Showing the changes to interface/src/scripting/TTSScriptingInterface.cpp (1 of 5 files). Show all
Unified · Split
| @@ -0,0 +1,163 @@ | |||
| 1 | // | ||
| 2 | // TTSScriptingInterface.cpp | ||
| 3 | // libraries/audio-client/src/scripting | ||
| 4 | // | ||
| 5 | // Created by Zach Fox on 2018-10-10. | ||
| 6 | // Copyright 2018 High Fidelity, Inc. | ||
| 7 | // | ||
| 8 | // Distributed under the Apache License, Version 2.0. | ||
| 9 | // See the accompanying file LICENSE or http://www.apache.org/licenses/LICENSE-2.0.html | ||
| 10 | // | ||
| 11 | |||
| 12 | #include "TTSScriptingInterface.h" | ||
| 13 | #include "avatar/AvatarManager.h" | ||
| 14 | |||
| 15 | TTSScriptingInterface::TTSScriptingInterface() { | ||
| 16 | // | ||
| 17 | // Create text to speech engine | ||
| 18 | // | ||
| 19 | HRESULT hr = m_tts.CoCreateInstance(CLSID_SpVoice); | ||
| 20 | if (FAILED(hr)) { | ||
| 21 | qDebug() << "Text-to-speech engine creation failed."; | ||
| 22 | } | ||
| 23 | |||
| 24 | // | ||
| 25 | // Get token corresponding to default voice | ||
| 26 | // | ||
| 27 | hr = SpGetDefaultTokenFromCategoryId(SPCAT_VOICES, &m_voiceToken, FALSE); | ||
| 28 | if (FAILED(hr)) { | ||
| 29 | qDebug() << "Can't get default voice token."; | ||
| 30 | } | ||
| 31 | |||
| 32 | // | ||
| 33 | // Set default voice | ||
| 34 | // | ||
| 35 | hr = m_tts->SetVoice(m_voiceToken); | ||
| 36 | if (FAILED(hr)) { | ||
| 37 | qDebug() << "Can't set default voice."; | ||
| 38 | } | ||
| 39 | } | ||
| 40 | |||
| 41 | TTSScriptingInterface::~TTSScriptingInterface() { | ||
| 42 | } | ||
| 43 | |||
| 44 | class ReleaseOnExit { | ||
| 45 | public: | ||
| 46 | ReleaseOnExit(IUnknown* p) : m_p(p) {} | ||
| 47 | ~ReleaseOnExit() { | ||
| 48 | if (m_p) { | ||
| 49 | m_p->Release(); | ||
| 50 | } | ||
| 51 | } | ||
| 52 | |||
| 53 | private: | ||
| 54 | IUnknown* m_p; | ||
| 55 | }; | ||
| 56 | |||
| 57 | void TTSScriptingInterface::testTone(const bool& alsoInject) { | ||
| 58 | QByteArray byteArray(480000, 0); | ||
| 59 | _lastSoundByteArray.resize(0); | ||
| 60 | _lastSoundByteArray.resize(480000); | ||
| 61 | |||
| 62 | int32_t a = 0; | ||
| 63 | int16_t* samples = reinterpret_cast<int16_t*>(byteArray.data()); | ||
| 64 | for (a = 0; a < 240000; a++) { | ||
| 65 | int16_t temp = (glm::sin(glm::radians((float)a))) * 32768; | ||
| 66 | samples[a] = temp; | ||
| 67 | } | ||
| 68 | emit ttsSampleCreated(_lastSoundByteArray); | ||
| 69 | |||
| 70 | if (alsoInject) { | ||
| 71 | AudioInjectorOptions options; | ||
| 72 | options.position = DependencyManager::get<AvatarManager>()->getMyAvatarPosition(); | ||
| 73 | |||
| 74 | _lastSoundAudioInjector = AudioInjector::playSound(_lastSoundByteArray, options); | ||
| 75 | } | ||
| 76 | } | ||
| 77 | |||
| 78 | void TTSScriptingInterface::speakText(const QString& textToSpeak, const bool& alsoInject) { | ||
| 79 | WAVEFORMATEX fmt; | ||
| 80 | fmt.wFormatTag = WAVE_FORMAT_PCM; | ||
| 81 | fmt.nSamplesPerSec = 24000; | ||
| 82 | fmt.wBitsPerSample = 16; | ||
| 83 | fmt.nChannels = 1; | ||
| 84 | fmt.nBlockAlign = fmt.nChannels * fmt.wBitsPerSample / 8; | ||
| 85 | fmt.nAvgBytesPerSec = fmt.nSamplesPerSec * fmt.nBlockAlign; | ||
| 86 | fmt.cbSize = 0; | ||
| 87 | |||
| 88 | IStream* pStream = NULL; | ||
| 89 | |||
| 90 | ISpStream* pSpStream = nullptr; | ||
| 91 | HRESULT hr = CoCreateInstance(CLSID_SpStream, nullptr, CLSCTX_ALL, __uuidof(ISpStream), (void**)&pSpStream); | ||
| 92 | if (FAILED(hr)) { | ||
| 93 | qDebug() << "CoCreateInstance failed."; | ||
| 94 | } | ||
| 95 | ReleaseOnExit rSpStream(pSpStream); | ||
| 96 | |||
| 97 | pStream = SHCreateMemStream(NULL, 0); | ||
| 98 | if (nullptr == pStream) { | ||
| 99 | qDebug() << "SHCreateMemStream failed."; | ||
| 100 | } | ||
| 101 | |||
| 102 | hr = pSpStream->SetBaseStream(pStream, SPDFID_WaveFormatEx, &fmt); | ||
| 103 | if (FAILED(hr)) { | ||
| 104 | qDebug() << "Can't set base stream."; | ||
| 105 | } | ||
| 106 | |||
| 107 | hr = m_tts->SetOutput(pSpStream, true); | ||
| 108 | if (FAILED(hr)) { | ||
| 109 | qDebug() << "Can't set output stream."; | ||
| 110 | } | ||
| 111 | |||
| 112 | ReleaseOnExit rStream(pStream); | ||
| 113 | |||
| 114 | ULONG streamNumber; | ||
| 115 | hr = m_tts->Speak(reinterpret_cast<LPCWSTR>(textToSpeak.utf16()), SPF_IS_XML | SPF_ASYNC | SPF_PURGEBEFORESPEAK, | ||
| 116 | &streamNumber); | ||
| 117 | if (FAILED(hr)) { | ||
| 118 | qDebug() << "Speak failed."; | ||
| 119 | } | ||
| 120 | |||
| 121 | m_tts->WaitUntilDone(-1); | ||
| 122 | |||
| 123 | hr = pSpStream->GetBaseStream(&pStream); | ||
| 124 | if (FAILED(hr)) { | ||
| 125 | qDebug() << "Couldn't get base stream."; | ||
| 126 | } | ||
| 127 | |||
| 128 | hr = IStream_Reset(pStream); | ||
| 129 | if (FAILED(hr)) { | ||
| 130 | qDebug() << "Couldn't reset stream."; | ||
| 131 | } | ||
| 132 | |||
| 133 | ULARGE_INTEGER StreamSize; | ||
| 134 | StreamSize.LowPart = 0; | ||
| 135 | hr = IStream_Size(pStream, &StreamSize); | ||
| 136 | |||
| 137 | DWORD dwSize = StreamSize.QuadPart; | ||
| 138 | char* buf1 = new char[dwSize + 1]; | ||
| 139 | memset(buf1, 0, dwSize + 1); | ||
| 140 | |||
| 141 | hr = IStream_Read(pStream, buf1, dwSize); | ||
| 142 | if (FAILED(hr)) { | ||
| 143 | qDebug() << "Couldn't read from stream."; | ||
| 144 | } | ||
| 145 | |||
| 146 | _lastSoundByteArray.resize(0); | ||
| 147 | _lastSoundByteArray.append(buf1, dwSize); | ||
| 148 | |||
| 149 | emit ttsSampleCreated(_lastSoundByteArray); | ||
| 150 | |||
| 151 | if (alsoInject) { | ||
| 152 | AudioInjectorOptions options; | ||
| 153 | options.position = DependencyManager::get<AvatarManager>()->getMyAvatarPosition(); | ||
| 154 | |||
| 155 | _lastSoundAudioInjector = AudioInjector::playSound(_lastSoundByteArray, options); | ||
| 156 | } | ||
| 157 | } | ||
| 158 | |||
| 159 | void TTSScriptingInterface::stopLastSpeech() { | ||
| 160 | if (_lastSoundAudioInjector) { | ||
| 161 | _lastSoundAudioInjector->stop(); | ||
| 162 | } | ||
| 163 | } | ||