path: root/rotord/src/libavwrapper.cpp

#include "libavwrapper.h"

extern Poco::Mutex mutex; //application wide mutex
static Poco::Mutex mutex;

#include <stdexcept>
#include <iostream>
#include <cassert>

using namespace std;

void libav::maybeInitFFMpegLib()
{
    if (b_is_one_time_inited)
        return;
    FFMS_Init(0, 0); //should do for all
    //av_register_all();
    //avcodec_register_all();
    avformat_network_init();
    b_is_one_time_inited = true;
}

 void libav::video_decoder::cleanup(){
    if (loaded) {
        mutex.lock();
        FFMS_DestroyVideoSource(source);
        mutex.unlock();
        loaded=false;
    }
}

bool libav::video_decoder::open(const std::string& filename){
    mutex.lock();
    loaded=false;
    FFMS_Index *index = FFMS_MakeIndex(filename.c_str(), 0, 0, NULL, NULL, FFMS_IEH_IGNORE, NULL, NULL, &err);
    if (index == NULL) {
        std::cerr<<"ffmpegsource: "<<err.Buffer<<std::endl;
        mutex.unlock();
        return false;
    }
    int trackno = FFMS_GetFirstTrackOfType(index, FFMS_TYPE_VIDEO, &err);
    if (trackno < 0) {
        std::cerr<<"ffmpegsource: "<<err.Buffer<<std::endl;
        mutex.unlock();
        return false;
    }
    source = FFMS_CreateVideoSource(filename.c_str(), trackno, index, 1, FFMS_SEEK_NORMAL, &err);
    if (source == NULL) {
        std::cerr<<"ffmpegsource: "<<err.Buffer<<std::endl;
        mutex.unlock();
        return false;
    }
    FFMS_DestroyIndex(index);
    props = FFMS_GetVideoProperties(source);
    const FFMS_Frame *propframe = FFMS_GetFrame(source, 0, &err);
    w=propframe->EncodedWidth;
    h=propframe->EncodedHeight;
    //propframe->EncodedPixelFormat;
    
    if (FFMS_SetOutputFormatV2(source, pixfmts, propframe->EncodedWidth, propframe->EncodedHeight, FFMS_RESIZER_BICUBIC, &err)) {
        std::cerr<<"ffmpegsource: "<<err.Buffer<<std::endl;
        mutex.unlock();
        return false;
    }
    int framenumber = 0; /* valid until next call to FFMS_GetFrame* on the same video object */
    
    
    std::cerr<<"ffmpegsource: successfully opened "<<filename<<std::endl;
    loaded=true;
    mutex.unlock();
    return loaded;
}

bool libav::audio_decoder::open(const std::string& filename){
    mutex.lock();
    loaded=false;
    FFMS_Index *index = FFMS_MakeIndex(filename.c_str(),-1, 0, NULL, NULL, FFMS_IEH_IGNORE, NULL, NULL, &err);
    if (index == NULL) {
        std::cerr<<"ffmpegsource error making index for "<<filename<<":"<<err.Buffer<<std::endl;
        mutex.unlock();
        return false;
    }
    int trackno = FFMS_GetFirstTrackOfType(index, FFMS_TYPE_AUDIO, &err);
    if (trackno < 0) {
        std::cerr<<"ffmpegsource error finding audio track in "<<filename<<":"<<err.Buffer<<std::endl;
        mutex.unlock();
        return false;
    }
    std::cerr<<"ffmpegsource found audio track "<<trackno<<" in "<<filename<<":"<<err.Buffer<<std::endl;
    source = FFMS_CreateAudioSource(filename.c_str(), trackno, index, FFMS_DELAY_TIME_ZERO,  &err);
    if (source == NULL) {
        std::cerr<<"ffmpegsource error creating audio source from "<<filename<<":"<<err.Buffer<<std::endl;
        mutex.unlock();
        return false;
    }
    FFMS_DestroyIndex(index);
    props = FFMS_GetAudioProperties(source);

    std::cerr<<"ffmpegsource: successfully opened "<<filename<<std::endl;
    loaded=true;
    mutex.unlock();
    return loaded;
}
 void libav::audio_decoder::cleanup(){
    if (loaded) {
        mutex.lock();
        FFMS_DestroyAudioSource(source);
        mutex.unlock();
        loaded=false;
    }
}

bool libav::exporter::setup(int w,int h, int bitRate, int frameRate, std::string container, bool _fragmentation){

    maybeInitFFMpegLib();
    
    fragmentation=_fragmentation;

    this->w=w;
    this->h=h;
    this->bitRate=bitRate;
    this->frameRate=frameRate;
    this->container=container;

    if (NULL != sws_ctx) {
        sws_freeContext(sws_ctx);
        sws_ctx = NULL;
    }

    sws_ctx = sws_getContext(w, h, AV_PIX_FMT_RGB24,
                             w, h, AV_PIX_FMT_YUV420P,
                             sws_flags, NULL, NULL, NULL);

    return true;
}

bool libav::exporter::record(std::string filename){

    printf("exporter 1 - using filename '%s'.\n",filename.c_str());

    // allocate the output media context //
    avformat_alloc_output_context2(&oc, NULL, NULL, filename.c_str());

    printf("exporter 2.\n");

    if (!oc) {
        printf("Could not deduce output format from file extension: using MPEG.\n");
        avformat_alloc_output_context2(&oc, NULL, "mpeg", filename.c_str());
    }
    if (!oc) {
        return false;
    }
    fmt = oc->oformat;

    printf("exporter 3.\n");

    // Add the audio and video streams using the default format codecs
    // * and initialize the codecs. //
    video_st = NULL;
    audio_st = NULL;

    fmt->video_codec=AV_CODEC_ID_H264; //AV_CODEC_ID_MPEG4

    if (fmt->video_codec != AV_CODEC_ID_NONE) {
        video_st = add_stream(oc, &video_codec, fmt->video_codec);
    }
    if (fmt->audio_codec != AV_CODEC_ID_NONE) {
        //audio_st = add_stream(oc, &audio_codec, fmt->audio_codec);
        audio_st = add_stream(oc, &audio_codec, fmt->audio_codec);
    }

    printf("exporter 4.\n");

    //set initial video params
    video_st->codec->width=w;
    video_st->codec->height=h;
    video_st->codec->time_base.num = 1;//codecCtx->ticks_per_frame;
    video_st->codec->time_base.den = frameRate;
    video_st->time_base = video_st->codec->time_base;
        //audioStream->time_base = codecCtx->time_base; //???has the capability of crashing

    video_st->codec->gop_size = 75; /* emit one intra frame every ten frames */
    video_st->codec->pix_fmt = PIX_FMT_YUV420P;


    //video_st->codec->bit_rate = bitRate; //need to deal with resolution etc
    //video_st->codec->rc_max_rate = bitRate; 
    //video_st->codec->rc_min_rate = 0; 

    //how to set profile????? 190913
    //this causes a crash
    //av_opt_set(video_st->codec->priv_data,"profile",FF_PROFILE_H264_CONSTRAINED_BASELINE, 0); 
    //
    // http://libav-users.943685.n4.nabble.com/Libav-user-encoding-bitrate-wrong-td4433740.html
    // http://libav-users.943685.n4.nabble.com/Libav-user-H264-encoding-set-target-average-bitrate-td3912529.html
    // http://trac.ffmpeg.org/wiki/x264EncodingGuide

    // Now that all the parameters are set, we can open the audio and
    // * video codecs and allocate the necessary encode buffers. //
    if (video_st)
        open_video(oc, video_codec, video_st);
    if (audio_st) {
        audioframesize=open_audio(oc, audio_codec, audio_st);
        audiostep=((float)audioframesize)/(audio_st->codec->sample_rate);
        std::cerr << "opened audio codec with "<<audioframesize<<" frame size and "<<audiostep<<" seconds per frame"<<std::endl;
    }

    printf("exporter 5.\n");


    av_dump_format(oc, 0, filename.c_str(), 1);

    // open the output file, if needed //
    if (!(fmt->flags & AVFMT_NOFILE)) {
        mutex.lock();
        int ret = avio_open(&oc->pb, filename.c_str(), AVIO_FLAG_WRITE);
        mutex.unlock();
        if (ret < 0) {
            std::cerr <<"Could not open " << filename.c_str() << std::endl;
            return false;
        }
    }

    AVDictionary *opts = NULL; // "create" an empty dictionary

    //THIS DOES SEEM TO SET CONTAINER OPTS= AS MOOV_SIZE MAKES THE MOVIE BANJAXED
    //av_dict_set(&opts, "moov_size", "20000", 0);
    if (fragmentation) {
        av_dict_set(&opts, "movflags","frag_keyframe", 0);
    }


    // Write the stream header, if any. //
    int ret = avformat_write_header(oc, &opts);
    if (ret < 0) {
        std::cerr <<"Error occurred when opening output file:" <<endl; // av_err2str(ret) << std::endl;
        return false;
    }
    //#include <libavformat/movenc.h>
    //mov_write_moov_tag(AVIOContext *pb, MOVMuxContext *mov, AVFormatContext *s)
    /*ret = mov_write_moov_tag(oc, NULL,NULL);
     if (ret < 0) {
        std::cerr <<"Error occurred when writing moov atom " <<endl; // << av_err2str(ret) << std::endl;
        return false;
    }*/

    if (frame)
        frame->pts = 0;

    outputframe=0;

    return true;
}
bool libav::exporter::encodeFrame(unsigned char *pixels,uint16_t *samples){
    // Compute current audio and video time. //
    if (audio_st)
        audio_pts = (double)audio_st->pts.val * audio_st->time_base.num / audio_st->time_base.den;
    else
        audio_pts = 0.0;

    if (video_st)
        video_pts = (double)video_st->pts.val * video_st->time_base.num /
                    video_st->time_base.den;
    else
        video_pts = 0.0;

    // write interleaved audio and video frames //
    if (!video_st || (video_st && audio_st && audio_pts < video_pts)) {
        write_audio_frame(oc, audio_st, samples);
    } else {
        write_video_frame(oc, video_st, pixels);

        frame->pts += av_rescale_q(1, video_st->codec->time_base, video_st->time_base);
    }

    //std::cerr << "encoded frame " << outputframe << std::endl;
    outputframe++;

    return true;
}
bool libav::exporter::encodeFrame(unsigned char *pixels,AVPacket *audio){
    // Compute current audio and video time. //
    if (audio_st)
        audio_pts = (double)audio_st->pts.val * audio_st->time_base.num / audio_st->time_base.den;
    else
        audio_pts = 0.0;

    if (video_st)
        video_pts = (double)video_st->pts.val * video_st->time_base.num /
                    video_st->time_base.den;
    else
        video_pts = 0.0;

    // write interleaved audio and video frames //
    if (!video_st || (video_st && audio_st && audio_pts < video_pts)) {
        write_audio_frame(oc, audio_st, audio);
    } else {
        write_video_frame(oc, video_st, pixels);

        frame->pts += av_rescale_q(1, video_st->codec->time_base, video_st->time_base);
    }

    //std::cerr << "encoded frame " << outputframe << std::endl;
    outputframe++;

    return true;
}
bool libav::exporter::encodeFrame(unsigned char *pixels){
    video_pts = (double)video_st->pts.val * video_st->time_base.num / video_st->time_base.den;
    write_video_frame(oc, video_st, pixels);
    frame->pts += av_rescale_q(1, video_st->codec->time_base, video_st->time_base);
    outputframe++;
    return true;
}
bool libav::exporter::encodeFrame(uint16_t *samples){
    audio_pts = (double)audio_st->pts.val * audio_st->time_base.num / audio_st->time_base.den;
    write_audio_frame(oc, audio_st, samples);
    return true;
}
void libav::exporter::finishRecord(){

    av_write_trailer(oc);
    // Close each codec. //
    if (video_st)
        close_video(oc, video_st);
    if (audio_st)
        close_audio(oc, audio_st);

    if (!(fmt->flags & AVFMT_NOFILE)) {
        // Close the output file. //
        mutex.lock();
        avio_close(oc->pb);
        mutex.unlock();
    }

    // free the stream //
    avformat_free_context(oc);
}

AVStream* libav::exporter::add_stream(AVFormatContext *oc, AVCodec **codec,enum AVCodecID codec_id)
    {
        AVCodecContext *c;
        AVStream *st;

        // find the encoder //
        *codec = avcodec_find_encoder(codec_id);
        if (!(*codec)) {
            //fprintf(stderr, "Could not find encoder for '%s'\n",
            //        avcodec_get_name(codec_id));
            exit(1);
        }

        st = avformat_new_stream(oc, *codec);
        if (!st) {
            //fprintf(stderr, "Could not allocate stream\n");
            exit(1);
        }
        st->id = oc->nb_streams-1;
        c = st->codec;

        switch ((*codec)->type) {
        case AVMEDIA_TYPE_AUDIO:
            st->id = 1;
            c->sample_fmt  = AV_SAMPLE_FMT_S16;
            c->bit_rate    = 256000;
            c->sample_rate = 44100;
            c->channels    = 2;
            c->channel_layout=AV_CH_LAYOUT_STEREO;
            break;

        case AVMEDIA_TYPE_VIDEO:
            c->codec_id = codec_id;

            c->bit_rate = bitRate; //need to deal with resolution etc

            c->rc_max_rate = bitRate; 
            c->rc_min_rate = 0; 
            //c->rc_buffer_size = Profile()->m_videoMaxVopSize;  //??


            // Resolution must be a multiple of two. //
            //c->width    = 352;
            //c->height   = 288;


            // timebase: This is the fundamental unit of time (in seconds) in terms
            // * of which frame timestamps are represented. For fixed-fps content,
            // * timebase should be 1/framerate and timestamp increments should be
            // * identical to 1. //
            c->time_base.den = frameRate;
            c->time_base.num = 1;
            c->gop_size      = 12; // emit one intra frame every twelve frames at most //
            c->pix_fmt       = AV_PIX_FMT_YUV420P; //ADDED HARDCODED TJR 280513
            if (c->codec_id == AV_CODEC_ID_MPEG2VIDEO) {
                // just for testing, we also add B frames //
                c->max_b_frames = 2;
            }
            if (c->codec_id == AV_CODEC_ID_MPEG1VIDEO) {
                // Needed to avoid using macroblocks in which some coeffs overflow.
                // * This does not happen with normal video, it just happens here as
                // * the motion of the chroma plane does not match the luma plane. //
                c->mb_decision = 2;
            }

        break;

        default:
            break;
        }

        // Some formats want stream headers to be separate. //
        if (oc->oformat->flags & AVFMT_GLOBALHEADER)
            c->flags |= CODEC_FLAG_GLOBAL_HEADER;

        return st;
    }

void libav::exporter::open_video(AVFormatContext *oc, AVCodec *codec, AVStream *st)
    {
        int ret;
        AVCodecContext *c = st->codec;

        // open the codec //
        mutex.lock();
        ret = avcodec_open2(c, codec, NULL); //OPTIONS CAN GO HERE
        mutex.unlock();
        if (ret < 0) {
            //fprintf(stderr, "Could not open video codec: %s\n", av_err2str(ret));
            exit(1);
        }

        // allocate and init a re-usable frame //
        frame = avcodec_alloc_frame();
        // moved to constructor and freeing in destructor -- stills crashes the same
        if (!frame) {
            //fprintf(stderr, "Could not allocate video frame\n");
            exit(1);
        }

        // Allocate the encoded raw picture. //
        ret = avpicture_alloc(&dst_picture, c->pix_fmt, c->width, c->height);
        if (ret < 0) {
            //fprintf(stderr, "Could not allocate picture: %s\n", av_err2str(ret));
            exit(1);
        }

        // If the output format is not YUV420P, then a temporary YUV420P
        // * picture is needed too. It is then converted to the required
        // * output format. //
        if (c->pix_fmt != AV_PIX_FMT_YUV420P) {
            ret = avpicture_alloc(&src_picture, AV_PIX_FMT_RGB24, c->width, c->height);
            if (ret < 0) {
                //fprintf(stderr, "Could not allocate temporary picture: %s\n",
                //        av_err2str(ret));
                exit(1);
            }
        }



        // copy data and linesize picture pointers to frame //
        *((AVPicture *)frame) = dst_picture;

        outPixels = (uint8_t*)av_malloc(avpicture_get_size(PIX_FMT_YUV420P, st->codec->width,st->codec->height));
    }

    int libav::exporter::open_audio(AVFormatContext *oc, AVCodec *codec, AVStream *st)
    {
        AVCodecContext *c;
        int ret;

        c = st->codec;

        // open it //
        mutex.lock();
        ret = avcodec_open2(c, codec, NULL);
        mutex.unlock();
        if (ret < 0) {
            //fprintf(stderr, "Could not open audio codec: %s\n", av_err2str(ret));
            exit(1);
        }

        // init signal generator //
        t     = 0;
        tincr = 2 * M_PI * 110.0 / c->sample_rate;
        // increment frequency by 110 Hz per second //
        tincr2 = 2 * M_PI * 110.0 / c->sample_rate / c->sample_rate;

        if (c->codec->capabilities & CODEC_CAP_VARIABLE_FRAME_SIZE)
            audio_input_frame_size = 10000;
        else
            audio_input_frame_size = c->frame_size;

        /*
        samples = av_malloc(audio_input_frame_size *
                            av_get_bytes_per_sample(c->sample_fmt) *
                            c->channels);
        if (!samples) {
            //fprintf(stderr, "Could not allocate audio samples buffer\n");
            exit(1);
        }
        */
        return audio_input_frame_size;
    }

    void libav::exporter::write_audio_frame(AVFormatContext *oc, AVStream *st,uint16_t *samples)
    {
        AVCodecContext *c;
        AVPacket pkt = { 0 }; // data and size must be 0;
        AVFrame *frame = avcodec_alloc_frame(); 
        int got_packet, ret;

        //av_init_packet(&pkt);  111013 NOT NECESSARY
        c = st->codec;

        //get_audio_frame(samples, audio_input_frame_size, c->channels);
        frame->nb_samples = audio_input_frame_size;
        uint8_t *sampleptr;
        int bufsize=audio_input_frame_size * av_get_bytes_per_sample(c->sample_fmt) *c->channels;
        if (samples) {
            sampleptr=(uint8_t*)samples;
        }
        else {
            sampleptr=new uint8_t[bufsize];
            memset(sampleptr,0,bufsize);
        }

        avcodec_fill_audio_frame(frame, c->channels, c->sample_fmt,
                                 sampleptr,
                                 audio_input_frame_size *
                                 av_get_bytes_per_sample(c->sample_fmt) *
                                 c->channels, 0); //;
        //frame->sample_rate=44100; //hard coded input rate- nope, this doesn't help
        //frame->format=AV_SAMPLE_FMT_S16P;
        //?? why is ffmpeg reporting fltp as the sample format??? doesn't seem to have an effect to change this though
        ret = avcodec_encode_audio2(c, &pkt, frame, &got_packet);
        if (!samples) {
            delete[] sampleptr;
        }
        if (ret < 0) {
            //fprintf(stderr, "Error encoding audio frame: %s\n", av_err2str(ret));
            exit(1);
        }

        if (!got_packet)
            return;

        pkt.stream_index = st->index;

        //avcodec_free_frame(&frame); ///removed 091013


        // Write the compressed frame to the media file. //
        ret = av_interleaved_write_frame(oc, &pkt);
        av_free_packet(&pkt);
        if (ret != 0) {
            //fprintf(stderr, "Error while writing audio frame: %s\n",
            //        av_err2str(ret));
            exit(1);
        }
        
    }

    void libav::exporter::write_audio_frame(AVFormatContext *oc, AVStream *st,AVPacket *pkt)
    {
      
        pkt->stream_index = st->index;

        // Write the compressed frame to the media file. //
        int ret = av_interleaved_write_frame(oc, pkt);
        if (ret != 0) {
            //fprintf(stderr, "Error while writing audio frame: %s\n",
            //        av_err2str(ret));
            exit(1);
        }
        //avcodec_free_frame(&frame);
        av_free_packet(pkt); ///added 091013
    }

    void libav::exporter::close_audio(AVFormatContext *oc, AVStream *st)
    {
        mutex.lock();
        avcodec_close(st->codec);
        mutex.unlock();

    }

    void libav::exporter::write_video_frame(AVFormatContext *oc, AVStream *st, uint8_t *pixels)
    {
        int ret;

        AVCodecContext *c = st->codec;

        avpicture_fill(&src_picture, pixels, PIX_FMT_RGB24, c->width,c->height);
        //avpicture_fill(&dst_picture, outPixels, PIX_FMT_YUV420P, c->width,c->height);

        sws_scale(sws_ctx, src_picture.data, src_picture.linesize, 0, c->height, dst_picture.data, dst_picture.linesize);
        //fill_yuv_image(&dst_picture, frame_count, c->width, c->height);
        if (oc->oformat->flags & AVFMT_RAWPICTURE) {
            // Raw video case - directly store the picture in the packet //
            AVPacket pkt;
            //av_init_packet(&pkt); ///removed 101013  NOT NECESSARY

            pkt.flags        |= AV_PKT_FLAG_KEY;
            pkt.stream_index  = st->index;
            pkt.data          = dst_picture.data[0];
            pkt.size          = sizeof(AVPicture);

            ret = av_interleaved_write_frame(oc, &pkt);
            av_free_packet(&pkt); ///added 091013
        } else {
            AVPacket pkt = { 0 };
            int got_packet;
           
            ret = avcodec_encode_video2(c, &pkt, frame, &got_packet);
            if (ret < 0) {
                //fprintf(stderr, "Error encoding video frame: %s\n", av_err2str(ret));
                exit(1);
            }
            // If size is zero, it means the image was buffered. //

            if (!ret && got_packet && pkt.size) {
                pkt.stream_index = st->index;

                // Write the compressed frame to the media file. //
                ret = av_interleaved_write_frame(oc, &pkt);
            } else {
                ret = 0;
            }
            av_free_packet(&pkt); ///added 091013
        }

        


        if (ret != 0) {
            //fprintf(stderr, "Error while writing video frame: %s\n", av_err2str(ret));
            exit(1);
        }
        frame_count++;

        //avcodec_free_frame(&frame);
    }

    void libav::exporter::close_video(AVFormatContext *oc, AVStream *st)
    {
        mutex.lock();
        //avcodec_close(st->codec); //change 0706 to trace 2nd render issue
        avcodec_close(audio_st->codec);
        avcodec_close(video_st->codec);
        //
        //


        //av_free(src_picture.data[0]); //removed to explore weird 2nd render crash.. seems to WORK -- seems that the picture data is owned elsewhere
        av_free(dst_picture.data[0]);
        av_free(frame); //removed to explore crash 2nd time render
        //gives *** Error in `./rotord': corrupted double-linked list: 0x00007fd8b005bd60 ***
        //where is frame initialised???
        //moved to destructor


        av_free(outPixels); //SIGSEV here???
        mutex.unlock();
    }