AuidioStreamPlayer简易实现

首先可以先从下面列出的参考中进行对整个Audio Stream进行一定程度的理解

参考：

https://github.com/muhku/FreeStreamer
https://github.com/mattgallagher/AudioStreamer
http://msching.github.io/blog/2014/07/09/audio-in-ios-3/
http://msching.github.io/blog/2014/08/02/audio-in-ios-5/

主要数据结构：

typedef struct AudioStream{
    AudioFileStreamID                       mAudioFileStream;
    AudioQueueRef                           mQueue;
    AudioQueueBufferRef                     mBuffer[kNumbersOfBuffer];
    AudioStreamPacketDescription            mPacketDecs[kMaxPacketDescs];
    AudioStreamBasicDescription             mAudioStreamBasicDesc;
    unsigned int                            mFillBufferIndex;
    UInt32                                  mPacketBufferSize;
    size_t                                  mBytesFilled;
    size_t                                  mPacketsFilled;
    bool                                    mInused[kNumbersOfBuffer];
    NSInteger                               mBuffersUsed;

}AS;

typedef struct StreamFile{
    UInt32                                  mBitRate;
    NSInteger                               mFileLength;
    NSInteger                               mDataOffset;
    UInt64                                  mAudioDataByteCount;

}SF;
extern NSString * const ASStatusChangedNotification;
@interface FileStreamPlayer : NSObject{
    AS                                      audioStream;
    BOOL                                    discontinuous;
    NSThread                                *internalThread;
    NSURL                                   *url;
    NSDictionary                            *httpHeader;
    NSString                                *fileExtension;
    SF                                      streamFile;
    double                                  sampleRate;
    double                                  packetDuration;
    CFReadStreamRef                         stream;
    pthread_mutex_t                         queueBuffersMutex;
    pthread_cond_t                          queueBufferReadyCondition;
    AudioStreamState                        state;
    AudioStreamStopReason                   stopReason;
    NSNotificationCenter                    *notificationCenter;
}

首先，要实现AudioStream的获取，通过CFNetWorking的库来发送HTTP请求，并通过建立客户端接受数据将数据加入到当前RunLoop中：

CFHTTPMessageRef message= CFHTTPMessageCreateRequest(NULL,
                                                             (CFStringRef)@"GET",
                                                             (CFURLRef)url,
                                                             kCFHTTPVersion1_1);
        stream = CFReadStreamCreateForHTTPRequest(NULL, message);
        CFRelease(message);

        CFDictionaryRef proxySettings = CFNetworkCopySystemProxySettings();
        CFReadStreamSetProperty(stream, kCFStreamPropertyHTTPProxy, proxySettings);
        CFRelease(proxySettings);
        CFStreamClientContext context = {0, self, NULL, NULL, NULL};
        CFReadStreamSetClient(
                              stream,
                              kCFStreamEventHasBytesAvailable |
                              kCFStreamEventErrorOccurred |
                              kCFStreamEventEndEncountered,
                              ASReadStreamCallBack,
                              &context);
        CFReadStreamScheduleWithRunLoop(stream, CFRunLoopGetCurrent(), kCFRunLoopCommonModes);

这里会要求实现网络流数据获取后的回调函数：ReadStreamCallBack

static void ASReadStreamCallBack(CFReadStreamRef                aStream,
                                 CFStreamEventType              eventType,
                                 void*                          inClientInfo) {

    FileStreamPlayer* this = (FileStreamPlayer *)inClientInfo;

    [this handleReadFromStream:aStream eventType:eventType];
}

函数参数中要求：当前数据流，流事件类型，以及上下文对象

成功读取数据以后，通过：AudioFileStreamOpen将数据打开，可以理解成AudioFile读取时候打开一个数据流，然后进行解析

    AudioFileStreamOpen
    AudioFileStreamParseBytes

OpenStream中要求填入两个回调函数：PropertyListenerProc和ASPacketsProc，分别作为流信息和流解析分离帧的回调之后的分离帧解析中，每次都会回调刚刚填入的回调函数

static void ASPropertyListenerProc(//callback from audioStream header
                                   void *                           inClientData,
                                   AudioFileStreamID                inAudioFileStream,
                                   AudioFileStreamPropertyID        inPropertyID,
                                   UInt32 *                         ioFlags) {

    FileStreamPlayer* this = (FileStreamPlayer *)inClientData;

    [this handleFileStreamPropertyChange:inAudioFileStream
                        fileStreamPropertyID:inPropertyID
                                     ioFlags:ioFlags];
}


static void ASPacketsProc(//callback from audioStream body
                          void *                            inClientData,
                          UInt32                            inNumberBytes,
                          UInt32                            inNumberPackets,
                          const void *                      inInputData,
                          AudioStreamPacketDescription      *inPacketDescriptions) {

    FileStreamPlayer* this = (FileStreamPlayer *)inClientData;

    [this handleAudioPackets:inInputData
                 numberBytes:inNumberBytes
               numberPackets:inNumberPackets
          packetDescriptions:inPacketDescriptions];
}

接下来的操作就是包括流信息解析后存入相应的数据结构中，利用Buffer存储塞到AudioQueue中去进行运行，设置相应的AudioQueue的回调函数对完成任务后的buffer进行再次调用：

AudioQueueNewOutput(&audioStream.mAudioStreamBasicDesc,
                        ASAudioQueueOutputCallback,
                        self, NULL, NULL, 0, &audioStream.mQueue);
static void ASAudioQueueOutputCallback(//callBack from Running Queue Buffers completed
                                       void*                    inClientData,
                                       AudioQueueRef            inAQ,
                                       AudioQueueBufferRef      inBuffer) {

    FileStreamPlayer* this = (FileStreamPlayer*)inClientData;

    [this handleCompleteBuffer:inAQ
                        buffer:inBuffer];
}

接下来就可以对AudioQueue进行各种操作了，播放暂停和停止等。另外，还要注意各个操作过程中线程资源之间的调用必须要用锁进行保护，在这里我主要参考了AudioStreamer，所以采用了互斥锁pthread_和@synchronized（self）。

最后，附上我的DEMO：

http://download.youkuaiyun.com/detail/atomic123/9323745