UnhandledPromiseRejectionWarning when calling SpeechClient.recognize

[en4letto]

I am using the @google-cloud/speech library and bundling an express app with webpack.

I had the proto error (as described here) and fixed it by passing {fallback: true} to the constructor.

Right now I get the error UnhandledPromiseRejectionWarning: TypeError: fetch is not a function, which is triggered when in my code I execute the recognize function of the client.
Here the related portions of my code:

// THIS IS THE CONFIG PASSED TO THE CONSTRUCTOR
authSTTConfig: {
        projectId: process.env.STT_PROJECT_ID,
        keyFilename: process.env.STT_KEY_FILENAME,
        fallback: true,
    }
//

...

class GCloudSpeechAPIv2 {
    constructor() {
        this.audioChunks = [];
        this.finalAudio = null;

        this.instanceSTT = new speech.SpeechClient(authSTTConfig);

        this.requestSTT = {
            // interimResults: true,
            config: cloudSTTConfig,
        };
    }

    async transcribeAudio() {
        this.requestSTT.audio = {
            content: this.finalAudio.toString("base64"),
        };

        // console.log("transcribeAudio");
        const [response] = await this.instanceSTT.recognize(this.requestSTT); // ERROR IS THROWN HERE
        const transcription = response.results
            .map((result) => result.alternatives[0].transcript)
            .join("");
        return transcription;
    }

...

Environment details

• OS: ubuntu 20.04
• Node.js version: v12.18.4
• npm version: 6.14.6
• @google-cloud/speech version: 4.1.5

Steps to reproduce

1. Implement the code to call the recognize function of the speech client
2. Bundle the application with webpack, add the {fallback: true} option

Thanks!

[en4letto]

A quick update: the problem seems to be related to node-fetch and webpack, here a more detailed analysis.

[b-loved-dreamer]

Hi @en4letto, upon trying your code, a UnhandledPromiseRejectionWarning: TypeError: Cannot read property 'toString' of null error was thrown. The 'toString' of null seems to be the causing root of the issue and this is because your finalAudio variable is never set to anything. After I assigned this variable, the UnhandledPromiseRejectionWarning went away. If you want to set your credentials explicitly, you can follow this sample: const client = new speech.SpeechClient({KeyFileName: 'path/to/file.json.'});. The request object is composed of config and audio objects. Please take a look at this sample.

[b-loved-dreamer]

Closing this issue. Please reopen if needed.