Re: [FFmpeg-devel] [PATCH] avformat/mov: Add support for still image AVIF parsing

[Vignesh Venkatasubramanian <vigneshv-at-google.com@ffmpeg.org>]

On Wed, Mar 16, 2022 at 5:41 AM Lynne <dev@lynne.ee> wrote:
>
> 15 Mar 2022, 22:38 by vigneshv-at-google.com@ffmpeg.org:
>
> > On Thu, Mar 3, 2022 at 3:58 PM Vignesh Venkatasubramanian
> > <vigneshv@google.com> wrote:
> >
> >>
> >> Add support for parsing AVIF still images. This patches supports
> >> AVIF still images that have exactly 1 item (i.e.) no alpha channel.
> >> Essentially, we will have to parse the "iloc" box and populate
> >> the mov index.
> >>
> >> With this patch, we can decode still AVIF images like so:
> >> ffmpeg -i image.avif image.png
> >>
> >> Partially fixes trac ticket #7621
> >>
> >> Signed-off-by: Vignesh Venkatasubramanian <vigneshv@google.com>
> >> ---
> >>  libavformat/isom.h |   1 +
> >>  libavformat/mov.c  | 142 +++++++++++++++++++++++++++++++++++++++++++++
> >>  2 files changed, 143 insertions(+)
> >>
> >> diff --git a/libavformat/isom.h b/libavformat/isom.h
> >> index 625dea8421..cc0a8e1ca9 100644
> >> --- a/libavformat/isom.h
> >> +++ b/libavformat/isom.h
> >> @@ -306,6 +306,7 @@ typedef struct MOVContext {
> >>  int have_read_mfra_size;
> >>  uint32_t mfra_size;
> >>  uint32_t max_stts_delta;
> >> +    int is_still_picture_avif;
> >>  } MOVContext;
> >>
> >>  int ff_mp4_read_descr_len(AVIOContext *pb);
> >> diff --git a/libavformat/mov.c b/libavformat/mov.c
> >> index 68b6d7f075..d260024e47 100644
> >> --- a/libavformat/mov.c
> >> +++ b/libavformat/mov.c
> >> @@ -1126,6 +1126,7 @@ static int mov_read_ftyp(MOVContext *c, AVIOContext *pb, MOVAtom atom)
> >>  c->isom = 1;
> >>  av_log(c->fc, AV_LOG_DEBUG, "ISO: File Type Major Brand: %.4s\n",(char *)&type);
> >>  av_dict_set(&c->fc->metadata, "major_brand", type, 0);
> >> +    c->is_still_picture_avif = !strncmp(type, "avif", 4);
> >>  minor_ver = avio_rb32(pb); /* minor version */
> >>  av_dict_set_int(&c->fc->metadata, "minor_version", minor_ver, 0);
> >>
> >> @@ -7173,6 +7174,146 @@ cleanup:
> >>  return ret;
> >>  }
> >>
> >> +static int rb_size(AVIOContext *pb, uint64_t* value, int size)
> >> +{
> >> +    if (size == 0) {
> >> +        *value = 0;
> >> +    } else if (size == 1) {
> >> +        *value = avio_r8(pb);
> >> +    } else if (size == 2) {
> >> +        *value = avio_rb16(pb);
> >> +    } else if (size == 4) {
> >> +        *value = avio_rb32(pb);
> >> +    } else if (size == 8) {
> >> +        *value = avio_rb64(pb);
> >> +    } else {
> >> +        return -1;
> >> +    }
> >> +    return size;
> >> +}
> >> +
> >> +static int mov_read_iloc(MOVContext *c, AVIOContext *pb, MOVAtom atom)
> >> +{
> >> +    int version, offset_size, length_size, base_offset_size, index_size;
> >> +    int item_count, extent_count;
> >> +    uint64_t base_offset, extent_offset, extent_length;
> >> +    int i, j;
> >> +    uint8_t value;
> >> +    AVStream *st;
> >> +    MOVStreamContext *sc;
> >> +
> >> +    if (!c->is_still_picture_avif) {
> >> +        // * For non-avif, we simply ignore the iloc box.
> >> +        // * For animated avif, we don't care about the iloc box as all the
> >> +        //   necessary information can be found in the moov box.
> >> +        return 0;
> >> +    }
> >> +
> >> +    if (c->fc->nb_streams != 0) {
> >> +        av_log(c->fc, AV_LOG_INFO, "Duplicate iloc box found\n");
> >> +        return 0;
> >> +    }
> >> +
> >> +    st = avformat_new_stream(c->fc, NULL);
> >> +    if (!st) return AVERROR(ENOMEM);
> >> +    st->id = c->fc->nb_streams;
> >> +    sc = av_mallocz(sizeof(MOVStreamContext));
> >> +    if (!sc) return AVERROR(ENOMEM);
> >> +
> >> +    st->priv_data = sc;
> >> +    st->codecpar->codec_type = AVMEDIA_TYPE_VIDEO;
> >> +    st->codecpar->codec_id = AV_CODEC_ID_AV1;
> >> +    sc->ffindex = st->index;
> >> +    c->trak_index = st->index;
> >> +    st->avg_frame_rate.num = st->avg_frame_rate.den = 1;
> >> +    st->time_base.num = st->time_base.den = 1;
> >> +    st->nb_frames = 1;
> >> +    sc->time_scale = 1;
> >> +    sc = st->priv_data;
> >> +    sc->pb = c->fc->pb;
> >> +    sc->pb_is_copied = 1;
> >> +
> >> +    version = avio_r8(pb);
> >> +    avio_rb24(pb);  // flags.
> >> +
> >> +    value = avio_r8(pb);
> >> +    offset_size = (value >> 4) & 0xF;
> >> +    length_size = value & 0xF;
> >> +    value = avio_r8(pb);
> >> +    base_offset_size = (value >> 4) & 0xF;
> >> +    index_size = (version == 0) ? 0 : (value & 0xF);
> >> +    if (index_size != 0) {
> >> +        return AVERROR_PATCHWELCOME;
> >> +    }
> >> +    item_count = (version < 2) ? avio_rb16(pb) : avio_rb32(pb);
> >> +    if (item_count > 1) {
> >> +        // For still AVIF images, we only support one item. Second item will
> >> +        // generally be found for AVIF images with alpha channel. We don't
> >> +        // support them as of now.
> >> +        return AVERROR_PATCHWELCOME;
> >> +    }
> >> +
> >> +    // Populate the necessary fields used by mov_build_index.
> >> +    sc->stsc_count = item_count;
> >> +    sc->stsc_data = av_malloc_array(item_count, sizeof(*sc->stsc_data));
> >> +    if (!sc->stsc_data) {
> >> +        return AVERROR(ENOMEM);
> >> +    }
> >> +    sc->stsc_data[0].first = 1;
> >> +    sc->stsc_data[0].count = 1;
> >> +    sc->stsc_data[0].id = 1;
> >> +    sc->chunk_count = item_count;
> >> +    sc->chunk_offsets = av_malloc_array(item_count, sizeof(*sc->chunk_offsets));
> >> +    if (!sc->chunk_offsets) {
> >> +        return AVERROR(ENOMEM);
> >> +    }
> >> +    sc->sample_count = item_count;
> >> +    sc->sample_sizes = av_malloc_array(item_count, sizeof(*sc->sample_sizes));
> >> +    if (!sc->sample_sizes) {
> >> +        return AVERROR(ENOMEM);
> >> +    }
> >> +    sc->stts_count = item_count;
> >> +    sc->stts_data = av_malloc_array(item_count, sizeof(*sc->stts_data));
> >> +    if (!sc->stts_data) {
> >> +        return AVERROR(ENOMEM);
> >> +    }
> >> +    sc->stts_data[0].count = 1;
> >> +    sc->stts_data[0].duration = 0;  // Not used for still images. But needed by mov_build_index.
> >> +
> >> +    for (i = 0; i < item_count; ++i) {
> >> +        (version < 2) ? avio_rb16(pb) : avio_rb32(pb);  // item_id;
> >> +        if (version > 0) {
> >> +            avio_rb16(pb);  // construction_method.
> >> +        }
> >> +        avio_rb16(pb);  // data_reference_index.
> >> +        if (rb_size(pb, &base_offset, base_offset_size) < 0) {
> >> +            return AVERROR_INVALIDDATA;
> >> +        }
> >> +        extent_count = avio_rb16(pb);
> >> +        if (extent_count > 1) {
> >> +            // For still AVIF images, we only support one extent item.
> >> +            return AVERROR_PATCHWELCOME;
> >> +        }
> >> +        for (j = 0; j < extent_count; ++j) {
> >> +            if (rb_size(pb, &extent_offset, offset_size) < 0 ||
> >> +                rb_size(pb, &extent_length, length_size) < 0) {
> >> +                return AVERROR_INVALIDDATA;
> >> +            }
> >> +            sc->sample_sizes[0] = extent_length;
> >> +            sc->chunk_offsets[0] = base_offset + extent_offset;
> >> +        }
> >> +    }
> >> +
> >> +    mov_build_index(c, st);
> >> +
> >> +    // For still AVIF images, the iloc box contains all the necessary
> >> +    // information that would generally be provided by the moov box. So simply
> >> +    // mark that we have found the moov box so that parsing can continue.
> >> +    c->found_moov = 1;
> >> +
> >> +    return atom.size;
> >> +}
> >> +
> >>  static const MOVParseTableEntry mov_default_parse_table[] = {
> >>  { MKTAG('A','C','L','R'), mov_read_aclr },
> >>  { MKTAG('A','P','R','G'), mov_read_avid },
> >> @@ -7272,6 +7413,7 @@ static const MOVParseTableEntry mov_default_parse_table[] = {
> >>  { MKTAG('d','v','v','C'), mov_read_dvcc_dvvc },
> >>  { MKTAG('d','v','w','C'), mov_read_dvcc_dvvc },
> >>  { MKTAG('k','i','n','d'), mov_read_kind },
> >> +{ MKTAG('i','l','o','c'), mov_read_iloc },
> >>  { 0, NULL }
> >>  };
> >>
> >> --
> >> 2.35.1.616.g0bdcbb4464-goog
> >>
> >
> > Ping on this patch please.
> >
>
> Fix the coding style, please.

I am sorry, can you please be a bit more specific? I was trying to
follow the code style that is consistent with the rest of the file.

> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".



-- 
Vignesh
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".