libavformat/idcin.c
4120a53a
 /*
ffa5ed24
  * id Quake II CIN File Demuxer
41ed7ab4
  * Copyright (c) 2003 The FFmpeg project
4120a53a
  *
b78e7197
  * This file is part of FFmpeg.
  *
  * FFmpeg is free software; you can redistribute it and/or
4120a53a
  * modify it under the terms of the GNU Lesser General Public
  * License as published by the Free Software Foundation; either
b78e7197
  * version 2.1 of the License, or (at your option) any later version.
4120a53a
  *
b78e7197
  * FFmpeg is distributed in the hope that it will be useful,
4120a53a
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  * Lesser General Public License for more details.
  *
  * You should have received a copy of the GNU Lesser General Public
b78e7197
  * License along with FFmpeg; if not, write to the Free Software
5509bffa
  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
4120a53a
  */
 
 /**
ba87f080
  * @file
ffa5ed24
  * id Quake II CIN file demuxer by Mike Melanson (melanson@pcisys.net)
  * For more information about the id CIN format, visit:
4120a53a
  *   http://www.csse.monash.edu.au/~timf/
  *
  * CIN is a somewhat quirky and ill-defined format. Here are some notes
  * for anyone trying to understand the technical details of this format:
  *
  * The format has no definite file signature. This is problematic for a
  * general-purpose media player that wants to automatically detect file
  * types. However, a CIN file does start with 5 32-bit numbers that
  * specify audio and video parameters. This demuxer gets around the lack
  * of file signature by performing sanity checks on those parameters.
41ed7ab4
  * Probabilistically, this is a reasonable solution since the number of
4120a53a
  * valid combinations of the 5 parameters is a very small subset of the
  * total 160-bit number space.
  *
  * Refer to the function idcin_probe() for the precise A/V parameters
  * that this demuxer allows.
  *
  * Next, each audio and video frame has a duration of 1/14 sec. If the
  * audio sample rate is a multiple of the common frequency 22050 Hz it will
  * divide evenly by 14. However, if the sample rate is 11025 Hz:
  *   11025 (samples/sec) / 14 (frames/sec) = 787.5 (samples/frame)
  * The way the CIN stores audio in this case is by storing 787 sample
  * frames in the first audio frame and 788 sample frames in the second
  * audio frame. Therefore, the total number of bytes in an audio frame
  * is given as:
  *   audio frame #0: 787 * (bytes/sample) * (# channels) bytes in frame
  *   audio frame #1: 788 * (bytes/sample) * (# channels) bytes in frame
  *   audio frame #2: 787 * (bytes/sample) * (# channels) bytes in frame
  *   audio frame #3: 788 * (bytes/sample) * (# channels) bytes in frame
  *
ffa5ed24
  * Finally, not all id CIN creation tools agree on the resolution of the
4120a53a
  * color palette, apparently. Some creation tools specify red, green, and
  * blue palette components in terms of 6-bit VGA color DAC values which
  * range from 0..63. Other tools specify the RGB components as full 8-bit
  * values that range from 0..255. Since there are no markers in the file to
  * differentiate between the two variants, this demuxer uses the following
  * heuristic:
  *   - load the 768 palette bytes from disk
  *   - assume that they will need to be shifted left by 2 bits to
  *     transform them from 6-bit values to 8-bit values
  *   - scan through all 768 palette bytes
  *     - if any bytes exceed 63, do not shift the bytes at all before
  *       transmitting them to the video decoder
  */
 
0651e892
 #include "libavutil/channel_layout.h"
b0c96e06
 #include "libavutil/imgutils.h"
6a5d31ac
 #include "libavutil/intreadwrite.h"
4120a53a
 #include "avformat.h"
c3f9ebf7
 #include "internal.h"
4120a53a
 
 #define HUFFMAN_TABLE_SIZE (64 * 1024)
40e8e497
 #define IDCIN_FPS 14
4120a53a
 
 typedef struct IdcinDemuxContext {
     int video_stream_index;
     int audio_stream_index;
     int audio_chunk_size1;
     int audio_chunk_size2;
ccc0ffb1
     int block_align;
4120a53a
 
     /* demux state variables */
     int current_audio_chunk;
     int next_chunk_is_video;
     int audio_present;
7040e479
     int64_t first_pkt_pos;
4120a53a
 } IdcinDemuxContext;
 
4d8875ec
 static int idcin_probe(const AVProbeData *p)
4120a53a
 {
06bede95
     unsigned int number, sample_rate;
4c439f6e
     unsigned int w, h;
     int i;
4120a53a
 
     /*
ffa5ed24
      * This is what you could call a "probabilistic" file check: id CIN
4120a53a
      * files don't have a definite file signature. In lieu of such a marker,
      * perform sanity checks on the 5 32-bit header fields:
      *  width, height: greater than 0, less than or equal to 1024
      * audio sample rate: greater than or equal to 8000, less than or
      *  equal to 48000, or 0 for no audio
      * audio sample width (bytes/sample): 0 for no audio, or 1 or 2
      * audio channels: 0 for no audio, or 1 or 2
      */
 
8466ab59
     /* check we have enough data to do all checks, otherwise the
        0-padding may cause a wrong recognition */
4c439f6e
     if (p->buf_size < 20 + HUFFMAN_TABLE_SIZE + 12)
8466ab59
         return 0;
 
4120a53a
     /* check the video width */
4c439f6e
     w = AV_RL32(&p->buf[0]);
     if ((w == 0) || (w > 1024))
4120a53a
        return 0;
 
     /* check the video height */
4c439f6e
     h = AV_RL32(&p->buf[4]);
     if ((h == 0) || (h > 1024))
4120a53a
        return 0;
 
     /* check the audio sample rate */
06bede95
     sample_rate = AV_RL32(&p->buf[8]);
     if (sample_rate && (sample_rate < 8000 || sample_rate > 48000))
4120a53a
         return 0;
 
     /* check the audio bytes/sample */
fead30d4
     number = AV_RL32(&p->buf[12]);
06bede95
     if (number > 2 || sample_rate && !number)
4120a53a
         return 0;
 
     /* check the audio channels */
fead30d4
     number = AV_RL32(&p->buf[16]);
06bede95
     if (number > 2 || sample_rate && !number)
4120a53a
         return 0;
 
4c439f6e
     i = 20 + HUFFMAN_TABLE_SIZE;
     if (AV_RL32(&p->buf[i]) == 1)
         i += 768;
 
     if (i+12 > p->buf_size || AV_RL32(&p->buf[i+8]) != w*h)
         return 1;
 
a5f88736
     /* return half certainty since this check is a bit sketchy */
e0f8be64
     return AVPROBE_SCORE_EXTENSION;
4120a53a
 }
 
6e9651d1
 static int idcin_read_header(AVFormatContext *s)
4120a53a
 {
471fe57e
     AVIOContext *pb = s->pb;
e4141433
     IdcinDemuxContext *idcin = s->priv_data;
4120a53a
     AVStream *st;
     unsigned int width, height;
     unsigned int sample_rate, bytes_per_sample, channels;
5d045046
     int ret;
4120a53a
 
     /* get the 5 header parameters */
e63a3628
     width = avio_rl32(pb);
     height = avio_rl32(pb);
     sample_rate = avio_rl32(pb);
     bytes_per_sample = avio_rl32(pb);
     channels = avio_rl32(pb);
4120a53a
 
5d045046
     if (s->pb->eof_reached) {
         av_log(s, AV_LOG_ERROR, "incomplete header\n");
         return s->pb->error ? s->pb->error : AVERROR_EOF;
     }
 
b0c96e06
     if (av_image_check_size(width, height, 0, s) < 0)
         return AVERROR_INVALIDDATA;
     if (sample_rate > 0) {
         if (sample_rate < 14 || sample_rate > INT_MAX) {
             av_log(s, AV_LOG_ERROR, "invalid sample rate: %u\n", sample_rate);
             return AVERROR_INVALIDDATA;
         }
         if (bytes_per_sample < 1 || bytes_per_sample > 2) {
             av_log(s, AV_LOG_ERROR, "invalid bytes per sample: %u\n",
                    bytes_per_sample);
             return AVERROR_INVALIDDATA;
         }
         if (channels < 1 || channels > 2) {
             av_log(s, AV_LOG_ERROR, "invalid channels: %u\n", channels);
             return AVERROR_INVALIDDATA;
         }
12c2530b
         idcin->audio_present = 1;
     } else {
         /* if sample rate is 0, assume no audio */
         idcin->audio_present = 0;
b0c96e06
     }
 
3b3bbdd3
     st = avformat_new_stream(s, NULL);
4120a53a
     if (!st)
769e10f0
         return AVERROR(ENOMEM);
c3f9ebf7
     avpriv_set_pts_info(st, 33, 1, IDCIN_FPS);
ccc0ffb1
     st->start_time = 0;
4120a53a
     idcin->video_stream_index = st->index;
9200514a
     st->codecpar->codec_type = AVMEDIA_TYPE_VIDEO;
     st->codecpar->codec_id = AV_CODEC_ID_IDCIN;
     st->codecpar->codec_tag = 0;  /* no fourcc */
     st->codecpar->width = width;
     st->codecpar->height = height;
4120a53a
 
     /* load up the Huffman tables into extradata */
323b8c95
     if ((ret = ff_get_extradata(s, st->codecpar, pb, HUFFMAN_TABLE_SIZE)) < 0)
5d045046
         return ret;
4120a53a
 
12c2530b
     if (idcin->audio_present) {
4120a53a
         idcin->audio_present = 1;
3b3bbdd3
         st = avformat_new_stream(s, NULL);
4120a53a
         if (!st)
769e10f0
             return AVERROR(ENOMEM);
ccc0ffb1
         avpriv_set_pts_info(st, 63, 1, sample_rate);
         st->start_time = 0;
4120a53a
         idcin->audio_stream_index = st->index;
9200514a
         st->codecpar->codec_type = AVMEDIA_TYPE_AUDIO;
         st->codecpar->codec_tag = 1;
         st->codecpar->channels = channels;
         st->codecpar->channel_layout = channels > 1 ? AV_CH_LAYOUT_STEREO :
                                                       AV_CH_LAYOUT_MONO;
         st->codecpar->sample_rate = sample_rate;
         st->codecpar->bits_per_coded_sample = bytes_per_sample * 8;
         st->codecpar->bit_rate = sample_rate * bytes_per_sample * 8 * channels;
         st->codecpar->block_align = idcin->block_align = bytes_per_sample * channels;
4120a53a
         if (bytes_per_sample == 1)
9200514a
             st->codecpar->codec_id = AV_CODEC_ID_PCM_U8;
4120a53a
         else
9200514a
             st->codecpar->codec_id = AV_CODEC_ID_PCM_S16LE;
4120a53a
 
         if (sample_rate % 14 != 0) {
             idcin->audio_chunk_size1 = (sample_rate / 14) *
             bytes_per_sample * channels;
             idcin->audio_chunk_size2 = (sample_rate / 14 + 1) *
                 bytes_per_sample * channels;
         } else {
             idcin->audio_chunk_size1 = idcin->audio_chunk_size2 =
                 (sample_rate / 14) * bytes_per_sample * channels;
         }
         idcin->current_audio_chunk = 0;
12c2530b
     }
4120a53a
 
     idcin->next_chunk_is_video = 1;
7040e479
     idcin->first_pkt_pos = avio_tell(s->pb);
4120a53a
 
     return 0;
 }
 
 static int idcin_read_packet(AVFormatContext *s,
                              AVPacket *pkt)
 {
     int ret;
     unsigned int command;
     unsigned int chunk_size;
e4141433
     IdcinDemuxContext *idcin = s->priv_data;
471fe57e
     AVIOContext *pb = s->pb;
4120a53a
     int i;
     int palette_scale;
2a2bbcb0
     unsigned char r, g, b;
     unsigned char palette_buffer[768];
2d8591c2
     uint32_t palette[256];
4120a53a
 
d34ec64a
     if (avio_feof(s->pb))
5d045046
         return s->pb->error ? s->pb->error : AVERROR_EOF;
4120a53a
 
     if (idcin->next_chunk_is_video) {
e63a3628
         command = avio_rl32(pb);
4120a53a
         if (command == 2) {
6f3e0b21
             return AVERROR(EIO);
4120a53a
         } else if (command == 1) {
             /* trigger a palette change */
5d045046
             ret = avio_read(pb, palette_buffer, 768);
             if (ret < 0) {
                 return ret;
             } else if (ret != 768) {
                 av_log(s, AV_LOG_ERROR, "incomplete packet\n");
6f3e0b21
                 return AVERROR(EIO);
5d045046
             }
4120a53a
             /* scale the palette as necessary */
             palette_scale = 2;
             for (i = 0; i < 768; i++)
2a2bbcb0
                 if (palette_buffer[i] > 63) {
4120a53a
                     palette_scale = 0;
                     break;
                 }
 
2a2bbcb0
             for (i = 0; i < 256; i++) {
                 r = palette_buffer[i * 3    ] << palette_scale;
                 g = palette_buffer[i * 3 + 1] << palette_scale;
                 b = palette_buffer[i * 3 + 2] << palette_scale;
32c49389
                 palette[i] = (0xFFU << 24) | (r << 16) | (g << 8) | (b);
                 if (palette_scale == 2)
                     palette[i] |= palette[i] >> 6 & 0x30303;
2a2bbcb0
             }
4120a53a
         }
 
5d045046
         if (s->pb->eof_reached) {
             av_log(s, AV_LOG_ERROR, "incomplete packet\n");
             return s->pb->error ? s->pb->error : AVERROR_EOF;
         }
e63a3628
         chunk_size = avio_rl32(pb);
33f58c36
         if (chunk_size < 4 || chunk_size > INT_MAX - 4) {
             av_log(s, AV_LOG_ERROR, "invalid chunk size: %u\n", chunk_size);
             return AVERROR_INVALIDDATA;
         }
4120a53a
         /* skip the number of decoded bytes (always equal to width * height) */
45a8a02a
         avio_skip(pb, 4);
4120a53a
         chunk_size -= 4;
115329f1
         ret= av_get_packet(pb, pkt, chunk_size);
044a950d
         if (ret < 0)
             return ret;
5d045046
         else if (ret != chunk_size) {
             av_log(s, AV_LOG_ERROR, "incomplete packet\n");
             return AVERROR(EIO);
         }
2d8591c2
         if (command == 1) {
             uint8_t *pal;
 
             pal = av_packet_new_side_data(pkt, AV_PKT_DATA_PALETTE,
                                           AVPALETTE_SIZE);
2dc9bcad
             if (!pal) {
ee884c30
                 return AVERROR(ENOMEM);
b805c725
             }
2d8591c2
             memcpy(pal, palette, AVPALETTE_SIZE);
49543373
             pkt->flags |= AV_PKT_FLAG_KEY;
2d8591c2
         }
4120a53a
         pkt->stream_index = idcin->video_stream_index;
ccc0ffb1
         pkt->duration     = 1;
4120a53a
     } else {
         /* send out the audio chunk */
         if (idcin->current_audio_chunk)
             chunk_size = idcin->audio_chunk_size2;
         else
             chunk_size = idcin->audio_chunk_size1;
2692067a
         ret= av_get_packet(pb, pkt, chunk_size);
044a950d
         if (ret < 0)
             return ret;
4120a53a
         pkt->stream_index = idcin->audio_stream_index;
ccc0ffb1
         pkt->duration     = chunk_size / idcin->block_align;
4120a53a
 
         idcin->current_audio_chunk ^= 1;
     }
 
     if (idcin->audio_present)
         idcin->next_chunk_is_video ^= 1;
 
06deaf8a
     return 0;
4120a53a
 }
 
7040e479
 static int idcin_read_seek(AVFormatContext *s, int stream_index,
                            int64_t timestamp, int flags)
 {
     IdcinDemuxContext *idcin = s->priv_data;
 
     if (idcin->first_pkt_pos > 0) {
d1923d15
         int64_t ret = avio_seek(s->pb, idcin->first_pkt_pos, SEEK_SET);
7040e479
         if (ret < 0)
             return ret;
         ff_update_cur_dts(s, s->streams[idcin->video_stream_index], 0);
         idcin->next_chunk_is_video = 1;
         idcin->current_audio_chunk = 0;
         return 0;
     }
     return -1;
 }
 
66355be3
 AVInputFormat ff_idcin_demuxer = {
dfc2c4d9
     .name           = "idcin",
6774247a
     .long_name      = NULL_IF_CONFIG_SMALL("id Cinematic"),
dfc2c4d9
     .priv_data_size = sizeof(IdcinDemuxContext),
     .read_probe     = idcin_probe,
     .read_header    = idcin_read_header,
     .read_packet    = idcin_read_packet,
7040e479
     .read_seek      = idcin_read_seek,
     .flags          = AVFMT_NO_BYTE_SEEK,
4120a53a
 };