Subversion Repositories Kolibri OS

Rev

Blame | Last modification | View Log | RSS feed

  1. /*
  2.  * Copyright (c) 2012 Stefano Sabatini
  3.  *
  4.  * Permission is hereby granted, free of charge, to any person obtaining a copy
  5.  * of this software and associated documentation files (the "Software"), to deal
  6.  * in the Software without restriction, including without limitation the rights
  7.  * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
  8.  * copies of the Software, and to permit persons to whom the Software is
  9.  * furnished to do so, subject to the following conditions:
  10.  *
  11.  * The above copyright notice and this permission notice shall be included in
  12.  * all copies or substantial portions of the Software.
  13.  *
  14.  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  15.  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  16.  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
  17.  * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  18.  * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
  19.  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
  20.  * THE SOFTWARE.
  21.  */
  22.  
  23. /**
  24.  * @file
  25.  * Demuxing and decoding example.
  26.  *
  27.  * Show how to use the libavformat and libavcodec API to demux and
  28.  * decode audio and video data.
  29.  * @example demuxing_decoding.c
  30.  */
  31.  
  32. #include <libavutil/imgutils.h>
  33. #include <libavutil/samplefmt.h>
  34. #include <libavutil/timestamp.h>
  35. #include <libavformat/avformat.h>
  36.  
  37. static AVFormatContext *fmt_ctx = NULL;
  38. static AVCodecContext *video_dec_ctx = NULL, *audio_dec_ctx;
  39. static int width, height;
  40. static enum AVPixelFormat pix_fmt;
  41. static AVStream *video_stream = NULL, *audio_stream = NULL;
  42. static const char *src_filename = NULL;
  43. static const char *video_dst_filename = NULL;
  44. static const char *audio_dst_filename = NULL;
  45. static FILE *video_dst_file = NULL;
  46. static FILE *audio_dst_file = NULL;
  47.  
  48. static uint8_t *video_dst_data[4] = {NULL};
  49. static int      video_dst_linesize[4];
  50. static int video_dst_bufsize;
  51.  
  52. static int video_stream_idx = -1, audio_stream_idx = -1;
  53. static AVFrame *frame = NULL;
  54. static AVPacket pkt;
  55. static int video_frame_count = 0;
  56. static int audio_frame_count = 0;
  57.  
  58. /* The different ways of decoding and managing data memory. You are not
  59.  * supposed to support all the modes in your application but pick the one most
  60.  * appropriate to your needs. Look for the use of api_mode in this example to
  61.  * see what are the differences of API usage between them */
  62. enum {
  63.     API_MODE_OLD                  = 0, /* old method, deprecated */
  64.     API_MODE_NEW_API_REF_COUNT    = 1, /* new method, using the frame reference counting */
  65.     API_MODE_NEW_API_NO_REF_COUNT = 2, /* new method, without reference counting */
  66. };
  67.  
  68. static int api_mode = API_MODE_OLD;
  69.  
  70. static int decode_packet(int *got_frame, int cached)
  71. {
  72.     int ret = 0;
  73.     int decoded = pkt.size;
  74.  
  75.     *got_frame = 0;
  76.  
  77.     if (pkt.stream_index == video_stream_idx) {
  78.         /* decode video frame */
  79.         ret = avcodec_decode_video2(video_dec_ctx, frame, got_frame, &pkt);
  80.         if (ret < 0) {
  81.             fprintf(stderr, "Error decoding video frame (%s)\n", av_err2str(ret));
  82.             return ret;
  83.         }
  84.  
  85.         if (*got_frame) {
  86.  
  87.             if (frame->width != width || frame->height != height ||
  88.                 frame->format != pix_fmt) {
  89.                 /* To handle this change, one could call av_image_alloc again and
  90.                  * decode the following frames into another rawvideo file. */
  91.                 fprintf(stderr, "Error: Width, height and pixel format have to be "
  92.                         "constant in a rawvideo file, but the width, height or "
  93.                         "pixel format of the input video changed:\n"
  94.                         "old: width = %d, height = %d, format = %s\n"
  95.                         "new: width = %d, height = %d, format = %s\n",
  96.                         width, height, av_get_pix_fmt_name(pix_fmt),
  97.                         frame->width, frame->height,
  98.                         av_get_pix_fmt_name(frame->format));
  99.                 return -1;
  100.             }
  101.  
  102.             printf("video_frame%s n:%d coded_n:%d pts:%s\n",
  103.                    cached ? "(cached)" : "",
  104.                    video_frame_count++, frame->coded_picture_number,
  105.                    av_ts2timestr(frame->pts, &video_dec_ctx->time_base));
  106.  
  107.             /* copy decoded frame to destination buffer:
  108.              * this is required since rawvideo expects non aligned data */
  109.             av_image_copy(video_dst_data, video_dst_linesize,
  110.                           (const uint8_t **)(frame->data), frame->linesize,
  111.                           pix_fmt, width, height);
  112.  
  113.             /* write to rawvideo file */
  114.             fwrite(video_dst_data[0], 1, video_dst_bufsize, video_dst_file);
  115.         }
  116.     } else if (pkt.stream_index == audio_stream_idx) {
  117.         /* decode audio frame */
  118.         ret = avcodec_decode_audio4(audio_dec_ctx, frame, got_frame, &pkt);
  119.         if (ret < 0) {
  120.             fprintf(stderr, "Error decoding audio frame (%s)\n", av_err2str(ret));
  121.             return ret;
  122.         }
  123.         /* Some audio decoders decode only part of the packet, and have to be
  124.          * called again with the remainder of the packet data.
  125.          * Sample: fate-suite/lossless-audio/luckynight-partial.shn
  126.          * Also, some decoders might over-read the packet. */
  127.         decoded = FFMIN(ret, pkt.size);
  128.  
  129.         if (*got_frame) {
  130.             size_t unpadded_linesize = frame->nb_samples * av_get_bytes_per_sample(frame->format);
  131.             printf("audio_frame%s n:%d nb_samples:%d pts:%s\n",
  132.                    cached ? "(cached)" : "",
  133.                    audio_frame_count++, frame->nb_samples,
  134.                    av_ts2timestr(frame->pts, &audio_dec_ctx->time_base));
  135.  
  136.             /* Write the raw audio data samples of the first plane. This works
  137.              * fine for packed formats (e.g. AV_SAMPLE_FMT_S16). However,
  138.              * most audio decoders output planar audio, which uses a separate
  139.              * plane of audio samples for each channel (e.g. AV_SAMPLE_FMT_S16P).
  140.              * In other words, this code will write only the first audio channel
  141.              * in these cases.
  142.              * You should use libswresample or libavfilter to convert the frame
  143.              * to packed data. */
  144.             fwrite(frame->extended_data[0], 1, unpadded_linesize, audio_dst_file);
  145.         }
  146.     }
  147.  
  148.     /* If we use the new API with reference counting, we own the data and need
  149.      * to de-reference it when we don't use it anymore */
  150.     if (*got_frame && api_mode == API_MODE_NEW_API_REF_COUNT)
  151.         av_frame_unref(frame);
  152.  
  153.     return decoded;
  154. }
  155.  
  156. static int open_codec_context(int *stream_idx,
  157.                               AVFormatContext *fmt_ctx, enum AVMediaType type)
  158. {
  159.     int ret, stream_index;
  160.     AVStream *st;
  161.     AVCodecContext *dec_ctx = NULL;
  162.     AVCodec *dec = NULL;
  163.     AVDictionary *opts = NULL;
  164.  
  165.     ret = av_find_best_stream(fmt_ctx, type, -1, -1, NULL, 0);
  166.     if (ret < 0) {
  167.         fprintf(stderr, "Could not find %s stream in input file '%s'\n",
  168.                 av_get_media_type_string(type), src_filename);
  169.         return ret;
  170.     } else {
  171.         stream_index = ret;
  172.         st = fmt_ctx->streams[stream_index];
  173.  
  174.         /* find decoder for the stream */
  175.         dec_ctx = st->codec;
  176.         dec = avcodec_find_decoder(dec_ctx->codec_id);
  177.         if (!dec) {
  178.             fprintf(stderr, "Failed to find %s codec\n",
  179.                     av_get_media_type_string(type));
  180.             return AVERROR(EINVAL);
  181.         }
  182.  
  183.         /* Init the decoders, with or without reference counting */
  184.         if (api_mode == API_MODE_NEW_API_REF_COUNT)
  185.             av_dict_set(&opts, "refcounted_frames", "1", 0);
  186.         if ((ret = avcodec_open2(dec_ctx, dec, &opts)) < 0) {
  187.             fprintf(stderr, "Failed to open %s codec\n",
  188.                     av_get_media_type_string(type));
  189.             return ret;
  190.         }
  191.         *stream_idx = stream_index;
  192.     }
  193.  
  194.     return 0;
  195. }
  196.  
  197. static int get_format_from_sample_fmt(const char **fmt,
  198.                                       enum AVSampleFormat sample_fmt)
  199. {
  200.     int i;
  201.     struct sample_fmt_entry {
  202.         enum AVSampleFormat sample_fmt; const char *fmt_be, *fmt_le;
  203.     } sample_fmt_entries[] = {
  204.         { AV_SAMPLE_FMT_U8,  "u8",    "u8"    },
  205.         { AV_SAMPLE_FMT_S16, "s16be", "s16le" },
  206.         { AV_SAMPLE_FMT_S32, "s32be", "s32le" },
  207.         { AV_SAMPLE_FMT_FLT, "f32be", "f32le" },
  208.         { AV_SAMPLE_FMT_DBL, "f64be", "f64le" },
  209.     };
  210.     *fmt = NULL;
  211.  
  212.     for (i = 0; i < FF_ARRAY_ELEMS(sample_fmt_entries); i++) {
  213.         struct sample_fmt_entry *entry = &sample_fmt_entries[i];
  214.         if (sample_fmt == entry->sample_fmt) {
  215.             *fmt = AV_NE(entry->fmt_be, entry->fmt_le);
  216.             return 0;
  217.         }
  218.     }
  219.  
  220.     fprintf(stderr,
  221.             "sample format %s is not supported as output format\n",
  222.             av_get_sample_fmt_name(sample_fmt));
  223.     return -1;
  224. }
  225.  
  226. int main (int argc, char **argv)
  227. {
  228.     int ret = 0, got_frame;
  229.  
  230.     if (argc != 4 && argc != 5) {
  231.         fprintf(stderr, "usage: %s [-refcount=<old|new_norefcount|new_refcount>] "
  232.                 "input_file video_output_file audio_output_file\n"
  233.                 "API example program to show how to read frames from an input file.\n"
  234.                 "This program reads frames from a file, decodes them, and writes decoded\n"
  235.                 "video frames to a rawvideo file named video_output_file, and decoded\n"
  236.                 "audio frames to a rawaudio file named audio_output_file.\n\n"
  237.                 "If the -refcount option is specified, the program use the\n"
  238.                 "reference counting frame system which allows keeping a copy of\n"
  239.                 "the data for longer than one decode call. If unset, it's using\n"
  240.                 "the classic old method.\n"
  241.                 "\n", argv[0]);
  242.         exit(1);
  243.     }
  244.     if (argc == 5) {
  245.         const char *mode = argv[1] + strlen("-refcount=");
  246.         if      (!strcmp(mode, "old"))            api_mode = API_MODE_OLD;
  247.         else if (!strcmp(mode, "new_norefcount")) api_mode = API_MODE_NEW_API_NO_REF_COUNT;
  248.         else if (!strcmp(mode, "new_refcount"))   api_mode = API_MODE_NEW_API_REF_COUNT;
  249.         else {
  250.             fprintf(stderr, "unknow mode '%s'\n", mode);
  251.             exit(1);
  252.         }
  253.         argv++;
  254.     }
  255.     src_filename = argv[1];
  256.     video_dst_filename = argv[2];
  257.     audio_dst_filename = argv[3];
  258.  
  259.     /* register all formats and codecs */
  260.     av_register_all();
  261.  
  262.     /* open input file, and allocate format context */
  263.     if (avformat_open_input(&fmt_ctx, src_filename, NULL, NULL) < 0) {
  264.         fprintf(stderr, "Could not open source file %s\n", src_filename);
  265.         exit(1);
  266.     }
  267.  
  268.     /* retrieve stream information */
  269.     if (avformat_find_stream_info(fmt_ctx, NULL) < 0) {
  270.         fprintf(stderr, "Could not find stream information\n");
  271.         exit(1);
  272.     }
  273.  
  274.     if (open_codec_context(&video_stream_idx, fmt_ctx, AVMEDIA_TYPE_VIDEO) >= 0) {
  275.         video_stream = fmt_ctx->streams[video_stream_idx];
  276.         video_dec_ctx = video_stream->codec;
  277.  
  278.         video_dst_file = fopen(video_dst_filename, "wb");
  279.         if (!video_dst_file) {
  280.             fprintf(stderr, "Could not open destination file %s\n", video_dst_filename);
  281.             ret = 1;
  282.             goto end;
  283.         }
  284.  
  285.         /* allocate image where the decoded image will be put */
  286.         width = video_dec_ctx->width;
  287.         height = video_dec_ctx->height;
  288.         pix_fmt = video_dec_ctx->pix_fmt;
  289.         ret = av_image_alloc(video_dst_data, video_dst_linesize,
  290.                              width, height, pix_fmt, 1);
  291.         if (ret < 0) {
  292.             fprintf(stderr, "Could not allocate raw video buffer\n");
  293.             goto end;
  294.         }
  295.         video_dst_bufsize = ret;
  296.     }
  297.  
  298.     if (open_codec_context(&audio_stream_idx, fmt_ctx, AVMEDIA_TYPE_AUDIO) >= 0) {
  299.         audio_stream = fmt_ctx->streams[audio_stream_idx];
  300.         audio_dec_ctx = audio_stream->codec;
  301.         audio_dst_file = fopen(audio_dst_filename, "wb");
  302.         if (!audio_dst_file) {
  303.             fprintf(stderr, "Could not open destination file %s\n", audio_dst_filename);
  304.             ret = 1;
  305.             goto end;
  306.         }
  307.     }
  308.  
  309.     /* dump input information to stderr */
  310.     av_dump_format(fmt_ctx, 0, src_filename, 0);
  311.  
  312.     if (!audio_stream && !video_stream) {
  313.         fprintf(stderr, "Could not find audio or video stream in the input, aborting\n");
  314.         ret = 1;
  315.         goto end;
  316.     }
  317.  
  318.     /* When using the new API, you need to use the libavutil/frame.h API, while
  319.      * the classic frame management is available in libavcodec */
  320.     if (api_mode == API_MODE_OLD)
  321.         frame = avcodec_alloc_frame();
  322.     else
  323.         frame = av_frame_alloc();
  324.     if (!frame) {
  325.         fprintf(stderr, "Could not allocate frame\n");
  326.         ret = AVERROR(ENOMEM);
  327.         goto end;
  328.     }
  329.  
  330.     /* initialize packet, set data to NULL, let the demuxer fill it */
  331.     av_init_packet(&pkt);
  332.     pkt.data = NULL;
  333.     pkt.size = 0;
  334.  
  335.     if (video_stream)
  336.         printf("Demuxing video from file '%s' into '%s'\n", src_filename, video_dst_filename);
  337.     if (audio_stream)
  338.         printf("Demuxing audio from file '%s' into '%s'\n", src_filename, audio_dst_filename);
  339.  
  340.     /* read frames from the file */
  341.     while (av_read_frame(fmt_ctx, &pkt) >= 0) {
  342.         AVPacket orig_pkt = pkt;
  343.         do {
  344.             ret = decode_packet(&got_frame, 0);
  345.             if (ret < 0)
  346.                 break;
  347.             pkt.data += ret;
  348.             pkt.size -= ret;
  349.         } while (pkt.size > 0);
  350.         av_free_packet(&orig_pkt);
  351.     }
  352.  
  353.     /* flush cached frames */
  354.     pkt.data = NULL;
  355.     pkt.size = 0;
  356.     do {
  357.         decode_packet(&got_frame, 1);
  358.     } while (got_frame);
  359.  
  360.     printf("Demuxing succeeded.\n");
  361.  
  362.     if (video_stream) {
  363.         printf("Play the output video file with the command:\n"
  364.                "ffplay -f rawvideo -pix_fmt %s -video_size %dx%d %s\n",
  365.                av_get_pix_fmt_name(pix_fmt), width, height,
  366.                video_dst_filename);
  367.     }
  368.  
  369.     if (audio_stream) {
  370.         enum AVSampleFormat sfmt = audio_dec_ctx->sample_fmt;
  371.         int n_channels = audio_dec_ctx->channels;
  372.         const char *fmt;
  373.  
  374.         if (av_sample_fmt_is_planar(sfmt)) {
  375.             const char *packed = av_get_sample_fmt_name(sfmt);
  376.             printf("Warning: the sample format the decoder produced is planar "
  377.                    "(%s). This example will output the first channel only.\n",
  378.                    packed ? packed : "?");
  379.             sfmt = av_get_packed_sample_fmt(sfmt);
  380.             n_channels = 1;
  381.         }
  382.  
  383.         if ((ret = get_format_from_sample_fmt(&fmt, sfmt)) < 0)
  384.             goto end;
  385.  
  386.         printf("Play the output audio file with the command:\n"
  387.                "ffplay -f %s -ac %d -ar %d %s\n",
  388.                fmt, n_channels, audio_dec_ctx->sample_rate,
  389.                audio_dst_filename);
  390.     }
  391.  
  392. end:
  393.     avcodec_close(video_dec_ctx);
  394.     avcodec_close(audio_dec_ctx);
  395.     avformat_close_input(&fmt_ctx);
  396.     if (video_dst_file)
  397.         fclose(video_dst_file);
  398.     if (audio_dst_file)
  399.         fclose(audio_dst_file);
  400.     if (api_mode == API_MODE_OLD)
  401.         avcodec_free_frame(&frame);
  402.     else
  403.         av_frame_free(&frame);
  404.     av_free(video_dst_data[0]);
  405.  
  406.     return ret < 0;
  407. }
  408.