Re: [PATCH v6 0/3] MR3369: mf/tests: Test sample processing for MPEG4 media sink.

[Nikolay Sivov (＠nsivov)]

Nikolay Sivov (@nsivov) commented about dlls/mf/tests/mf.c:
> HRESULT (WINAPI *pMFCreateDXGIDeviceManager)(UINT *token, IMFDXGIDeviceManager **manager);
>  BOOL has_video_processor;
>
> +const static BYTE test_h264_header[] =
> +{
> - 0x00, 0x00, 0x01, 0x67, 0x64, 0x00, 0x14, 0xac, 0xd9, 0x46, 0x36, 0xc0,
> - 0x5a, 0x83, 0x03, 0x03, 0x52, 0x80, 0x00, 0x00, 0x03, 0x00, 0x80, 0x00,
> - 0x00, 0x03, 0x01, 0x47, 0x8a, 0x14, 0xcb, 0x00, 0x00, 0x01, 0x68, 0xeb,
> - 0xec, 0xb2, 0x2c,
>
> +};
> -
>
> +const static BYTE test_h264_frame[] =

Do we need this? Maybe it's enough to check the header or some elements of it actually being h264 stream, but full frame, it seems too much. Maybe we could use media foundation itself to check that output file type is detected as h264 when played back.

In my opinion this is a good example that tests are getting out of hand, when test callback is made to handle every possible case we need. I think it's fine to have a new mpeg4.c for tests.