On 10/11/17 12:27, Diego Biurrun wrote:
> On Thu, Nov 09, 2017 at 01:07:47AM +0000, Mark Thompson wrote:
>> --- a/libavcodec/cbs.h
>> +++ b/libavcodec/cbs.h
>> @@ -25,6 +25,19 @@
>>
>> +/*
>> + * This defines a framework for converting between a coded bitstream
>> + * and structures defining all individual syntax elements found in
>> + * such a stream.
>> + *
>> + * Conversion in both directions is possible. Given a coded bitstream
>> + * (any meaningful fragment), it can be parsed and decomposed into
>> + * syntax elements stored in a set of codec-specific structures.
>> + * Similarly, given a set of those same codec-specific structures the
>> + * syntax elements can be serialised and combined to create a coded
>> + * bitstream.
>> + */
>> +
>> struct CodedBitstreamType;
>
> This looks like it should be Doxygen in one form or another, probably
> some sort of section.
Would it actually end up somewhere? This isn't public API.
>> @@ -114,7 +127,7 @@ typedef struct CodedBitstreamFragment {
>> /**
>> * Number of units in this fragment.
>> *
>> - * This may be zero if the fragment only exists in bistream form
>> + * This may be zero if the fragment only exists in bitstream form
>
> Always a classic :)
>
>> --- a/libavcodec/cbs_internal.h
>> +++ b/libavcodec/cbs_internal.h
>> @@ -32,6 +32,9 @@ typedef struct CodedBitstreamType {
>>
>> // Split frag->data into coded bitstream units, creating the
>> // frag->units array. Fill data but not content on each unit.
>> + // header is set if the fragment came from a header block, which
>> + // may require different parsing for some codecs (e.g. the AVCC
>> + // header in H.264).
>
> I'd capitalize "header" here.
It's referring to the argument "header", capitalising it would be wrong.
Maybe "The header argument should be set if the fragment ..." to move it away
from the start of the sentence?
- Mark
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel