_convert_to_utf8 (str, 128, enc, &tmp, &err);
+ if (converted_str != NULL) {
+ g_free (str);
+ str = converted_str;
+ }
+ }
+
format = gst_sub_parse_data_format_autodetect (str);
g_free (str);
GST_DEBUG ("MPSub format detected");
caps = SUB_CAPS;
break;
+#ifndef GST_DISABLE_XML
case GST_SUB_PARSE_FORMAT_SAMI:
GST_DEBUG ("SAMI (time-based) format detected");
caps = SAMI_CAPS;
break;
+#endif
case GST_SUB_PARSE_FORMAT_TMPLAYER:
GST_DEBUG ("TMPlayer (time based) format detected");
caps = TMP_CAPS;
GST_DEBUG ("SubViewer format detected");
caps = SUB_CAPS;
break;
+ case GST_SUB_PARSE_FORMAT_DKS:
+ GST_DEBUG ("DKS format detected");
+ caps = DKS_CAPS;
+ break;
+ case GST_SUB_PARSE_FORMAT_QTTEXT:
+ GST_DEBUG ("QTtext format detected");
+ caps = QTTEXT_CAPS;
+ break;
default:
case GST_SUB_PARSE_FORMAT_UNKNOWN:
GST_DEBUG ("no subtitle format detected");
static gboolean
plugin_init (GstPlugin * plugin)
{
- static gchar *sub_exts[] = { "srt", "sub", "mpsub", "mdvd", "smi", "txt",
- NULL
- };
+ static const gchar *sub_exts[] =
+ { "srt", "sub", "mpsub", "mdvd", "smi", "txt", "dks", NULL };
GST_DEBUG_CATEGORY_INIT (sub_parse_debug, "subparse", 0, ".sub parser");
if (!gst_type_find_register (plugin, "subparse_typefind", GST_RANK_MARGINAL,
- gst_subparse_type_find, sub_exts, SUB_CAPS, NULL, NULL))
+ gst_subparse_type_find, (gchar **) sub_exts, SUB_CAPS, NULL, NULL))
return FALSE;
if (!gst_element_register (plugin, "subparse",