一个简单的MicrosoftSpeechSDK5.1例子.doc

一个简单的 Microsoft Speech SDK 5.1 例子 刚刚从 Microsoft 网站下栽了 Speech SDK 5.1 和中日文发声补丁，参考 Speech SDK文档写了几个小程序，尽管我也是刚刚接触 Speech若干小时的初学者，还是拿出来共享吧： Speech SDK 采用 COM 的形式封装，和大多数COM对象一样，下面是一个似曾相识的初始化： ???? CoInitialize(NULL); ???? CLSIDFromProgID(LSAPI.SpVoice, CLSID_SpVoice); ???? ISpVoice * pSpVoice = NULL; ????if (FAILED(CoCreateInstance(CLSID_SpVoice, NULL, ??????? CLSCTX_INPROC_SERVER, IID_ISpVoice, (void **)pSpVoice))) ???? { ??????? cout Failed to create instance of ISpVoice! endl; ??????? return -1; ???? } ????现在我们用它来说 “Hello World”，非常简单： pSpVoice-Speak(LHello World!, SPF_DEFAULT, NULL); 最后，清除这一切 pSpVoice-Release(); CoUninitialize(); 很容易，是不是？下面我们稍微改变一下： IEnumSpObjectTokens *pSpEnumTokens = NULL; if (SUCCEEDED(SpEnumTokens(SPCAT_VOICES, Llanguage=409, NULL, pSpEnumTokens))) { ?????? ISpObjectToken *pSpToken = NULL; ?????? while (SUCCEEDED(pSpEnumTokens-Next(1, pSpToken, NULL)) ?????? ?????? pSpToken != NULL) ?????? { ?????? ?????? pSpVoice-SetVoice(pSpToken); ?????? ?????? pSpVoice-Speak(LHello World!, SPF_DEFAULT, NULL); ?????? ?????? pSpToken-Release(); ?????? } ?????? pSpEnumTokens-Release(); } 如果只需要一种 Voice，SpFindBestToken 更加方便： ISpObjectToken * pSpObjectToken = NULL; if (SUCCEEDED(SpFindBestToken(SPCAT_VOICES, Lgender=female, ?????? NULL, pSpObjectToken))) { ?????? pSpVoice-SetVoice(pSpObjectToken); ?????? pSpVoice-Speak(LHello World!, SPF_DEFAULT, NULL); ?????? pSpObjectToken-Release(); } 现在改为使用中文： ISpObjectToken * pSpObjectToken = NULL; if (SUCCEEDED(SpFindBestToken(SPCAT_VOICES, Llanguage=804, ?????? NULL, pSpObjectToken))) { ?????? pSpVoice-SetVoice(pSpObjectToken); ?????? pSpVoice-Speak(L世界你好！, SPF_DEFAULT, NULL); ?????? pSpObjectToken-Release(); } ISpVoice::Speak 可以允许朗读的句子中包含 XML 标记，其实以上这一切用 XML 标记将变得十分简单： pSpVoice-Speak(Llang langid=\804\世界你好！/lang, SPF_DEFAULT, NULL); ???? pSpVoice-Speak(Lvoice required=\gender=female\Hello World!/voice, SPF_DEFAULT, NULL); 最后，以一段简单的报时结束这篇文章： SYSTEMTIME st; GetLocalTime(st); ISpObjectToken * pSpObjectToken = NULL; if (S