基于live555的RtspClient

术业专攻


基于live555的RtspClient

前言

很早之前做的一个监控客户端使用的一个庞大的SDK，现在把RtspClient这端独立出来。后面可能还会写一些列关于音视频相关的Blog，包括视频解码、视频显示、在YUV数据中叠加图片等等。

音视频系列

正文

开发环境

  • Visual Studio 2013
  • Windows 10

开发工具

  • Live555源码

开发过程

一、编译Live555库

使用vs2013编译live555库网上应该有一大把资料。
参考这个Live555编译,
使用方法二，编译出来4个静态库：

  • BasicUsageEnvironmentlib.lib
  • groupsocklib.lib
  • liveMedialib.lib
  • UsageEnvironmentlib.lib

二、开发RtspClient项目

在编译好了live555的库后，用vs创建项目引入静态库，这里就不详细的介绍了，主要来讲讲RtspClient的代码。
live555的源码里面，testProgs\testRTSPClient.cpp 和 testProgs\openRTSP.cpp 详细的介绍了具体流程。
那我们慢慢根据代码来慢慢探索live555 RtspClient 整个连接过程。

创建RTSPClient对象。

1
2
3
m_pScheduler = BasicTaskScheduler::createNew();
m_pEnv = BasicUsageEnvironment::createNew(*m_pScheduler);
m_pRtspClient = RTSPClientEx::createNew(*m_pEnv, m_szUrl.c_str(), 0, "RTSP Client");

首先是创建TaskScheduler对象和UsageEnvironment对象,然后根据RTSP地址m_szUrl 创建一个RTSPClient对象，一个RTSPClient对象代表一个RTSP客户端。

发送DESCRIBE命令

1
2
3
4
5
6
7
Authenticator auth;
auth.setUsernameAndPassword(m_szAccount.c_str(), m_szPassword.c_str());
m_pRtspClient->sendDescribeCommand(continueAfterDESCRIBE, &auth);
if (!wait_Live555_response(wait_live555_time))
{
    continue;
}

调用sendDescribeCommand函数发送DESCRIBE命令，回调函数是continueAfterDESCRIBE函数，在收到RTSP服务器端对DESCRIBE命令的回复时调用。大家看到这里可能发现下面的wait_Live555_response怎么和testRTSPClient.cpp里面的不同了，这里我是参考了vlc代码里面的实现。这在里调用wait_Live555_response堵塞线程等待continueAfterDESCRIBE返回结果才执行下一步。wait_Live555_response这个函数会在接下来的代码中发挥同样的作用 wait_live555_time 是超时时间。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
static void TaskInterruptRTSP(void *p_private)
{
	CRtspPullClass *pPullClass = (CRtspPullClass*)p_private;
	pPullClass->m_event_rtsp = 0xff;
}

bool CRtspPullClass::wait_Live555_response(int i_timeout)
{
	TaskToken task;
	m_event_rtsp = 0;
	if (i_timeout > 0)
	{
		task = m_pScheduler->scheduleDelayedTask(i_timeout * 1000,
			TaskInterruptRTSP,
			this);
	}
	m_event_rtsp = 0;
	m_b_error = true;
	m_ilive555_ret = 0;
	m_pScheduler->doEventLoop(&m_event_rtsp);
	if (i_timeout > 0)
		m_pScheduler->unscheduleDelayedTask(task);
	return !m_b_error;
}

上面是wait_Live555_response()的具体实现,先通过scheduleDelayedTask设置一个超时回调，然后doEventLoop堵塞Live555的线程，这里m_event_rtsp = 0的时候调用doEventLoop就是堵塞，如果当m_event_rtsp = 1的时候这里就继续往下执行。

然后让我们来看看continueAfterDESCRIBE()回调里的实现

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
static void continueAfterDESCRIBE(RTSPClient* rtspClient, int resultCode, char* resultString)
{
	CRtspPullClass* pPullClass = ((RTSPClientEx*)rtspClient)->GetPullClass();
	UsageEnvironment& env = rtspClient->envir();
	pPullClass->m_ilive555_ret = resultCode;
	if (resultCode == 0)
	{
		char* sdpDescription = resultString;
		free(pPullClass->m_p_sdp);
		pPullClass->m_p_sdp = NULL;
		if (sdpDescription)
		{
			pPullClass->m_p_sdp = strdup(sdpDescription);
			pPullClass->m_b_error = false;
		}
		else
			pPullClass->m_b_error = true;
	}
	else
		pPullClass->m_b_error = true;
	delete[] resultString;
	pPullClass->m_event_rtsp = 1;
}

如果回调返回成功的话就会获取到sdp信息。然后我们就会更具sdp信息创建一个MediaSession对象，MediaSession表示客户端请求服务器端某个媒体资源的会话。MediaSubsession，表示MediaSession的子会话，创建MediaSession的同时也创建了包含的MediaSubsession对象。然后客户端对服务器端的每个ServerMediaSubsession发送SETUP命令请求建立连接。让我们看接下来创建MediaSession的代码

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
typedef struct _VIDEO_PARAM
{
	char codec[256];
	int width;
	int height;
	int colorbits;
	int framerate;
	int bitrate;
	char vol_data[256];
	int vol_length;
}VIDEO_PARAM;



typedef struct  _AUDIO_PARAM
{
	char codec[256];
	int samplerate;
	int bitspersample;
	int channels;
	int framerate;
	int bitrate;
}AUDIO_PARAM;


typedef struct  __STREAM_AV_PARAM
{
	unsigned char	ProtocolName[32];
	short  bHaveVideo;//0 表示没有视频参数
	short  bHaveAudio;//0 表示没有音频参数
	VIDEO_PARAM videoParam;//视频参数
	AUDIO_PARAM audioParam;//音频参数
	char		szUrlInfo[512];//注意长度
}STREAM_AV_PARAM;

if (!(m_pSession = MediaSession::createNew(*m_pEnv, m_p_sdp)))
{
	continue;
}
if (m_pSession->hasSubsessions() ==false)
{
	continue;
}
m_dwLastRecvDataTick = GetTickCount();
STREAM_AV_PARAM * pAvParam = &m_avParam;
memset(pAvParam, 0, sizeof(STREAM_AV_PARAM));
memset(&m_AudioParam, 0, sizeof(AUDIO_PARAM));
memset(&m_VideoParam, 0, sizeof(VIDEO_PARAM));
sprintf((char *)pAvParam->ProtocolName, "%s", "AjVisionHD");
MediaSubsession* subsession = NULL;
MediaSubsessionIterator iter(*m_pSession);

STREAM_AV_PARAM 是一个包含媒体流参数的结构体，因为在解码的时候需要知道音视频的类型是H.265 还是H.264 是AAC还是G.711。音频的采样率、码率、通道数。视频的长宽、码率、和帧率。
取得MediaSubsession，然后客户端对服务器端的每个MediaSubsessionn发送SETUP命令请求建立连接。

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
while ((subsession = iter.next()) != NULL)
{
	if (subsession->initiate() == false)
	{
		continue;
	}
	if (subsession->rtpSource() != NULL)
	{
		desiredPortNum = subsession->clientPortNum();
		if (desiredPortNum != 0)
		{
			subsession->setClientPortNum(desiredPortNum);
			desiredPortNum += 2;
		}
		subsession->rtpSource()->setPacketReorderingThresholdTime(200000);
		if (strcmp(subsession->mediumName(), "video") == 0)
		{
			int nIsH264 = -1;
			if ((strcmp(subsession->codecName(), "h264") == 0)
				|| (strcmp(subsession->codecName(), "H264") == 0))
				nIsH264 = 1;
			else if ((strcmp(subsession->codecName(), "h265") == 0)
				|| (strcmp(subsession->codecName(), "H265") == 0))
				nIsH264 = 0;
			if (nIsH264 == -1)
				continue;
			VIDEO_PARAM *pVideoParam = &m_VideoParam;
			memset(pVideoParam, 0, sizeof(VIDEO_PARAM));
			strcpy(pVideoParam->codec, subsession->codecName());
			pVideoParam->width = subsession->videoWidth();
			pVideoParam->height = subsession->videoHeight();
			pVideoParam->framerate = subsession->videoFPS();
			pVideoParam->bitrate = 0;
			if (subsession->fmtp_config() != NULL)
			{
				unsigned configLen;
				unsigned char* configData = parseGeneralConfigStr(subsession->fmtp_config(), configLen);
				if (configData != NULL)
				{
					memcpy(pVideoParam->vol_data, configData, configLen);
					pVideoParam->vol_length = configLen;
					delete[] configData; configData = NULL;
				}
			}
			pAvParam->bHaveVideo = 1;
			memcpy(&pAvParam->videoParam, pVideoParam, sizeof(VIDEO_PARAM));
			if (m_pVideoSink)
			{
				CVideoStreamSink::close(m_pVideoSink);
				m_pVideoSink = NULL;
			}
			try
			{
				m_pVideoSink = new CVideoStreamSink(*m_pEnv, nIsH264, this, (LONG)this, 4 * 1024 * 1024);
			}
			catch (...)
			{
				m_pVideoSink = NULL;
				continue;
			}
			if (m_pVideoSink)
			{
				m_pVideoSink->SetStreamName("");
				if ( m_nLinkMode== 2)
					m_pRtspClient->sendSetupCommand(*subsession, default_live555_callback, false, false, true);
				else
					m_pRtspClient->sendSetupCommand(*subsession, default_live555_callback, false, m_nLinkMode, false);
				m_dwLastRecvDataTick = GetTickCount();
				if (!IsRunning())
					break;
				if (!wait_Live555_response(wait_live555_time))
				{
					continue;
				}
				m_dwLastRecvDataTick = GetTickCount();
				if (m_nLinkMode == 2)
				{//组播
					subsession->setDestinations(subsession->connectionEndpointAddress());
				}
				subsession->miscPtr = m_pRtspClient;
				m_pVideoSink->startPlaying(*(subsession->readSource()), subsessionAfterPlaying, subsession);
				if (subsession->rtcpInstance() != NULL)
			    	subsession->rtcpInstance()->setByeHandler(subsessionByeHandler, subsession);
				subsession->sink = m_pVideoSink;
			}

		}
		else if (strcmp(subsession->mediumName(), "audio") == 0)
		{
			AUDIO_PARAM *pAudioParam = &m_AudioParam;
			memset(pAudioParam, 0, sizeof(AUDIO_PARAM));
			strcpy(pAudioParam->codec, subsession->codecName());
			DebugLog("CRtspPullClass::RtspLinkThread() %s recv audio type = %s", m_szUrl.c_str(), subsession->codecName());
			if (strstr(pAudioParam->codec, "PCMU") || strstr(pAudioParam->codec, "PCMA")) // for G711
			{
				pAudioParam->bitrate = 64000;
				pAudioParam->bitspersample = 16;
				pAudioParam->channels = 1;//subsession->numChannels();
				pAudioParam->framerate = 8;
				pAudioParam->samplerate = subsession->rtpTimestampFrequency();

			}
			else //for AAC
			{
				unsigned configLen;
				unsigned char* configData = parseGeneralConfigStr(subsession->fmtp_config(), configLen);
				if (configData != NULL)
				{
					unsigned char samplingFrequencyIndex = (configData[0] & 0x07) << 1;
					samplingFrequencyIndex |= (configData[1] & 0x80) >> 7;
					if (samplingFrequencyIndex > 15)
					{
						samplingFrequencyIndex = 0;
					}

					pAudioParam->bitspersample = 16;
					pAudioParam->samplerate = samplingFrequencyTable[samplingFrequencyIndex];
					pAudioParam->channels = 2;
					pAudioParam->bitrate = pAudioParam->samplerate * pAudioParam->bitspersample / 8;
					pAudioParam->framerate = 8;
					delete[] configData; configData = NULL;
				}
				else
				{
					pAudioParam->bitrate = 16000;
					pAudioParam->bitspersample = 16;
					pAudioParam->channels = 2;
					pAudioParam->framerate = 16;
					pAudioParam->samplerate = 16000;
				}
			}
			pAvParam->bHaveAudio = 1;
			memcpy(&pAvParam->audioParam, pAudioParam, sizeof(AUDIO_PARAM));
			if (m_pAudioSink)
			{
				CAudioStreamSink::close(m_pAudioSink);
				m_pAudioSink = NULL;
			}
			try
			{
				m_pAudioSink = new CAudioStreamSink(*m_pEnv, this, (LONG)this, 64 * 1024);
			}
			catch (...)
			{
				CAudioStreamSink::close(m_pAudioSink);
				m_pAudioSink = NULL;
				continue;
			}
			if (m_pAudioSink)
			{
				if (m_nLinkMode == 2)
				{
					m_pRtspClient->sendSetupCommand(*subsession, default_live555_callback, false, false, true);//,TRUE);
				}
				else
					m_pRtspClient->sendSetupCommand(*subsession, default_live555_callback, false, m_nLinkMode);//,TRUE);			
				m_dwLastRecvDataTick = GetTickCount();
				if (!IsRunning())
					break;
				if (!wait_Live555_response(wait_live555_time))
				{
					if (m_pAudioSink)
					{
						CAudioStreamSink::close(m_pAudioSink);
						m_pAudioSink = NULL;
					}
					subsession->sink = NULL;
					continue;
				}
				if (m_nLinkMode == 2)
				{//组播
					subsession->setDestinations(subsession->connectionEndpointAddress());
				}
				subsession->miscPtr = m_pRtspClient;
				m_pAudioSink->startPlaying(*(subsession->readSource()), subsessionAfterPlaying, subsession);
				if (subsession->rtcpInstance() != NULL)
					subsession->rtcpInstance()->setByeHandler(subsessionByeHandler, subsession);
				subsession->sink = m_pAudioSink;
				m_dwLastRecvDataTick = GetTickCount();
			}
		}
	}
}
if (m_pVideoSink || m_pAudioSink)
{
	m_pRtspClient->sendPlayCommand(*m_pSession, default_live555_callback, 0, -1, 1);
	if (!wait_Live555_response(wait_live555_time))
	{
		continue;
	}
	m_isStoping = 0;
	m_pScheduler->doEventLoop(&m_isStoping);
}

在上面的代码中，首先调用MediaSubsession的initiate函数初始化MediaSubsession，然后通过MediaSubsession获取音视频的参数填充结构体。最后为MediaSubsession创建MediaSink对象来请求和保存服务器端发送的数据。然后对MediaSubsession发送SETUP命令。同样调用wait_Live555_response等待回调返回数据，然后调用MediaSink::startPlaying函数开始准备播放对应的MediaSubsession。如果建立连接成功了最后会发送Play命令请求开始传送数据同样等到回复后调用m_pScheduler->doEventLoop(&m_isStoping)堵塞线程;

下面的代码是MediaSink里面的实现了

1
2
3
4
5
6
7
8
9
10


void CMediaSinkSink::afterGettingFrame(void* clientData, unsigned frameSize,
 unsigned /*numTruncatedBytes*/,
 struct timeval presentationTime,
 unsigned /*durationInMicroseconds*/) 
{
	CVideoStreamSink* sink = (CVideoStreamSink*)clientData;	
	fSource->getNextFrame(fBuffer, fBufferSize,afterGettingFrame, this,onSourceClosure, this);
}

在DummySink::afterGettingFrame函数中只是简单地打印出了某个MediaSubsession接收到了多少字节的数据，然后接着利用FramedSource去读取数据。可以看出，在RTSP客户端，Live555也是在MediaSink和FramedSource之间形成了一个循环，不停地从服务器端读取数据。

参考文献

最后的话