On Tue, Apr 12, 2011 at 03:35:04PM +0200, Max Horn wrote: > Hi Kostya, > > thanks for your feedback. Attached is a revised patch, in which I hopefully > addressed all issues you pointed out. If I forgot something / introduced new > badness, please let me know. > > Cheers, > Max
> From 861668cde2dc632c405a97f00786fee79b4bf0bc Mon Sep 17 00:00:00 2001 > From: Max Horn <[email protected]> > Date: Tue, 12 Apr 2011 13:59:39 +0200 > Subject: [PATCH] add xWMA demuxer > > --- [...] > diff --git a/libavformat/xwma.c b/libavformat/xwma.c > new file mode 100644 > index 0000000..00cd711 > --- /dev/null > +++ b/libavformat/xwma.c > @@ -0,0 +1,253 @@ > +/* > + * xWMA demuxer > + * Copyright (c) 2011 Max Horn > + * > + * This file is part of Libav. > + * > + * Libav is free software; you can redistribute it and/or > + * modify it under the terms of the GNU Lesser General Public > + * License as published by the Free Software Foundation; either > + * version 2.1 of the License, or (at your option) any later version. > + * > + * Libav is distributed in the hope that it will be useful, > + * 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 > + * License along with Libav; if not, write to the Free Software > + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 > USA > + */ > + > +#include "avformat.h" > +#include "riff.h" > + > +/* > + * Demuxer for xWMA, a Microsoft audio container used by XAudio 2. > + */ > + > +typedef struct { > + int64_t data_end; > +} XWMAContext; > + > +static int xwma_probe(AVProbeData *p) > +{ > + if (!memcmp(p->buf, "RIFF", 4) && !memcmp(p->buf + 8, "XWMA", 4)) > + return AVPROBE_SCORE_MAX; > + return 0; > +} > + > +static int xwma_read_header(AVFormatContext *s, AVFormatParameters *ap) > +{ > + int64_t size, av_uninit(data_size); > + uint32_t dpds_table_size = 0; > + uint32_t *dpds_table = 0; > + unsigned int tag; > + AVIOContext *pb = s->pb; > + AVStream *st; > + XWMAContext *xwma = s->priv_data; > + int i; > + > + /* The following code is mostly copied from wav.c, with some > + * minor alterations. > + */ > + > + /* check RIFF header */ > + tag = avio_rl32(pb); > + if (tag != MKTAG('R', 'I', 'F', 'F')) > + return -1; > + avio_rl32(pb); /* file size */ > + tag = avio_rl32(pb); > + if (tag != MKTAG('X', 'W', 'M', 'A')) > + return -1; > + > + /* parse fmt header */ > + tag = avio_rl32(pb); > + if (tag != MKTAG('f', 'm', 't', ' ')) > + return -1; > + size = avio_rl32(pb); > + if (size < 0) > + return -1; not possible since avio_rl32() will return unsigned value (and it will fit into int64_t) > + st = av_new_stream(s, 0); > + if (!st) > + return AVERROR(ENOMEM); > + > + ff_get_wav_header(pb, st->codec, size); > + st->need_parsing = AVSTREAM_PARSE_NONE; > + > + /* All xWMA files I have seen contained WMAv2 data. If there are files > + * using WMA Pro or some other codec, then we need to figure out the > right > + * extradata for that. Thus, ask the user for feedback, but try to go on > + * anyway. > + */ > + if (st->codec->codec_id != CODEC_ID_WMAV2) { > + av_log(s, AV_LOG_WARNING, "unexpected codec (tag 0x04%x; id %d)\n", > + st->codec->codec_tag, st->codec->codec_id); > + av_log_ask_for_sample(s, NULL); > + } else { > + /* In all xWMA files I have seen, there is no extradata. But the WMA > + * codecs require extradata, so we provide our own fake extradata. > + * > + * First, check that there really was no extradata in the header. If > + * there was, then try to use, after asking the the user to provide a > + * sample of this unusual file. > + */ > + if (st->codec->extradata_size != 0) { > + /* Surprise, surprise: We *did* get some extradata. No idea > + * if it will work, but just go on and try it, after asking > + * the user for a sample. > + */ > + av_log(s, AV_LOG_WARNING, "unexpected extradata (%d bytes)\n", > + st->codec->extradata_size); > + av_log_ask_for_sample(s, NULL); > + } else { > + st->codec->extradata_size = 6; > + st->codec->extradata = av_mallocz(6 + > FF_INPUT_BUFFER_PADDING_SIZE); > + if (!st->codec->extradata) > + return AVERROR(ENOMEM); > + > + /* setup extradata with our experimentally obtained value */ > + st->codec->extradata[4] = 31; > + } > + } > + > + /* set the sample rate */ > + av_set_pts_info(st, 64, 1, st->codec->sample_rate); > + > + /* parse the remaining RIFF chunks */ > + for (;;) { > + if (pb->eof_reached) > + return -1; > + /* read next chunk tag */ > + tag = avio_rl32(pb); > + size = avio_rl32(pb); > + if (tag == MKTAG('d', 'a', 't', 'a')) { > + /* We assume that the data chunk comes last. */ > + break; > + } else if (tag == MKTAG('d','p','d','s')) { > + /* Quoting the MSDN xWMA docs on the dpds chunk: "Contains the > + * decoded packet cumulative data size array, each element is the > + * number of bytes accumulated after the corresponding xWMA > packet > + * is decoded in order" > + * > + * Each packet has size equal to st->codec->block_align, which in > + * all cases I saw so far was always 2230. Thus, we can use the > + * dpds data to compute a seeking index. > + */ > + > + /* Error out if there is more than one dpds chunk. */ > + if (dpds_table) { > + av_log(s, AV_LOG_ERROR, "two dpds chunks present\n"); > + return -1; > + } > + > + /* Compute the number of entries in the dpds chunk. */ > + if (size % 4 != 0) { /* Size should be divisible by four */ "if (size & 3)" looks better to me > + av_log(s, AV_LOG_WARNING, "dpds chunk size "PRId64" not > divisible by 4\n", size); > + } > + dpds_table_size = size / 4; someone would complain about possible security hole if sizeof(uint32_t) > 4, but maybe a general check for insanely large dpds_table_size values would be good (like >= INT_MAX / 4 and also == 0). > + /* Allocate some temporary storage to keep the dpds data around. > + * for processing later on. > + */ > + dpds_table = av_malloc(dpds_table_size * sizeof(uint32_t)); > + if (!dpds_table) { > + return AVERROR(ENOMEM); > + } > + > + for (i = 0; i < dpds_table_size; ++i) { > + dpds_table[i] = avio_rl32(pb); > + size -= 4; > + } > + } > + avio_skip(pb, size); > + } > + > + /* Determine overall data length */ > + if (size < 0) > + return -1; > + if (!size) { > + xwma->data_end = INT64_MAX; > + } else > + xwma->data_end = avio_tell(pb) + size; > + > + > + if (dpds_table && dpds_table_size) { > + int64_t cur_pos; > + const uint32_t bytes_per_sample > + = (st->codec->channels * st->codec->bits_per_coded_sample) > >> 3; > + > + /* Estimate the duration from the total number of output bytes. */ > + const uint64_t total_decoded_bytes = dpds_table[dpds_table_size - 1]; > + st->duration = total_decoded_bytes / bytes_per_sample; why not st->duration = dpds_table_size > + /* Use the dpds data to build a seek table. We can only do this > after > + * we know the offset to the data chunk, as we need that to determine > + * the actual offset to each input block. > + * Note: If we allowed ourselves to assume that the data chunk always > + * follows immediately after the dpds block, we could of course guess > + * the data block's start offset already while reading the dpds > chunk. > + * I decided against that, just in case other chunks ever are > + * discovered. > + */ > + cur_pos = avio_tell(pb); > + for (i = 0; i < dpds_table_size; ++i) { > + /* From the number of output bytes that would accumulate in the > + * output buffer after decoding the first (i+1) packets, we > compute > + * an offset / timestamp pair. > + */ > + av_add_index_entry(st, > + cur_pos + (i+1) * st->codec->block_align, /* > pos */ > + dpds_table[i] / bytes_per_sample, /* > timestamp */ > + st->codec->block_align, /* > size */ > + 0, /* > duration */ > + AVINDEX_KEYFRAME); > + } > + } else if (st->codec->bit_rate) { > + /* No dpds chunk was present (or only an empty one), so estimate > + * the total duration using the average bits per sample and the > + * total data length. > + */ > + st->duration = (size<<3) * st->codec->sample_rate / > st->codec->bit_rate; > + } > + > + av_free(dpds_table); > + > + return 0; > +} > + > +static int xwma_read_packet(AVFormatContext *s, AVPacket *pkt) > +{ > + int ret, size; > + int64_t left; > + AVStream *st; > + XWMAContext *xwma = s->priv_data; > + > + /* Only one stream in all xWMA files I've seen so far. */ maybe it's better to drop that comment, we don't expect multiple streams in most audio files (except designed for that) > + st = s->streams[0]; > + > + left = xwma->data_end - avio_tell(s->pb); > + if (left <= 0) { > + return AVERROR_EOF; > + } > + > + /* read a single block; the default block size is 2230. */ > + size = (st->codec->block_align > 1) ? st->codec->block_align : 2230; > + size = FFMIN(size, left); > + > + ret = av_get_packet(s->pb, pkt, size); > + if (ret < 0) > + return ret; > + > + pkt->stream_index = 0; > + return ret; > +} > + > +AVInputFormat ff_xwma_demuxer = { > + "xwma", > + NULL_IF_CONFIG_SMALL("Microsoft xWMA"), > + sizeof(XWMAContext), > + xwma_probe, > + xwma_read_header, > + xwma_read_packet, > +}; Formatting and natual language nits are left to Diego. _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
