[FFmpeg-devel] [PATCH] avformat/mov: add ignore_hoov option to ignore hoov atom

["zhilizhao(赵志立)"]

> On Dec 23, 2021, at 3:10 PM, Steven Liu <lq at chinaffmpeg.org> wrote:
> 
> Try to get context from the moov atom when the hoov before moov atom,
> because the streams info get a possible incorrect when there have both
> hoov and moov atom. So add and ignore_hoov option for try to get
> moov context by user.
> 
> Signed-off-by: Steven Liu <lq at chinaffmpeg.org>
> ---
> doc/demuxers.texi  |  6 ++++++
> libavformat/isom.h |  2 ++
> libavformat/mov.c  | 14 ++++++++++++--
> 3 files changed, 20 insertions(+), 2 deletions(-)
> 
> diff --git a/doc/demuxers.texi b/doc/demuxers.texi
> index cab8a7072c..d23c118210 100644
> --- a/doc/demuxers.texi
> +++ b/doc/demuxers.texi
> @@ -660,6 +660,11 @@ to demuxing linearly from the beginning. Default is true.
> Ignore any edit list atoms. The demuxer, by default, modifies the stream index to reflect the
> timeline described by the edit list. Default is false.
> 
> + at item ignore_hoov
> +Ignore hoov atom. The demuxer, by default, use hoov atom when there have hoov before moov atom.
> +You can try ignore hoov atom try the moov atom.
> +Default is false.
> +
> @item advanced_editlist
> Modify the stream index to reflect the timeline described by the edit list. @code{ignore_editlist}
> must be set to false for this option to be effective.
> diff --git a/libavformat/isom.h b/libavformat/isom.h
> index ef8f19b18c..6aa2b40c94 100644
> --- a/libavformat/isom.h
> +++ b/libavformat/isom.h
> @@ -255,6 +255,7 @@ typedef struct MOVContext {
>     AVFormatContext *fc;
>     int time_scale;
>     int64_t duration;     ///< duration of the longest track
> +    int found_hoov;       ///< 'hoov' atom has been found
>     int found_moov;       ///< 'moov' atom has been found
>     int found_mdat;       ///< 'mdat' atom has been found
>     int found_hdlr_mdta;  ///< 'hdlr' atom with type 'mdta' has been found
> @@ -273,6 +274,7 @@ typedef struct MOVContext {
>     unsigned int nb_chapter_tracks;
>     int use_absolute_path;
>     int ignore_editlist;
> +    int ignore_hoov;
>     int advanced_editlist;
>     int ignore_chapters;
>     int seek_individually;
> diff --git a/libavformat/mov.c b/libavformat/mov.c
> index 2aed6e80ef..cd775b219f 100644
> --- a/libavformat/mov.c
> +++ b/libavformat/mov.c
> @@ -1159,8 +1159,11 @@ static int mov_read_moov(MOVContext *c, AVIOContext *pb, MOVAtom atom)
> 
>     if (c->found_moov) {
>         av_log(c->fc, AV_LOG_WARNING, "Found duplicated MOOV Atom. Skipped it\n");
> -        avio_skip(pb, atom.size);
> -        return 0;
> +        if (!c->found_hoov) {
> +            avio_skip(pb, atom.size);
> +            return 0;
> +        }
> +        c->found_hoov = 0;
>     }
> 
>     if ((ret = mov_read_default(c, pb, atom)) < 0)
> @@ -4239,6 +4242,9 @@ static int mov_read_trak(MOVContext *c, AVIOContext *pb, MOVAtom atom)
>     MOVStreamContext *sc;
>     int ret;
> 
> +    if (c->found_hoov)
> +        return 0;
> +

Does it break files with only hoov boxes?

>     st = avformat_new_stream(c->fc, NULL);
>     if (!st) return AVERROR(ENOMEM);
>     st->id = -1;
> @@ -7329,6 +7335,8 @@ static int mov_read_default(MOVContext *c, AVIOContext *pb, MOVAtom atom)
>                 a.size >= 8 &&
>                 c->fc->strict_std_compliance < FF_COMPLIANCE_STRICT) {
>                 uint32_t type;
> +                if (c->ignore_hoov)
> +                    c->found_hoov = 1;
>                 avio_skip(pb, 4);
>                 type = avio_rl32(pb);
>                 if (avio_feof(pb))
> @@ -8541,6 +8549,8 @@ static const AVOption mov_options[] = {
>         0, 1, FLAGS},
>     {"ignore_editlist", "Ignore the edit list atom.", OFFSET(ignore_editlist), AV_OPT_TYPE_BOOL, {.i64 = 0},
>         0, 1, FLAGS},
> +    {"ignore_hoov", "Ignore the hoov atom.", OFFSET(ignore_hoov), AV_OPT_TYPE_BOOL, {.i64 = 0},
> +        0, 1, FLAGS},

TBH, add an option is too much effort to workaround such broken files. I prefer

a.type == MKTAG('h','o','o','v') && c->fc->strict_std_compliance < FF_COMPLIANCE_UNOFFICIAL

But it will break those samples if user doesn't lower strict_std_compliance.

>     {"advanced_editlist",
>         "Modify the AVIndex according to the editlists. Use this option to decode in the order specified by the edits.",
>         OFFSET(advanced_editlist), AV_OPT_TYPE_BOOL, {.i64 = 1},
> -- 
> 2.25.0
> 
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
> 
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".