Re: [FFmpeg-devel] [PATCH V2] [RFC] GSoC: FLIF16 Image format parser

2020-03-02 Thread Nicolas George
Anamitra Ghorui (12020-03-02): > According to the specification of the file format, there is no mention of an > upper bound for the integer: https://flif.info/spec.html#_part_1_main_header But they are mapped to FFmpeg values: width and height are ints, frame counts are uint64_t at best. Therefore

Re: [FFmpeg-devel] [PATCH V2] [RFC] GSoC: FLIF16 Image format parser

2020-03-02 Thread Anamitra Ghorui
Hello, > Is there an upper bound for the size of these integers? In my > experience, varints in multimedia code are used to save a few bits in > the file than to allow very large integers. You could make the code much > simpler if you had an upper bound. > > Also, there is probably already simila

Re: [FFmpeg-devel] [PATCH V2] [RFC] GSoC: FLIF16 Image format parser

2020-03-02 Thread Anton Khirnov
Quoting Nicolas George (2020-03-01 19:59:11) > > > +} varint_t; > > Names ending in _t are reserved by the C standard. FFmpeg uses types > with capitals and FF prefix for private APIs. FF prefix is not needed for structs. -- Anton Khirnov ___ ffmpeg-

Re: [FFmpeg-devel] [PATCH V2] [RFC] GSoC: FLIF16 Image format parser

2020-03-01 Thread Nicolas George
Anamitra Ghorui (12020-03-01): > Hello, > The parser can now deduce the structure of the format upto the point where the > compressed bitstream actually starts. It also includes functions for handling > the variable integers (varints). The parsing code has not been tested yet and > I'll be doing th