[Bug 739281] video-blend: correct improper logic calculations

[GStreamer (bugzilla.gnome.org)]

https://bugzilla.gnome.org/show_bug.cgi?id=739281
  GStreamer | gst-plugins-base | unspecified

--- Comment #22 from Tim-Philipp Müller <t.i.m at zen.co.uk> 2014-11-02 15:38:06 UTC ---
Let's do this systematically :)

I would like to see the existing test extended into a full set of unit tests
for the following situations:

1) for an overlay rectangle that is smaller in width+height than the video
surface it is rendered on:
1a) if rendered completely left of video surface (x + overlay_width <= 0), no
blending takes place
1b) if rendered completely right of video surface (x >= video_width), no
blending takes place
1c) if rendered completely above video surface (y + overlay_height <= 0), no
blending takes place
1d) if rendered completely below video surface (y >= video_height), no blending
takes place
1e) if rendered partially left of video surface (x < 0 && -x < overlay_width),
overlay pixels overlay_width+x to overlay_width are blended.
1f) if rendered partially right of video surface, ...
1g) if rendered partially above video surface ...
1h) if rendered partially below video surface..

2) for an overlay rectangle that is wider than video_width, make sure things
still work properly with negative x and x < video_width

3) for an overlay rectangle that is higher than video_height, ...

(sorry!) :)

Ideally as a separate test. Should be less code than it sounds hopefully, just
a bit of copy'n'paste. We need to fill up the overlay with appropriate values
that allow us to test if the right bits are blittet at the right spot later
(where stuff is blitted).

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