javascript - Mediastream 管道到 NodeJS socket.io 流到 Google Speech API 并流回响应

Question

我想使用Google Speech API实现语音到文本，但在我的前端我不太明白我应该做什么，我在后端和前端都使用Socket.io Stream。

前端（Javascript）

bindSendAudioMessage() {
    let me = this;

    me.sendAudioMessageButton = me.ele.find('#send-audio-message-btn');

    me.sendAudioMessageButton.off('click').one('click', async function () {
        let stream = await navigator.mediaDevices.getUserMedia({ audio : true});
        me.recordingStarted(stream);
    });
},
recordingStarted: function (inputStream) {
    let serverStream = ss.createStream();
    ss(chatBox.socketIO).emit('speech-to-text', serverStream);
    inputStream.pipe(serverStream);
    ss(chatBox.socketIO).on('speech-text', function (stream) {
        console.log('receiving something');
        console.log(stream);
        stream.on('data', data => {
            console.log(data);
        })
    })
},

后端（NodeJS）

// Imports the Google Cloud client library
const speech = require('@google-cloud/speech');

// Creates a client
const client = new speech.SpeechClient();
SocketStream(socket).on('speech-to-text', function (inputStream) {
    console.log(inputStream);
    const request = {
        config: {
            encoding: 'LINEAR16',
            sampleRateHertz: 16000,
            languageCode: 'en-US',
        },
        interimResults: false, // If you want interim results, set this to true
        single_utterance: true,
    };

    // Create a recognize stream
    const recognizeStream = client
        .streamingRecognize(request)
        .on('error', console.error)
        .on('data', data =>
            process.stdout.write(
                data.results[0] && data.results[0].alternatives[0]
                    ? `Transcript: ${data.results[0].alternatives[0].transcript}\n`
                    : `\n\nReached transcription time limit, press Ctrl+C\n`
            )
        );

    let outputStream = SocketStream.createStream();
    SocketStream(socket).emit('speech-text', outputStream);

    // Pipe inputStream to recognizeStream then to outputStream
    inputStream.pipe(recognizeStream).pipe(outputStream);
})

我确信我在流 API 中缺少一些东西，我知道的一个问题是navigator.mediaDevices.getUserMedia({ audio : true})会给我一个MediaStream与SocketIO Stream.

• 如何准备音频MediaStream以将其流式传输到SocketIO Stream？

• 当我从 Google API 获得响应时，如何流回响应？

• 这条线inputStream.pipe(recognizeStream).pipe(outputStream);有意义吗？

标签： javascriptnode.jsgetusermediagoogle-speech-apimediastream