Libav
|
00001 /* 00002 * Core Audio Format demuxer 00003 * Copyright (c) 2007 Justin Ruggles 00004 * Copyright (c) 2009 Peter Ross 00005 * 00006 * This file is part of FFmpeg. 00007 * 00008 * FFmpeg is free software; you can redistribute it and/or 00009 * modify it under the terms of the GNU Lesser General Public 00010 * License as published by the Free Software Foundation; either 00011 * version 2.1 of the License, or (at your option) any later version. 00012 * 00013 * FFmpeg is distributed in the hope that it will be useful, 00014 * but WITHOUT ANY WARRANTY; without even the implied warranty of 00015 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU 00016 * Lesser General Public License for more details. 00017 * 00018 * You should have received a copy of the GNU Lesser General Public 00019 * License along with FFmpeg; if not, write to the Free Software 00020 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA 00021 */ 00022 00028 #include "avformat.h" 00029 #include "riff.h" 00030 #include "isom.h" 00031 #include "libavutil/intreadwrite.h" 00032 #include "caf.h" 00033 00034 typedef struct { 00035 int bytes_per_packet; 00036 int frames_per_packet; 00037 int64_t num_bytes; 00038 00039 int64_t packet_cnt; 00040 int64_t frame_cnt; 00041 00042 int64_t data_start; 00043 int64_t data_size; 00044 } CaffContext; 00045 00046 static int probe(AVProbeData *p) 00047 { 00048 if (AV_RB32(p->buf) == MKBETAG('c','a','f','f') && AV_RB16(&p->buf[4]) == 1) 00049 return AVPROBE_SCORE_MAX; 00050 return 0; 00051 } 00052 00054 static int read_desc_chunk(AVFormatContext *s) 00055 { 00056 ByteIOContext *pb = s->pb; 00057 CaffContext *caf = s->priv_data; 00058 AVStream *st; 00059 int flags; 00060 00061 /* new audio stream */ 00062 st = av_new_stream(s, 0); 00063 if (!st) 00064 return AVERROR(ENOMEM); 00065 00066 /* parse format description */ 00067 st->codec->codec_type = AVMEDIA_TYPE_AUDIO; 00068 st->codec->sample_rate = av_int2dbl(get_be64(pb)); 00069 st->codec->codec_tag = get_be32(pb); 00070 flags = get_be32(pb); 00071 caf->bytes_per_packet = get_be32(pb); 00072 st->codec->block_align = caf->bytes_per_packet; 00073 caf->frames_per_packet = get_be32(pb); 00074 st->codec->channels = get_be32(pb); 00075 st->codec->bits_per_coded_sample = get_be32(pb); 00076 00077 /* calculate bit rate for constant size packets */ 00078 if (caf->frames_per_packet > 0 && caf->bytes_per_packet > 0) { 00079 st->codec->bit_rate = (uint64_t)st->codec->sample_rate * (uint64_t)caf->bytes_per_packet * 8 00080 / (uint64_t)caf->frames_per_packet; 00081 } else { 00082 st->codec->bit_rate = 0; 00083 } 00084 00085 /* determine codec */ 00086 if (st->codec->codec_tag == MKBETAG('l','p','c','m')) 00087 st->codec->codec_id = ff_mov_get_lpcm_codec_id(st->codec->bits_per_coded_sample, (flags ^ 0x2) | 0x4); 00088 else 00089 st->codec->codec_id = ff_codec_get_id(ff_codec_caf_tags, st->codec->codec_tag); 00090 return 0; 00091 } 00092 00094 static int read_kuki_chunk(AVFormatContext *s, int64_t size) 00095 { 00096 ByteIOContext *pb = s->pb; 00097 AVStream *st = s->streams[0]; 00098 00099 if (size < 0 || size > INT_MAX - FF_INPUT_BUFFER_PADDING_SIZE) 00100 return -1; 00101 00102 if (st->codec->codec_id == CODEC_ID_AAC) { 00103 /* The magic cookie format for AAC is an mp4 esds atom. 00104 The lavc AAC decoder requires the data from the codec specific 00105 description as extradata input. */ 00106 int strt, skip; 00107 MOVAtom atom; 00108 00109 strt = url_ftell(pb); 00110 ff_mov_read_esds(s, pb, atom); 00111 skip = size - (url_ftell(pb) - strt); 00112 if (skip < 0 || !st->codec->extradata || 00113 st->codec->codec_id != CODEC_ID_AAC) { 00114 av_log(s, AV_LOG_ERROR, "invalid AAC magic cookie\n"); 00115 return AVERROR_INVALIDDATA; 00116 } 00117 url_fskip(pb, skip); 00118 } else { 00119 st->codec->extradata = av_mallocz(size + FF_INPUT_BUFFER_PADDING_SIZE); 00120 if (!st->codec->extradata) 00121 return AVERROR(ENOMEM); 00122 get_buffer(pb, st->codec->extradata, size); 00123 st->codec->extradata_size = size; 00124 } 00125 00126 return 0; 00127 } 00128 00130 static int read_pakt_chunk(AVFormatContext *s, int64_t size) 00131 { 00132 ByteIOContext *pb = s->pb; 00133 AVStream *st = s->streams[0]; 00134 CaffContext *caf = s->priv_data; 00135 int64_t pos = 0, ccount; 00136 int num_packets, i; 00137 00138 ccount = url_ftell(pb); 00139 00140 num_packets = get_be64(pb); 00141 if (num_packets < 0 || INT32_MAX / sizeof(AVIndexEntry) < num_packets) 00142 return AVERROR_INVALIDDATA; 00143 00144 st->nb_frames = get_be64(pb); /* valid frames */ 00145 st->nb_frames += get_be32(pb); /* priming frames */ 00146 st->nb_frames += get_be32(pb); /* remainder frames */ 00147 00148 st->duration = 0; 00149 for (i = 0; i < num_packets; i++) { 00150 av_add_index_entry(s->streams[0], pos, st->duration, 0, 0, AVINDEX_KEYFRAME); 00151 pos += caf->bytes_per_packet ? caf->bytes_per_packet : ff_mp4_read_descr_len(pb); 00152 st->duration += caf->frames_per_packet ? caf->frames_per_packet : ff_mp4_read_descr_len(pb); 00153 } 00154 00155 if (url_ftell(pb) - ccount != size) { 00156 av_log(s, AV_LOG_ERROR, "error reading packet table\n"); 00157 return -1; 00158 } 00159 00160 caf->num_bytes = pos; 00161 return 0; 00162 } 00163 00165 static void read_info_chunk(AVFormatContext *s, int64_t size) 00166 { 00167 ByteIOContext *pb = s->pb; 00168 unsigned int i; 00169 unsigned int nb_entries = get_be32(pb); 00170 for (i = 0; i < nb_entries; i++) { 00171 char key[32]; 00172 char value[1024]; 00173 get_strz(pb, key, sizeof(key)); 00174 get_strz(pb, value, sizeof(value)); 00175 av_metadata_set2(&s->metadata, key, value, 0); 00176 } 00177 } 00178 00179 static int read_header(AVFormatContext *s, 00180 AVFormatParameters *ap) 00181 { 00182 ByteIOContext *pb = s->pb; 00183 CaffContext *caf = s->priv_data; 00184 AVStream *st; 00185 uint32_t tag = 0; 00186 int found_data, ret; 00187 int64_t size; 00188 00189 url_fskip(pb, 8); /* magic, version, file flags */ 00190 00191 /* audio description chunk */ 00192 if (get_be32(pb) != MKBETAG('d','e','s','c')) { 00193 av_log(s, AV_LOG_ERROR, "desc chunk not present\n"); 00194 return AVERROR_INVALIDDATA; 00195 } 00196 size = get_be64(pb); 00197 if (size != 32) 00198 return AVERROR_INVALIDDATA; 00199 00200 ret = read_desc_chunk(s); 00201 if (ret) 00202 return ret; 00203 st = s->streams[0]; 00204 00205 /* parse each chunk */ 00206 found_data = 0; 00207 while (!url_feof(pb)) { 00208 00209 /* stop at data chunk if seeking is not supported or 00210 data chunk size is unknown */ 00211 if (found_data && (caf->data_size < 0 || url_is_streamed(pb))) 00212 break; 00213 00214 tag = get_be32(pb); 00215 size = get_be64(pb); 00216 if (url_feof(pb)) 00217 break; 00218 00219 switch (tag) { 00220 case MKBETAG('d','a','t','a'): 00221 url_fskip(pb, 4); /* edit count */ 00222 caf->data_start = url_ftell(pb); 00223 caf->data_size = size < 0 ? -1 : size - 4; 00224 if (caf->data_size > 0 && !url_is_streamed(pb)) 00225 url_fskip(pb, caf->data_size); 00226 found_data = 1; 00227 break; 00228 00229 /* magic cookie chunk */ 00230 case MKBETAG('k','u','k','i'): 00231 if (read_kuki_chunk(s, size)) 00232 return AVERROR_INVALIDDATA; 00233 break; 00234 00235 /* packet table chunk */ 00236 case MKBETAG('p','a','k','t'): 00237 if (read_pakt_chunk(s, size)) 00238 return AVERROR_INVALIDDATA; 00239 break; 00240 00241 case MKBETAG('i','n','f','o'): 00242 read_info_chunk(s, size); 00243 break; 00244 00245 default: 00246 #define _(x) ((x) >= ' ' ? (x) : ' ') 00247 av_log(s, AV_LOG_WARNING, "skipping CAF chunk: %08X (%c%c%c%c)\n", 00248 tag, _(tag>>24), _((tag>>16)&0xFF), _((tag>>8)&0xFF), _(tag&0xFF)); 00249 #undef _ 00250 case MKBETAG('f','r','e','e'): 00251 if (size < 0) 00252 return AVERROR_INVALIDDATA; 00253 url_fskip(pb, size); 00254 break; 00255 } 00256 } 00257 00258 if (!found_data) 00259 return AVERROR_INVALIDDATA; 00260 00261 if (caf->bytes_per_packet > 0 && caf->frames_per_packet > 0) { 00262 if (caf->data_size > 0) 00263 st->nb_frames = (caf->data_size / caf->bytes_per_packet) * caf->frames_per_packet; 00264 } else if (st->nb_index_entries) { 00265 st->codec->bit_rate = st->codec->sample_rate * caf->data_size * 8 / 00266 st->duration; 00267 } else { 00268 av_log(s, AV_LOG_ERROR, "Missing packet table. It is required when " 00269 "block size or frame size are variable.\n"); 00270 return AVERROR_INVALIDDATA; 00271 } 00272 s->file_size = url_fsize(pb); 00273 s->file_size = FFMAX(0, s->file_size); 00274 00275 av_set_pts_info(st, 64, 1, st->codec->sample_rate); 00276 st->start_time = 0; 00277 00278 /* position the stream at the start of data */ 00279 if (caf->data_size >= 0) 00280 url_fseek(pb, caf->data_start, SEEK_SET); 00281 00282 return 0; 00283 } 00284 00285 #define CAF_MAX_PKT_SIZE 4096 00286 00287 static int read_packet(AVFormatContext *s, AVPacket *pkt) 00288 { 00289 ByteIOContext *pb = s->pb; 00290 AVStream *st = s->streams[0]; 00291 CaffContext *caf = s->priv_data; 00292 int res, pkt_size = 0, pkt_frames = 0; 00293 int64_t left = CAF_MAX_PKT_SIZE; 00294 00295 if (url_feof(pb)) 00296 return AVERROR(EIO); 00297 00298 /* don't read past end of data chunk */ 00299 if (caf->data_size > 0) { 00300 left = (caf->data_start + caf->data_size) - url_ftell(pb); 00301 if (left <= 0) 00302 return AVERROR(EIO); 00303 } 00304 00305 pkt_frames = caf->frames_per_packet; 00306 pkt_size = caf->bytes_per_packet; 00307 00308 if (pkt_size > 0 && pkt_frames == 1) { 00309 pkt_size = (CAF_MAX_PKT_SIZE / pkt_size) * pkt_size; 00310 pkt_size = FFMIN(pkt_size, left); 00311 pkt_frames = pkt_size / caf->bytes_per_packet; 00312 } else if (st->nb_index_entries) { 00313 if (caf->packet_cnt < st->nb_index_entries - 1) { 00314 pkt_size = st->index_entries[caf->packet_cnt + 1].pos - st->index_entries[caf->packet_cnt].pos; 00315 pkt_frames = st->index_entries[caf->packet_cnt + 1].timestamp - st->index_entries[caf->packet_cnt].timestamp; 00316 } else if (caf->packet_cnt == st->nb_index_entries - 1) { 00317 pkt_size = caf->num_bytes - st->index_entries[caf->packet_cnt].pos; 00318 pkt_frames = st->duration - st->index_entries[caf->packet_cnt].timestamp; 00319 } else { 00320 return AVERROR(EIO); 00321 } 00322 } 00323 00324 if (pkt_size == 0 || pkt_frames == 0 || pkt_size > left) 00325 return AVERROR(EIO); 00326 00327 res = av_get_packet(pb, pkt, pkt_size); 00328 if (res < 0) 00329 return res; 00330 00331 pkt->size = res; 00332 pkt->stream_index = 0; 00333 pkt->dts = pkt->pts = caf->frame_cnt; 00334 00335 caf->packet_cnt++; 00336 caf->frame_cnt += pkt_frames; 00337 00338 return 0; 00339 } 00340 00341 static int read_seek(AVFormatContext *s, int stream_index, 00342 int64_t timestamp, int flags) 00343 { 00344 AVStream *st = s->streams[0]; 00345 CaffContext *caf = s->priv_data; 00346 int64_t pos; 00347 00348 timestamp = FFMAX(timestamp, 0); 00349 00350 if (caf->frames_per_packet > 0 && caf->bytes_per_packet > 0) { 00351 /* calculate new byte position based on target frame position */ 00352 pos = caf->bytes_per_packet * timestamp / caf->frames_per_packet; 00353 if (caf->data_size > 0) 00354 pos = FFMIN(pos, caf->data_size); 00355 caf->packet_cnt = pos / caf->bytes_per_packet; 00356 caf->frame_cnt = caf->frames_per_packet * caf->packet_cnt; 00357 } else if (st->nb_index_entries) { 00358 caf->packet_cnt = av_index_search_timestamp(st, timestamp, flags); 00359 caf->frame_cnt = st->index_entries[caf->packet_cnt].timestamp; 00360 pos = st->index_entries[caf->packet_cnt].pos; 00361 } else { 00362 return -1; 00363 } 00364 00365 url_fseek(s->pb, pos + caf->data_start, SEEK_SET); 00366 return 0; 00367 } 00368 00369 AVInputFormat caf_demuxer = { 00370 "caf", 00371 NULL_IF_CONFIG_SMALL("Apple Core Audio Format"), 00372 sizeof(CaffContext), 00373 probe, 00374 read_header, 00375 read_packet, 00376 NULL, 00377 read_seek, 00378 .codec_tag = (const AVCodecTag*[]){ff_codec_caf_tags, 0}, 00379 };