[FFmpeg-devel] [PATCH] Add tests for various video filters

Stefano Sabatini stefasab at gmail.com
Sun Feb 26 17:00:33 CET 2012


On date Saturday 2012-02-25 14:41:41 +0100, Nicolas George encoded:
> Hi.
> 
> The attached patches add FATE tests for various uncovered video filters and
> sources.
> 
> I have tried to keep the tests very short while enabling features. Clément's
> coverage page (thanks a lot!) will show how much these tests are reliable,
> and they can be extended later if necessary.
> 
> My way of computing the ref file was:
> 
> 1. Check the filter visually with -re -f sdl.
> 2. Add the test with a dummy ref file.
> 3. Run it and see it fail.
> 4. Check the error log.
> 5. Copy-paste the ref from the failed run.
> 6. Re-run the test to be sure.
> 
> I do not know if it could be done more seriously.
> 
> Regards,
> 
> -- 
>   Nicolas George

> From 7dc27e886a5754f7c700d869d661e1869d74eb92 Mon Sep 17 00:00:00 2001
> From: Nicolas George <nicolas.george at normalesup.org>
> Date: Sat, 25 Feb 2012 12:54:52 +0100
> Subject: [PATCH 01/14] fate: add tests for testsrc.
> 
> This test uses the lavfi device and
> is ready for more similar tests.
> 
> Signed-off-by: Nicolas George <nicolas.george at normalesup.org>
[...]

> From 63d138f8ac6c43ac4fdac58a6b1120cdad1a3f75 Mon Sep 17 00:00:00 2001
> From: Nicolas George <nicolas.george at normalesup.org>
> Date: Sat, 25 Feb 2012 12:56:59 +0100
> Subject: [PATCH 02/14] fate: add tests for the setsar and setdar video
>  filters.
> 
> 
> Signed-off-by: Nicolas George <nicolas.george at normalesup.org>
> ---
>  tests/lavfi-regression.sh |    2 ++
>  tests/ref/lavfi/setdar    |    1 +
>  tests/ref/lavfi/setsar    |    1 +
>  3 files changed, 4 insertions(+), 0 deletions(-)
>  create mode 100644 tests/ref/lavfi/setdar
>  create mode 100644 tests/ref/lavfi/setsar
> 
> diff --git a/tests/lavfi-regression.sh b/tests/lavfi-regression.sh
> index 84fb757..ac0f3f8 100755
> --- a/tests/lavfi-regression.sh
> +++ b/tests/lavfi-regression.sh
> @@ -35,6 +35,8 @@ do_lavfi "crop_vflip"         "crop=iw-100:ih-100:100:100,vflip"
>  do_lavfi "null"               "null"
>  do_lavfi "scale200"           "scale=200:200"
>  do_lavfi "scale500"           "scale=500:500"
> +do_lavfi "setdar"             "setdar=16/9"
> +do_lavfi "setsar"             "setsar=16/11"
>  do_lavfi "vflip"              "vflip"
>  do_lavfi "vflip_crop"         "vflip,crop=iw-100:ih-100:100:100"
>  do_lavfi "vflip_vflip"        "vflip,vflip"
> diff --git a/tests/ref/lavfi/setdar b/tests/ref/lavfi/setdar
> new file mode 100644
> index 0000000..af7dcd6
> --- /dev/null
> +++ b/tests/ref/lavfi/setdar
> @@ -0,0 +1 @@
> +setdar              f94e744d160c65af967315f0145612a7
> diff --git a/tests/ref/lavfi/setsar b/tests/ref/lavfi/setsar
> new file mode 100644
> index 0000000..3102600
> --- /dev/null
> +++ b/tests/ref/lavfi/setsar
> @@ -0,0 +1 @@
> +setsar              f94e744d160c65af967315f0145612a7

This is maybe a bit overkill, considering that the raw video data is
not affected at all.

[...]
> From 509fdaa2665c7bb337795a848c9427d2e10227f8 Mon Sep 17 00:00:00 2001
> From: Nicolas George <nicolas.george at normalesup.org>
> Date: Sat, 25 Feb 2012 14:21:29 +0100
> Subject: [PATCH 14/14] fate: add test for the cellauto video source.
> 
> 
> Signed-off-by: Nicolas George <nicolas.george at normalesup.org>
> ---
>  tests/lavfi-regression.sh |    1 +
>  tests/ref/lavfi/cellauto  |    1 +
>  2 files changed, 2 insertions(+), 0 deletions(-)
>  create mode 100644 tests/ref/lavfi/cellauto
> 
> diff --git a/tests/lavfi-regression.sh b/tests/lavfi-regression.sh
> index a272265..650ecf5 100755
> --- a/tests/lavfi-regression.sh
> +++ b/tests/lavfi-regression.sh
> @@ -97,6 +97,7 @@ do_lavfi_lavd() {
>          $ENC_OPTS -vcodec rawvideo $* -f nut md5:
>  }
>  
> +do_lavfi_lavd "cellauto"             "cellauto=r=5:s=40x40:seed=42" -t 2

Not that I think that people care about my little pet, but OK if you care ;-).

Rest of the tests look fine to me, thanks.
-- 
FFmpeg = Fascinating and Faithful Mortal Ponderous Embarassing Gigant


More information about the ffmpeg-devel mailing list