/**
* SECTION:gstbaseparse
+ * @title: GstBaseParse
* @short_description: Base class for stream parsers
* @see_also: #GstBaseTransform
*
* into separate audio/video/whatever frames.
*
* It provides for:
- * <itemizedlist>
- * <listitem><para>provides one sink pad and one source pad</para></listitem>
- * <listitem><para>handles state changes</para></listitem>
- * <listitem><para>can operate in pull mode or push mode</para></listitem>
- * <listitem><para>handles seeking in both modes</para></listitem>
- * <listitem><para>handles events (SEGMENT/EOS/FLUSH)</para></listitem>
- * <listitem><para>
- * handles queries (POSITION/DURATION/SEEKING/FORMAT/CONVERT)
- * </para></listitem>
- * <listitem><para>handles flushing</para></listitem>
- * </itemizedlist>
+ *
+ * * provides one sink pad and one source pad
+ * * handles state changes
+ * * can operate in pull mode or push mode
+ * * handles seeking in both modes
+ * * handles events (SEGMENT/EOS/FLUSH)
+ * * handles queries (POSITION/DURATION/SEEKING/FORMAT/CONVERT)
+ * * handles flushing
*
* The purpose of this base class is to provide the basic functionality of
* a parser and share a lot of rather complex code.
*
- * Description of the parsing mechanism:
- * <orderedlist>
- * <listitem>
- * <itemizedlist><title>Set-up phase</title>
- * <listitem><para>
- * #GstBaseParse calls @start to inform subclass that data processing is
- * about to start now.
- * </para></listitem>
- * <listitem><para>
- * #GstBaseParse class calls @set_sink_caps to inform the subclass about
- * incoming sinkpad caps. Subclass could already set the srcpad caps
- * accordingly, but this might be delayed until calling
- * gst_base_parse_finish_frame() with a non-queued frame.
- * </para></listitem>
- * <listitem><para>
- * At least at this point subclass needs to tell the #GstBaseParse class
- * how big data chunks it wants to receive (min_frame_size). It can do
- * this with gst_base_parse_set_min_frame_size().
- * </para></listitem>
- * <listitem><para>
- * #GstBaseParse class sets up appropriate data passing mode (pull/push)
- * and starts to process the data.
- * </para></listitem>
- * </itemizedlist>
- * </listitem>
- * <listitem>
- * <itemizedlist>
- * <title>Parsing phase</title>
- * <listitem><para>
- * #GstBaseParse gathers at least min_frame_size bytes of data either
- * by pulling it from upstream or collecting buffers in an internal
- * #GstAdapter.
- * </para></listitem>
- * <listitem><para>
- * A buffer of (at least) min_frame_size bytes is passed to subclass with
- * @handle_frame. Subclass checks the contents and can optionally
- * return GST_FLOW_OK along with an amount of data to be skipped to find
- * a valid frame (which will result in a subsequent DISCONT).
- * If, otherwise, the buffer does not hold a complete frame,
- * @handle_frame can merely return and will be called again when additional
- * data is available. In push mode this amounts to an
- * additional input buffer (thus minimal additional latency), in pull mode
- * this amounts to some arbitrary reasonable buffer size increase.
- * Of course, gst_base_parse_set_min_frame_size() could also be used if a
- * very specific known amount of additional data is required.
- * If, however, the buffer holds a complete valid frame, it can pass
- * the size of this frame to gst_base_parse_finish_frame().
- * If acting as a converter, it can also merely indicate consumed input data
- * while simultaneously providing custom output data.
- * Note that baseclass performs some processing (such as tracking
- * overall consumed data rate versus duration) for each finished frame,
- * but other state is only updated upon each call to @handle_frame
- * (such as tracking upstream input timestamp).
- * </para><para>
- * Subclass is also responsible for setting the buffer metadata
- * (e.g. buffer timestamp and duration, or keyframe if applicable).
- * (although the latter can also be done by #GstBaseParse if it is
- * appropriately configured, see below). Frame is provided with
- * timestamp derived from upstream (as much as generally possible),
- * duration obtained from configuration (see below), and offset
- * if meaningful (in pull mode).
- * </para><para>
- * Note that @check_valid_frame might receive any small
- * amount of input data when leftover data is being drained (e.g. at EOS).
- * </para></listitem>
- * <listitem><para>
- * As part of finish frame processing,
- * just prior to actually pushing the buffer in question,
- * it is passed to @pre_push_frame which gives subclass yet one
- * last chance to examine buffer metadata, or to send some custom (tag)
- * events, or to perform custom (segment) filtering.
- * </para></listitem>
- * <listitem><para>
- * During the parsing process #GstBaseParseClass will handle both srcpad
- * and sinkpad events. They will be passed to subclass if @event or
- * @src_event callbacks have been provided.
- * </para></listitem>
- * </itemizedlist>
- * </listitem>
- * <listitem>
- * <itemizedlist><title>Shutdown phase</title>
- * <listitem><para>
- * #GstBaseParse class calls @stop to inform the subclass that data
- * parsing will be stopped.
- * </para></listitem>
- * </itemizedlist>
- * </listitem>
- * </orderedlist>
+ * # Description of the parsing mechanism:
+ *
+ * ## Set-up phase
+ *
+ * * #GstBaseParse calls @start to inform subclass that data processing is
+ * about to start now.
+ *
+ * * #GstBaseParse class calls @set_sink_caps to inform the subclass about
+ * incoming sinkpad caps. Subclass could already set the srcpad caps
+ * accordingly, but this might be delayed until calling
+ * gst_base_parse_finish_frame() with a non-queued frame.
+ *
+ * * At least at this point subclass needs to tell the #GstBaseParse class
+ * how big data chunks it wants to receive (min_frame_size). It can do
+ * this with gst_base_parse_set_min_frame_size().
+ *
+ * * #GstBaseParse class sets up appropriate data passing mode (pull/push)
+ * and starts to process the data.
+ *
+ * ## Parsing phase
+ *
+ * * #GstBaseParse gathers at least min_frame_size bytes of data either
+ * by pulling it from upstream or collecting buffers in an internal
+ * #GstAdapter.
+ *
+ * * A buffer of (at least) min_frame_size bytes is passed to subclass with
+ * @handle_frame. Subclass checks the contents and can optionally
+ * return GST_FLOW_OK along with an amount of data to be skipped to find
+ * a valid frame (which will result in a subsequent DISCONT).
+ * If, otherwise, the buffer does not hold a complete frame,
+ * @handle_frame can merely return and will be called again when additional
+ * data is available. In push mode this amounts to an
+ * additional input buffer (thus minimal additional latency), in pull mode
+ * this amounts to some arbitrary reasonable buffer size increase.
+ * Of course, gst_base_parse_set_min_frame_size() could also be used if a
+ * very specific known amount of additional data is required.
+ * If, however, the buffer holds a complete valid frame, it can pass
+ * the size of this frame to gst_base_parse_finish_frame().
+ * If acting as a converter, it can also merely indicate consumed input data
+ * while simultaneously providing custom output data.
+ * Note that baseclass performs some processing (such as tracking
+ * overall consumed data rate versus duration) for each finished frame,
+ * but other state is only updated upon each call to @handle_frame
+ * (such as tracking upstream input timestamp).
+ *
+ * Subclass is also responsible for setting the buffer metadata
+ * (e.g. buffer timestamp and duration, or keyframe if applicable).
+ * (although the latter can also be done by #GstBaseParse if it is
+ * appropriately configured, see below). Frame is provided with
+ * timestamp derived from upstream (as much as generally possible),
+ * duration obtained from configuration (see below), and offset
+ * if meaningful (in pull mode).
+ *
+ * Note that @check_valid_frame might receive any small
+ * amount of input data when leftover data is being drained (e.g. at EOS).
+ *
+ * * As part of finish frame processing,
+ * just prior to actually pushing the buffer in question,
+ * it is passed to @pre_push_frame which gives subclass yet one
+ * last chance to examine buffer metadata, or to send some custom (tag)
+ * events, or to perform custom (segment) filtering.
+ *
+ * * During the parsing process #GstBaseParseClass will handle both srcpad
+ * and sinkpad events. They will be passed to subclass if @event or
+ * @src_event callbacks have been provided.
+ *
+ * ## Shutdown phase
+ *
+ * * #GstBaseParse class calls @stop to inform the subclass that data
+ * parsing will be stopped.
*
* Subclass is responsible for providing pad template caps for
* source and sink pads. The pads need to be named "sink" and "src". It also
* and end of data processing.
*
* Things that subclass need to take care of:
- * <itemizedlist>
- * <listitem><para>Provide pad templates</para></listitem>
- * <listitem><para>
- * Fixate the source pad caps when appropriate
- * </para></listitem>
- * <listitem><para>
- * Inform base class how big data chunks should be retrieved. This is
- * done with gst_base_parse_set_min_frame_size() function.
- * </para></listitem>
- * <listitem><para>
- * Examine data chunks passed to subclass with @handle_frame and pass
- * proper frame(s) to gst_base_parse_finish_frame(), and setting src pad
- * caps and timestamps on frame.
- * </para></listitem>
- * <listitem><para>Provide conversion functions</para></listitem>
- * <listitem><para>
- * Update the duration information with gst_base_parse_set_duration()
- * </para></listitem>
- * <listitem><para>
- * Optionally passthrough using gst_base_parse_set_passthrough()
- * </para></listitem>
- * <listitem><para>
- * Configure various baseparse parameters using
- * gst_base_parse_set_average_bitrate(), gst_base_parse_set_syncable()
- * and gst_base_parse_set_frame_rate().
- * </para></listitem>
- * <listitem><para>
- * In particular, if subclass is unable to determine a duration, but
- * parsing (or specs) yields a frames per seconds rate, then this can be
- * provided to #GstBaseParse to enable it to cater for
- * buffer time metadata (which will be taken from upstream as much as
- * possible). Internally keeping track of frame durations and respective
- * sizes that have been pushed provides #GstBaseParse with an estimated
- * bitrate. A default @convert (used if not overridden) will then use these
- * rates to perform obvious conversions. These rates are also used to
- * update (estimated) duration at regular frame intervals.
- * </para></listitem>
- * </itemizedlist>
+ *
+ * * Provide pad templates
+ * * Fixate the source pad caps when appropriate
+ * * Inform base class how big data chunks should be retrieved. This is
+ * done with gst_base_parse_set_min_frame_size() function.
+ * * Examine data chunks passed to subclass with @handle_frame and pass
+ * proper frame(s) to gst_base_parse_finish_frame(), and setting src pad
+ * caps and timestamps on frame.
+ * * Provide conversion functions
+ * * Update the duration information with gst_base_parse_set_duration()
+ * * Optionally passthrough using gst_base_parse_set_passthrough()
+ * * Configure various baseparse parameters using
+ * gst_base_parse_set_average_bitrate(), gst_base_parse_set_syncable()
+ * and gst_base_parse_set_frame_rate().
+ *
+ * * In particular, if subclass is unable to determine a duration, but
+ * parsing (or specs) yields a frames per seconds rate, then this can be
+ * provided to #GstBaseParse to enable it to cater for
+ * buffer time metadata (which will be taken from upstream as much as
+ * possible). Internally keeping track of frame durations and respective
+ * sizes that have been pushed provides #GstBaseParse with an estimated
+ * bitrate. A default @convert (used if not overridden) will then use these
+ * rates to perform obvious conversions. These rates are also used to
+ * update (estimated) duration at regular frame intervals.
*
*/
static gboolean gst_base_parse_src_query_default (GstBaseParse * parse,
GstQuery * query);
-static void gst_base_parse_drain (GstBaseParse * parse);
-
static gint64 gst_base_parse_find_offset (GstBaseParse * parse,
GstClockTime time, gboolean before, GstClockTime * _ts);
static GstFlowReturn gst_base_parse_locate_time (GstBaseParse * parse,
}
}
-static GstBaseParseFrame *
+GstBaseParseFrame *
gst_base_parse_frame_copy (GstBaseParseFrame * frame)
{
GstBaseParseFrame *copy;
GST_INFO_OBJECT (parse,
"Chose default caps %" GST_PTR_FORMAT " for initial gap", default_caps);
- gst_caps_unref (sinkcaps);
+ if (sinkcaps)
+ gst_caps_unref (sinkcaps);
gst_caps_unref (caps);
return default_caps;
return TRUE;
}
+ if (parse->priv->upstream_format != GST_FORMAT_BYTES) {
+ /* don't do byte format conversions if we're not really parsing
+ * a raw elementary stream, since we don't really have BYTES
+ * position / duration info */
+ if (src_format == GST_FORMAT_BYTES || dest_format == GST_FORMAT_BYTES)
+ goto no_slaved_conversions;
+ }
+
/* need at least some frames */
if (!parse->priv->framecount)
goto no_framecount;
G_GUINT64_FORMAT, duration, bytes);
return FALSE;
}
-
+no_slaved_conversions:
+ {
+ GST_DEBUG_OBJECT (parse,
+ "Can't do format conversions when upstream format is not BYTES");
+ return FALSE;
+ }
}
static void
gst_base_parse_update_bitrates (GstBaseParse * parse, GstBaseParseFrame * frame)
{
guint64 data_len, frame_dur;
- gint overhead, frame_bitrate;
+ gint overhead;
+ guint frame_bitrate;
+ guint64 frame_bitrate64;
GstBuffer *buffer = frame->buffer;
overhead = frame->overhead;
/* duration should be valid by now,
* either set by subclass or maybe based on fps settings */
if (GST_BUFFER_DURATION_IS_VALID (buffer) && parse->priv->acc_duration != 0) {
+ guint64 avg_bitrate;
+
/* Calculate duration of a frame from buffer properties */
frame_dur = GST_BUFFER_DURATION (buffer);
- parse->priv->avg_bitrate = (8 * parse->priv->data_bytecount * GST_SECOND) /
- parse->priv->acc_duration;
+ avg_bitrate = gst_util_uint64_scale (GST_SECOND,
+ 8 * parse->priv->data_bytecount, parse->priv->acc_duration);
+ if (avg_bitrate > G_MAXUINT)
+ return;
+
+ parse->priv->avg_bitrate = avg_bitrate;
} else {
/* No way to figure out frame duration (is this even possible?) */
return;
if (parse->priv->bitrate) {
parse->priv->avg_bitrate = parse->priv->bitrate;
/* spread this (confirmed) info ASAP */
- if (parse->priv->posted_avg_bitrate != parse->priv->avg_bitrate)
+ if (parse->priv->post_avg_bitrate &&
+ parse->priv->posted_avg_bitrate != parse->priv->avg_bitrate)
parse->priv->tags_changed = TRUE;
}
- if (frame_dur)
- frame_bitrate = (8 * data_len * GST_SECOND) / frame_dur;
- else
+ if (!frame_dur)
+ return;
+
+ frame_bitrate64 = gst_util_uint64_scale (GST_SECOND, 8 * data_len, frame_dur);
+
+ if (frame_bitrate64 > G_MAXUINT)
return;
+ frame_bitrate = (guint) frame_bitrate64;
+
GST_LOG_OBJECT (parse, "frame bitrate %u, avg bitrate %u", frame_bitrate,
parse->priv->avg_bitrate);
return ret;
}
-/* gst_base_parse_drain:
+/**
+ * gst_base_parse_drain:
+ * @parse: a #GstBaseParse
*
* Drains the adapter until it is empty. It decreases the min_frame_size to
* match the current adapter size and calls chain method until the adapter
* is emptied or chain returns with error.
+ *
+ * Since: 1.12
*/
-static void
+void
gst_base_parse_drain (GstBaseParse * parse)
{
guint avail;
if (first) {
GST_BUFFER_FLAG_SET (buf, GST_BUFFER_FLAG_DISCONT);
first = FALSE;
+ } else {
+ /* likewise, subsequent buffers should never have DISCONT
+ * according to the "reverse fragment protocol", or such would
+ * confuse a downstream decoder
+ * (could be DISCONT due to aggregating upstream fragments by parsing) */
+ GST_BUFFER_FLAG_UNSET (buf, GST_BUFFER_FLAG_DISCONT);
}
/* iterate output queue an push downstream */
av = gst_adapter_available (parse->priv->adapter);
if (G_UNLIKELY (parse->priv->drain)) {
- GST_DEBUG_OBJECT (parse, "draining, data left: %u, min %u", av, min_size);
- /* pass all available data to subclass, not just the minimum,
- * but never pass less than the minimum required to the subclass */
- if (av >= min_size) {
- min_size = av;
- if (G_UNLIKELY (!min_size))
- goto done;
- } else if (av > 0) {
- GST_DEBUG_OBJECT (parse, "draining, but not enough data available, "
- "discarding %u bytes", av);
- gst_adapter_clear (parse->priv->adapter);
+ min_size = av;
+ GST_DEBUG_OBJECT (parse, "draining, data left: %d", min_size);
+ if (G_UNLIKELY (!min_size)) {
goto done;
}
}
ret = old_ret;
goto done;
}
- old_ret = ret;
+ if (old_ret == GST_FLOW_OK)
+ old_ret = ret;
}
done:
goto baseparse_push;
parse->priv->push_stream_start = TRUE;
+ /* In pull mode, upstream is BYTES */
+ parse->priv->upstream_format = GST_FORMAT_BYTES;
return gst_pad_start_task (sinkpad, (GstTaskFunction) gst_base_parse_loop,
sinkpad, NULL);
result = klass->stop (parse);
parse->priv->pad_mode = GST_PAD_MODE_NONE;
+ parse->priv->upstream_format = GST_FORMAT_UNDEFINED;
}
GST_DEBUG_OBJECT (parse, "activate return: %d", result);
return result;
gst_util_uint64_scale (GST_SECOND, fps_den * lead_out, fps_num);
/* aim for about 1.5s to estimate duration */
if (parse->priv->update_interval < 0) {
- parse->priv->update_interval = fps_num * 3 / (fps_den * 2);
+ guint64 interval = gst_util_uint64_scale (fps_num, 3,
+ G_GUINT64_CONSTANT (2) * fps_den);
+
+ parse->priv->update_interval = MIN (interval, G_MAXINT);
+
GST_LOG_OBJECT (parse, "estimated update interval to %d frames",
parse->priv->update_interval);
}
if (!res) {
/* Fall back on interpreting segment */
GST_OBJECT_LOCK (parse);
- if (format == GST_FORMAT_BYTES) {
+ /* Only reply BYTES if upstream is in BYTES already, otherwise
+ * we're not in charge */
+ if (format == GST_FORMAT_BYTES
+ && parse->priv->upstream_format == GST_FORMAT_BYTES) {
dest_value = parse->priv->offset;
res = TRUE;
} else if (format == parse->segment.format &&
res = TRUE;
}
GST_OBJECT_UNLOCK (parse);
- if (!res) {
+ if (!res && parse->priv->upstream_format == GST_FORMAT_BYTES) {
/* no precise result, upstream no idea either, then best estimate */
- /* priv->offset is updated in both PUSH/PULL modes */
+ /* priv->offset is updated in both PUSH/PULL modes, *iff* we're
+ * in charge of things */
res = gst_base_parse_convert (parse,
GST_FORMAT_BYTES, parse->priv->offset, format, &dest_value);
}
if (!gst_base_parse_get_duration (parse, GST_FORMAT_TIME, &duration)
|| duration == -1) {
/* seekable if we still have a chance to get duration later on */
- seekable =
- parse->priv->upstream_seekable && parse->priv->update_interval;
+ seekable = parse->priv->upstream_seekable &&
+ (parse->priv->update_interval > 0);
} else {
seekable = parse->priv->upstream_seekable;
GST_LOG_OBJECT (parse, "already determine upstream seekabled: %d",
gst_pad_push_event (parse->srcpad, gst_event_ref (fevent));
gst_pad_push_event (parse->sinkpad, fevent);
gst_base_parse_clear_queues (parse);
- } else {
- /* keep track of our position */
- seeksegment.base = gst_segment_to_running_time (&seeksegment,
- seeksegment.format, parse->segment.position);
}
memcpy (&parse->segment, &seeksegment, sizeof (GstSegment));