On 4/27/2025 5:50 PM, Andreas Rheinhardt wrote:
> James Almer:
>> apv_read_header() reads enough information that the generic demux code doesn't
>> attempt to read a frame to fill missing fields in codecpar, so make sure it's
>> set here.
>>
>> Signed-off-by: James Almer <jamrial@gmail.com>
>> ---
>>   libavformat/apvdec.c | 40 ++++++++++++++++++++++++++++++++++++++--
>>   1 file changed, 38 insertions(+), 2 deletions(-)
>>
>> diff --git a/libavformat/apvdec.c b/libavformat/apvdec.c
>> index e1ac34b003..9f8af35567 100644
>> --- a/libavformat/apvdec.c
>> +++ b/libavformat/apvdec.c
>> @@ -39,6 +39,11 @@ typedef struct APVHeaderInfo {
>>       uint8_t  chroma_format_idc;
>>       uint8_t  bit_depth_minus8;
>>   
>> +    uint8_t  color_primaries;
>> +    uint8_t  transfer_characteristics;
>> +    uint8_t  matrix_coefficients;
>> +    uint8_t  full_range_flag;
>> +
>>       enum AVPixelFormat pixel_format;
>>   } APVHeaderInfo;
>>   
>> @@ -111,6 +116,31 @@ static int apv_extract_header_info(APVHeaderInfo *info,
>>       if (zero != 0)
>>           return AVERROR_INVALIDDATA;
>>   
>> +    // Return if this function was called by apv_probe()
> 
> Why do you want to make probing less strict?

I'm not making it less strict, just ensuring apv_probe() behaves the 
exact same.
In any case, I'm withdrawing this patch in favor of a more proper 
approach. Will send it in a bit.