live555的客服端流程：建立任务计划对象－－建立环境对象－－处理用户输入的参数(RTSP地址)－－创建RTSPClient实例－－发出DESCRIBE--发出SETUP--发出PLAY－－进入Loop循环接收数据--发出TEARDOWN结束连接。
可以抽成3个函数接口：rtspOpen rtspRead rtspClose。
首先我们来分析rtspOpen的过程：
int rtspOpen(rtsp_object_t *p_obj, int tcpConnect)
{
... ...
TRACE1_DEC("BasicTaskScheduler::createNew !!!\n" );
if( ( p_sys->scheduler = BasicTaskScheduler::createNew() ) == NULL )
{
TRACE1_DEC("BasicTaskScheduler::createNew failed\n" );
goto error;
} if( !( p_sys->env = BasicUsageEnvironment::createNew(*p_sys->scheduler) ) )
{
TRACE1_DEC("BasicUsageEnvironment::createNew failed\n");
goto error;
} if( ( i_return = Connect( p_obj ) ) != RTSP_SUCCESS )
{
TRACE1_DEC( "Failed to connect with %s\n", p_obj->rtspURL );
goto error;
} if( p_sys->p_sdp == NULL )
{
TRACE1_DEC( "Failed to retrieve the RTSP Session Description\n" );
goto error;
} if( ( i_return = SessionsSetup( p_obj ) ) != RTSP_SUCCESS )
{
TRACE1_DEC( "Nothing to play for rtsp://%s\n", p_obj->rtspURL );
goto error;
} if( ( i_return = Play( p_obj ) ) != RTSP_SUCCESS )
goto error; ... ...
}
1> BasicTaskScheduler::createNew()
2> BasicUsageEnvironment::createNew()
3> connect
static int Connect( rtsp_object_t *p_demux )
{
... ...
sprintf(appName, "LibRTSP%d", p_demux->id);
if( ( p_sys->rtsp = RTSPClient::createNew( *p_sys->env, , appName, i_http_port ) ) == NULL )
{
TRACE1_DEC( "RTSPClient::createNew failed (%s)\n",
p_sys->env->getResultMsg() ); i_ret = RTSP_ERROR;
goto connect_error;
} psz_options = p_sys->rtsp->sendOptionsCmd( p_demux->rtspURL, psz_user, psz_pwd ); if(psz_options == NULL)
TRACE1_DEC("RTSP Option commend error!!\n"); delete [] psz_options; p_sdp = p_sys->rtsp->describeURL( p_demux->rtspURL );
... ...
}
connect中做了三件事：RTSPClient类的实例，发送“OPTIONS”请求，发送“describeURL”请求。
sendOptionsCmd()函数首先调用openConnectionFromURL()函数进程tcp连接，然后组包发送：
OPTIONS rtsp://120.90.0.50:8552/h264_ch2 RTSP/1.0
CSeq:
User-Agent: LibRTSP4 (LIVE555 Streaming Media v2008.04.02)
收到服务器的应答：
RTSP/1.0 OK
CSeq:
Date: Mon, May :: GMT
Public: OPTIONS, DESCRIBE, SETUP, TEARDOWN, PLAY, PAUSE
describeURL()函数首先也会调用openConnectionFromURL()函数进行TCP连接（这里可以看出先发OPTIONS请求，也可以先发describeURL请求），然后组包发送：
DESCRIBE rtsp://120.90.0.50:8552/h264_ch2 RTSP/1.0
CSeq:
Accept: application/sdp
User-Agent: LibRTSP4 (LIVE555 Streaming Media v2008.04.02)
收到服务器应答：
DESCRIBE rtsp://120.90.0.50:8552/h264_ch2 RTSP/1.0
CSeq:
Accept: application/sdp
User-Agent: LibRTSP4 (LIVE555 Streaming Media v2008.04.02) Received DESCRIBE response:
RTSP/1.0 OK
CSeq:
Date: Mon, May :: GMT
Content-Base: rtsp://192.168.103.51:8552/h264_ch2/
Content-Type: application/sdp
Content-Length: Need to read extra bytes
Read extra bytes: v=
o=- IN IP4 192.168.103.51
s=RTSP/RTP stream from NETRA
i=h264_ch2
t=
a=tool:LIVE555 Streaming Media v2008.04.02
a=type:broadcast
a=control:*
a=range:npt=-
a=x-qt-text-nam:RTSP/RTP stream from NETRA
a=x-qt-text-inf:h264_ch2
m=video RTP/AVP
c=IN IP4 0.0.0.0
a=rtpmap: H264/
a=fmtp: packetization-mode=;profile-level-id=;sprop-parameter-sets=h264
a=control:track1
m=audio RTP/AVP
c=IN IP4 0.0.0.0
a=rtpmap: PCMU//
a=control:track2
4> SessionsSetup
static int SessionsSetup( rtsp_object_t *p_demux )
{
... ...
// unsigned const thresh = 1000000;
if( !( p_sys->ms = MediaSession::createNew( *p_sys->env, p_sys->p_sdp ) ) )
{
TRACE1_DEC( "Could not create the RTSP Session: %s\n", p_sys->env->getResultMsg() );
return RTSP_ERROR;
} /* Initialise each media subsession */
iter = new MediaSubsessionIterator( *p_sys->ms );
while( ( sub = iter->next() ) != NULL )
{
... ...
bInit = sub->initiate(); if( !bInit )
{
TRACE1_DEC( "RTP subsession '%s/%s' failed (%s)\n",
sub->mediumName(), sub->codecName(), p_sys->env->getResultMsg() );
}
else
{
... ...
/* Issue the SETUP */
if( p_sys->rtsp )
{
if( !p_sys->rtsp->setupMediaSubsession( *sub, False, b_rtsp_tcp, False ) )
{
/* if we get an unsupported transport error, toggle TCP
* use and try again */
if( !strstr(p_sys->env->getResultMsg(),"461 Unsupported Transport")
|| !p_sys->rtsp->setupMediaSubsession( *sub, False, b_rtsp_tcp, False ) )
{
TRACE1_DEC( "SETUP of'%s/%s' failed %s\n", sub->mediumName(), sub->codecName(), p_sys->env->getResultMsg() );
continue;
}
}
} ... .../* Value taken from mplayer */
if( !strcmp( sub->mediumName(), "audio" ) )
{
if( !strcmp( sub->codecName(), "MP4A-LATM" ) )
{
... ...
}
else if( !strcmp( sub->codecName(), "PCMA" ) || !strcmp( sub->codecName(), "PCMU" ))
{
tk->fmt.i_extra = ;
tk->fmt.i_codec = RTSP_CODEC_PCMU;
}
}
else if( !strcmp( sub->mediumName(), "video" ) )
{
if( !strcmp( sub->codecName(), "H264" ) )
{
... ...
}
else if( !strcmp( sub->codecName(), "MP4V-ES" ) )
{
... ...
}
else if( !strcmp( sub->codecName(), "JPEG" ) )
{
tk->fmt.i_codec = RTSP_CODEC_MJPG;
}
}
... ...
}
}
... ...
}
这个函数做了四件事：创建MediaSession类的实例，创建MediaSubsessionIterator类的实例，MediaSubsession的初始化，发送"SETUP"请求。
创建MediaSession实例的同时，会调用initializeWithSDP()函数去解析SDP，解析出"s="相对应的fSessionName，解析出"s="相对应的fSessionName，解析出"i="相对应的fSessionDescription，解析出"c="相对应的connectionEndpointName，解析出"a=type:"相对应的fMediaSessionType等等。创建MediaSubsession类的实例，并且加入到fSubsessionsHead链表中，从上面的SDP描述来看，有两个MediaSubsession，一个video，一个audio。
创建MediaSubsessionIterator类的实例，并且调用reset函数，将fOurSession.fSubsessionsHead赋值给fNextPtr，也就是将链表的头结点赋值给fNextPtr。当执行while循环的时候，执行了两次，一次video，一次audio。
initiate函数，根据fSourceFilterAddr来判断是否是SSM，还是ASM，然后调用Groupsock的不同构造函数来创建实例fRTPSocket、fRTCPSocket；然后根据协议类型fProtocolName（这个值在sdp中的“m=”）来判断是基于udp还是rtp，我们只分析RTP，如果是RTP，则根据相应的编码类型fCodecName（这个值在sdp中的“a=rtpmap:”）来判断相应的fRTPSource，这里我们创建了H264和PCMU的RTPSource实例fRTPSource；创建RTCPInstance类的实例fRTCPInstance。
setupMediaSubsession()函数，主要是发送“SETUP”请求，通过SDP的描述，知道我们采用的是RTP协议，根据rtspOpen传入的参数streamUsingTCP来请求rtp是基于udp传输，还是tcp传输，如果是TCP传输，只能是单播，如果udp传输，则根据connectionEndpointName和传入的参数forceMulticastOnUnspecified来判断是否多播还是单播，我们的服务端值支持单播，而且传入的参数false，所以这里采用单播；组包发送“SETUP”请求：
SETUP rtsp://192.168.103.51:8552/h264_ch2/track1 RTSP/1.0
CSeq:
Transport: RTP/AVP;unicast;client_port=-
User-Agent: LibRTSP4 (LIVE555 Streaming Media v2008.04.02)
服务器应答：
RTSP/1.0 OK
CSeq:
Date: Mon, May :: GMT
Transport: RTP/AVP;unicast;destination=14.214.248.17;source=192.168.103.51;client_port=-;server_port=-
Session:
最后，如果采用TCP传输，则调用setStreamSocket()->RTPInterface::setStreamSocket()->addStreamSocket()函数将RTSP的socket值fInputSocketNum加入到fTCPStreams链表中；如果是UDP传输的话，组播地址为空，则用服务端地址保存到fDests中，如果组播地址不为空，则加入组播组。
... ...
if (streamUsingTCP) {
// Tell the subsession to receive RTP (and send/receive RTCP)
// over the RTSP stream:
if (subsession.rtpSource() != NULL)
subsession.rtpSource()->setStreamSocket(fInputSocketNum, subsession.rtpChannelId);
if (subsession.rtcpInstance() != NULL)
subsession.rtcpInstance()->setStreamSocket(fInputSocketNum, subsession.rtcpChannelId);
} else {
// Normal case.
// Set the RTP and RTCP sockets' destination address and port
// from the information in the SETUP response:
subsession.setDestinations(fServerAddress);
}
... ...
5> play
static int Play( rtsp_object_t *p_demux )
{
... ...
if( p_sys->rtsp )
{
/* The PLAY */
if( !p_sys->rtsp->playMediaSession( *p_sys->ms, p_sys->i_npt_start, -, ) )
{
TRACE1_DEC( "RTSP PLAY failed %s\n", p_sys->env->getResultMsg() );
return RTSP_ERROR;;
}
}
... ...return RTSP_SUCCESS;
}
playMediaSession()函数，就是发送“PLAY”请求：
PLAY rtsp://120.90.0.50:8552/h264_ch2/ RTSP/1.0
CSeq:
Session:
Range: npt=0.000-
User-Agent: LibRTSP4 (LIVE555 Streaming Media v2008.04.02)
服务器应答：
RTSP/1.0 OK
CSeq:
Date: Mon, May :: GMT
Range: npt=0.000-
Session:
RTP-Info: url=rtsp://192.168.103.51:8552/h264_ch2/track1;seq=63842;rtptime=1242931431,url=rtsp://192.168.103.51:8552/h264_ch2/track2;seq=432;rtptime=3179210581
接着我们分析rtspRead过程：
int rtspRead(rtsp_object_t *p_obj)
{
... ...
if(p_sys != NULL)
{
/* First warn we want to read data */
p_sys->event = ;
for( i = ; i < p_sys->i_track; i++ )
{
live_track_t *tk = p_sys->track[i];if( tk->waiting == )
{
tk->waiting = ;
tk->sub->readSource()->getNextFrame( tk->p_buffer, tk->i_buffer,
StreamRead, tk, StreamClose, tk );
}
} /* Create a task that will be called if we wait more than 300ms */
task = p_sys->scheduler->scheduleDelayedTask( , TaskInterrupt, p_obj ); /* Do the read */
p_sys->scheduler->doEventLoop( &p_sys->event ); /* remove the task */
p_sys->scheduler->unscheduleDelayedTask( task ); p_sys->b_error ? ret = RTSP_ERROR : ret = RTSP_SUCCESS;
} return ret;
}
这个函数首先要知道readSource()函数的fReadSource的值在哪里复制，在前面的initiate()函数里面有：
... ...
} else if (strcmp(fCodecName, "H264") == ) {
fReadSource = fRTPSource = H264VideoRTPSource::createNew(env(), fRTPSocket,
fRTPPayloadFormat,
fRTPTimestampFrequency);
} else if (strcmp(fCodecName, "JPEG") == ) { // motion JPEG
... ...
} else if ( strcmp(fCodecName, "PCMU") == // PCM u-law audio
|| strcmp(fCodecName, "GSM") == // GSM audio
|| strcmp(fCodecName, "PCMA") == // PCM a-law audio
|| strcmp(fCodecName, "L16") == // 16-bit linear audio
|| strcmp(fCodecName, "MP1S") == // MPEG-1 System Stream
|| strcmp(fCodecName, "MP2P") == // MPEG-2 Program Stream
|| strcmp(fCodecName, "L8") == // 8-bit linear audio
|| strcmp(fCodecName, "G726-16") == // G.726, 16 kbps
|| strcmp(fCodecName, "G726-24") == // G.726, 24 kbps
|| strcmp(fCodecName, "G726-32") == // G.726, 32 kbps
|| strcmp(fCodecName, "G726-40") == // G.726, 40 kbps
|| strcmp(fCodecName, "SPEEX") == // SPEEX audio
) {
createSimpleRTPSource = True;
useSpecialRTPoffset = ;
} else if (useSpecialRTPoffset >= ) {
... ...
} if (createSimpleRTPSource) {
char* mimeType = new char[strlen(mediumName()) + strlen(codecName()) + ] ;
sprintf(mimeType, "%s/%s", mediumName(), codecName());
fReadSource = fRTPSource = SimpleRTPSource::createNew(env(), fRTPSocket, fRTPPayloadFormat,
fRTPTimestampFrequency, mimeType,
(unsigned)useSpecialRTPoffset,
doNormalMBitRule);
delete[] mimeType;
}
}
如果是h264编码方式，则getNextFrame函数定义在FramedSource::getNextFrame：
void FramedSource::getNextFrame(unsigned char* to, unsigned maxSize,
afterGettingFunc* afterGettingFunc,
void* afterGettingClientData,
onCloseFunc* onCloseFunc,
void* onCloseClientData)
{
// Make sure we're not already being read:
if (fIsCurrentlyAwaitingData) {
envir() << "FramedSource[" << this << "]::getNextFrame(): attempting to read more than once at the same time!\n";
exit();
} fTo = to;
fMaxSize = maxSize;
fNumTruncatedBytes = ; // by default; could be changed by doGetNextFrame()
fDurationInMicroseconds = ; // by default; could be changed by doGetNextFrame()
fAfterGettingFunc = afterGettingFunc;
fAfterGettingClientData = afterGettingClientData;
fOnCloseFunc = onCloseFunc;
fOnCloseClientData = onCloseClientData;
fIsCurrentlyAwaitingData = True; doGetNextFrame();
}
doGetNextFrame()函数定义在MultiFramedRTPSource::doGetNextFrame()：
void MultiFramedRTPSource::doGetNextFrame()
{
if (!fAreDoingNetworkReads) {
// Turn on background read handling of incoming packets:
fAreDoingNetworkReads = True;
TaskScheduler::BackgroundHandlerProc* handler = (TaskScheduler::BackgroundHandlerProc*)&networkReadHandler;
fRTPInterface.startNetworkReading(handler);
} fSavedTo = fTo;
fSavedMaxSize = fMaxSize;
fFrameSize = ; // for now
fNeedDelivery = True; doGetNextFrame1();
}
doGetNextFrame1()函数定义在MultiFramedRTPSource::doGetNextFrame1()：
void MultiFramedRTPSource::doGetNextFrame1()
{
while (fNeedDelivery) {
// If we already have packet data available, then deliver it now.
Boolean packetLossPrecededThis;
BufferedPacket* nextPacket = fReorderingBuffer->getNextCompletedPacket(packetLossPrecededThis);
if (nextPacket == NULL) break; fNeedDelivery = False; if (nextPacket->useCount() == ) {
// Before using the packet, check whether it has a special header
// that needs to be processed:
unsigned specialHeaderSize;
if (!processSpecialHeader(nextPacket, specialHeaderSize)) {
// Something's wrong with the header; reject the packet:
fReorderingBuffer->releaseUsedPacket(nextPacket);
fNeedDelivery = True;
break;
}
nextPacket->skip(specialHeaderSize);
} // Check whether we're part of a multi-packet frame, and whether
// there was packet loss that would render this packet unusable:
if (fCurrentPacketBeginsFrame) {
if (packetLossPrecededThis || fPacketLossInFragmentedFrame) {
// We didn't get all of the previous frame.
// Forget any data that we used from it:
fTo = fSavedTo; fMaxSize = fSavedMaxSize;
fFrameSize = ;
}
fPacketLossInFragmentedFrame = False;
} else if (packetLossPrecededThis) {
// We're in a multi-packet frame, with preceding packet loss
fPacketLossInFragmentedFrame = True;
}
if (fPacketLossInFragmentedFrame) {
// This packet is unusable; reject it:
fReorderingBuffer->releaseUsedPacket(nextPacket);
fNeedDelivery = True;
break;
} // The packet is usable. Deliver all or part of it to our caller:
unsigned frameSize;
nextPacket->use(fTo, fMaxSize, frameSize, fNumTruncatedBytes,
fCurPacketRTPSeqNum, fCurPacketRTPTimestamp,
fPresentationTime, fCurPacketHasBeenSynchronizedUsingRTCP,
fCurPacketMarkerBit);
fFrameSize += frameSize; if (!nextPacket->hasUsableData()) {
// We're completely done with this packet now
fReorderingBuffer->releaseUsedPacket(nextPacket);
} if (fCurrentPacketCompletesFrame || fNumTruncatedBytes > ) {
// We have all the data that the client wants.
if (fNumTruncatedBytes > ) {
envir() << "MultiFramedRTPSource::doGetNextFrame1(): The total received frame size exceeds the client's buffer size ("
<< fSavedMaxSize << "). "<< fNumTruncatedBytes << " bytes of trailing data will be dropped!\n";
}
// Call our own 'after getting' function, so that the downstream object can consume the data:
if (fReorderingBuffer->isEmpty()) {
// Common case optimization: There are no more queued incoming packets, so this code will not get
// executed again without having first returned to the event loop. Call our 'after getting' function
// directly, because there's no risk of a long chain of recursion (and thus stack overflow):
afterGetting(this);
} else {
// Special case: Call our 'after getting' function via the event loop.
nextTask() = envir().taskScheduler().scheduleDelayedTask(, (TaskFunc*)FramedSource::afterGetting, this);
}
} else {
// This packet contained fragmented data, and does not complete
// the data that the client wants. Keep getting data:
fTo += frameSize; fMaxSize -= frameSize;
fNeedDelivery = True;
}
}
}
FramedSource::afterGetting(FramedSource* source) ：
void FramedSource::afterGetting(FramedSource* source)
{
source->fIsCurrentlyAwaitingData = False;
// indicates that we can be read again
// Note that this needs to be done here, in case the "fAfterFunc"
// called below tries to read another frame (which it usually will) if (source->fAfterGettingFunc != NULL) {
(*(source->fAfterGettingFunc))(source->fAfterGettingClientData,
source->fFrameSize,
source->fNumTruncatedBytes,
source->fPresentationTime,
source->fDurationInMicroseconds);
}
}
fAfterGettingFunc函数指针在FramedSource::getNextFrame()中被赋值afterGettingFunc，afterGettingFunc的值则是rtspRead()函数调用getNextFrame()函数时，传入的StreamRead()。这样就获取了一帧数据。
在MultiFramedRTPSource::doGetNextFrame()函数中，我们发现了fRTPInterface.startNetworkReading(handler)，这个函数主要做了什么作用？
void RTPInterface::startNetworkReading(TaskScheduler::BackgroundHandlerProc* handlerProc)
{
// Normal case: Arrange to read UDP packets:
envir().taskScheduler().turnOnBackgroundReadHandling(fGS->socketNum(), handlerProc, fOwner); // Also, receive RTP over TCP, on each of our TCP connections:
fReadHandlerProc = handlerProc;
for (tcpStreamRecord* streams = fTCPStreams; streams != NULL; streams = streams->fNext) {
// Get a socket descriptor for "streams->fStreamSocketNum":
SocketDescriptor* socketDescriptor = lookupSocketDescriptor(envir(), streams->fStreamSocketNum);
if (socketDescriptor == NULL) {
socketDescriptor = new SocketDescriptor(envir(), streams->fStreamSocketNum);
socketHashTable(envir())->Add((char const*)(long)(streams->fStreamSocketNum), socketDescriptor);
} // Tell it about our subChannel:
socketDescriptor->registerRTPInterface(streams->fStreamChannelId, this);
}
}
这个函数主要做了两个作用，一个是注册UDP socket的读取任务函数MultiFramedRTPSource::networkReadHandler()到任务队列，一个是注册TCP socket的读取任务函数SocketDescriptor::tcpReadHandler()到任务队列，最终还是会调用MultiFramedRTPSource::networkReadHandler()函数获取一帧数据。