[Webkit-unassigned] [Bug 30004] [GStreamer] Should handle BUFFERING messages

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=30004





--- Comment #17 from Xan Lopez <xan.lopez at gmail.com>  2010-01-22 02:06:21 PST ---
(From update of attachment 46262)
Just a few nitpicks on top of gustavo's comments.

>+#include <glib/gstdio.h>

You should not include anything other than the toplevel glib header.

> #include <gst/gst.h>
> #include <gst/interfaces/mixer.h>
> #include <gst/interfaces/xoverlay.h>
>@@ -48,6 +49,18 @@
> #include <math.h>
> #include <wtf/gtk/GOwnPtr.h>
> 

I have been informed that this stuff is not public and that you have to copy it
like this in order to use it, but a comment saying just that would be good
IMHO.

>+// GstPlayFlags flags from playbin2
>+typedef enum {
>+    GST_PLAY_FLAG_VIDEO         = 0x01,
>+    GST_PLAY_FLAG_AUDIO         = 0x02,
>+    GST_PLAY_FLAG_TEXT          = 0x04,
>+    GST_PLAY_FLAG_VIS           = 0x08,
>+    GST_PLAY_FLAG_SOFT_VOLUME   = 0x10,
>+    GST_PLAY_FLAG_NATIVE_AUDIO  = 0x20,
>+    GST_PLAY_FLAG_NATIVE_VIDEO  = 0x40,
>+    GST_PLAY_FLAG_DOWNLOAD      = 0x80
>+} GstPlayFlags;
>+



>+void MediaPlayerPrivate::processBufferingStats(GstMessage* message)
>+{
>+    GstBufferingMode mode;
>+
>+    gst_message_parse_buffering_stats(message, &mode, 0, 0, 0);
>+    if (mode != GST_BUFFERING_DOWNLOAD)
>+        return;
>+
>+    if (!m_startedBuffering) {
>+        m_startedBuffering = true;
>+
>+        if (m_fillTimeoutId > 0) {
>+            g_source_remove(m_fillTimeoutId);
>+            m_fillTimeoutId = 0;

I guess there's really no need to set the id to anything if you are going to
write over it in the next line.

>+        }
>+
>+        m_fillTimeoutId = g_timeout_add(200, (GSourceFunc) bufferingTimeoutCallback, this);
>+    }
>+}
>+
>+bool MediaPlayerPrivate::queryBufferingStats()
>+{
>+    GstQuery* query = gst_query_new_buffering(GST_FORMAT_PERCENT);
>+
>+    if (gst_element_query(m_playBin, query)) {
>+        gint64 start, stop;
>+        gdouble fillStatus;
>+
>+        gst_query_parse_buffering_range(query, 0, &start, &stop, 0);
>+
>+        if (stop != -1)
>+            fillStatus = 100.0 * stop / GST_FORMAT_PERCENT_MAX;
>+        else
>+            fillStatus = 100.0;
>+
>+        LOG_VERBOSE(Media, "Download buffer filled up to %f%%", fillStatus);
>+
>+        // Update maxTimeLoaded only if the media duration is
>+        // available. Otherwise we can't compute it.
>+        if (!m_mediaDuration) {
>+            float newDuration = duration();
>+            if (!isinf(newDuration)) {
>+                m_mediaDuration = newDuration;
>+                durationChanged();
>+            }
>+        }
>+
>+        if (m_mediaDuration) {
>+            m_maxTimeLoaded = static_cast<float>((fillStatus * m_mediaDuration) / 100.0);
>+            LOG_VERBOSE(Media, "Updated maxTimeLoaded: %f", m_maxTimeLoaded);
>+        }
>+
>+        if (fillStatus == 100.0) {
>+            // Buffering is done, remove the fill source from the main loop.
>+            m_fillTimeoutId = 0;

Unless I'm missing something you are not actually removing any source here.


>+            gst_query_unref(query);
>+
>+            // Media is now fully loaded. It will play even if network
>+            // connection is cut.
>+            m_networkState = MediaPlayer::Loaded;
>+            m_player->networkStateChanged();
>+
>+            return FALSE;
>+        }
>+
>+        if ((fillStatus > 0) && (m_readyState < MediaPlayer::HaveFutureData)) {
>+            // Buffering started, we should now have enough data to
>+            // start playback if it was requested.
>+            m_readyState = MediaPlayer::HaveFutureData;
>+            m_player->readyStateChanged();
>+            if (m_paused && m_playRequired) {
>+                m_playRequired = false;
>+                gst_element_set_state(m_playBin, GST_STATE_PLAYING);
>+            }
>+        }
>+    }
>+
>+    gst_query_unref(query);
>+    return TRUE;
>+}
>+

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.