[FFmpeg-devel] [PATCH] Split channel params from context in mlpdec.c

Benoit Fouet benoit.fouet
Wed Aug 13 09:36:45 CEST 2008


Michael Niedermayer wrote:
> On Tue, Aug 12, 2008 at 10:42:23PM -0300, Ramiro Polla wrote:
>   
>> On Tue, Aug 12, 2008 at 9:42 PM, Michael Niedermayer <michaelni at gmx.at> wrote:
>>     
>>> On Tue, Aug 12, 2008 at 09:21:34PM -0300, Ramiro Polla wrote:
>>>       
>>>> Hello,
>>>>
>>>> Attached patch splits channel params from MLP context into their own struct.
>>>>
>>>> I left quant_step_size where it is, waiting on the other thread about it...
>>>>
>>>> Ramiro Polla
>>>>         
>>>> Index: libavcodec/mlpdec.c
>>>> ===================================================================
>>>> --- libavcodec/mlpdec.c       (revision 14717)
>>>> +++ libavcodec/mlpdec.c       (working copy)
>>>> @@ -153,6 +153,20 @@
>>>>      int32_t     state[MAX_FILTER_ORDER];
>>>>  } FilterParams;
>>>>
>>>> +/** sample data coding information */
>>>> +typedef struct {
>>>> +    FilterParams filter_params[NUM_FILTERS];
>>>> +
>>>> +    //! Offset to apply to residual values.
>>>> +    int16_t     huff_offset;
>>>> +    //! sign/rounding-corrected version of huff_offset
>>>> +    int32_t     sign_huff_offset;
>>>> +    //! Which VLC codebook to use to read residuals.
>>>> +    uint8_t     codebook;
>>>> +    //! Size of residual suffix not encoded using VLC.
>>>> +    uint8_t     huff_lsbs;
>>>> +} ChannelParams;
>>>>         
>>> I think the doxy would be more readable to the right of vars instead of
>>> interleaved. (this is of course a seperate issue as the code is just
>>> moved around in this patch)
>>>       
>> I changed some that don't go too much beyond 80 columns. Is there a
>> way to have those comments split across many lines? I don't really
>> like going over 80 columns.
>>     
>
> int x; /**< aaaa
>             aaaa */
>
>
> IIRC
>
> the '<' just means the thing described is prior to the comment and
> without it its afterwards IIRC
>
>   

that's it

-- 
Benoit Fouet
Purple Labs S.A.
www.purplelabs.com




More information about the ffmpeg-devel mailing list