我在 GCS 事件触发的云函数中调用云语音转文本 API。
在云函数之外执行此操作(运行 node index.js
)完全没问题,但我的错误随后出现。
使用this doc ,我认为该错误是由于身份验证问题造成的,但我尝试了几件事,但现在不太确定。
我的代码是:
const {Storage} = require('@google-cloud/storage');
const storage = new Storage();
const nl = require('@google-cloud/language');
const client_nl = new nl.LanguageServiceClient();
const speech = require('@google-cloud/speech');
const client_speech = new speech.SpeechClient();
exports.getRecording = (data,context) => {
const file = data;
if (file.resourceState === 'not_exists') {
// Ignore file deletions
return true;
} else if (!new RegExp(/\.(wav|mp3)/g).test(file.name)) {
// Ignore changes to non-audio files
return true;
}
console.log(`Analyzing gs://${file.bucket}/${file.name}`);
const bucket = storage.bucket(file.bucket);
const audio = {
uri: 'gs://${file.bucket}/${file.name}'
};
// Configure audio settings for BoF recordings
const audioConfig = {
encoding: 'LINEAR16',
sampleRateHertz: 44100,
languageCode: 'fr-FR'
};
const request = {
audio: audio,
config: audioConfig,
};
return client_speech.recognize(request)
.then(([transcription]) => {
const filename = `analysis.json`;
console.log(`Saving gs://${file.bucket}/${filename}`);
return bucket
.file(filename)
.save(JSON.stringify(transcription, null, 2));
});
然后我部署:
gcloud 函数部署 getRecording --runtime nodejs10 --trigger-resource trigger-bucket-id --trigger-event google.storage.object.finalize --service-account my-service-account
我尝试过:
导出 GOOGLE_APPLICATION_CREDENTIALS=/path/to/file/keyfile.json
- 添加了一个
config.json
文件,其中包含"GOOGLE_APPLICATION_CREDENTIALS":"./keyfile.json"
,以及根项目中的 key 文件和index.js
中的require('./config.json')
- 添加了选项 json
const options = {
projectId: 'my-project-id',
keyFilename: './key-file.json'
};
const client_speech = new speech.SpeechClient(options);
我不断收到此错误,需要一些帮助
D getRecording 573287126069013 2019-06-07 15:03:09.609 Function execution started
getRecording 573287126069013 2019-06-07 15:03:09.789 Analyzing gs://my-bucket/audio_trimed.wav
D getRecording 573287126069013 2019-06-07 15:03:10.979 Function execution took 1372 ms, finished with status: 'error'
E getRecording 573291408785013 2019-06-07 15:03:11.990 Error: Requested entity was not found.
at Http2CallStream.call.on (/srv/functions/node_modules/@grpc/grpc-js/build/src/client.js:101:45)
at Http2CallStream.emit (events.js:194:15)
at Http2CallStream.EventEmitter.emit (domain.js:459:23)
at Http2CallStream.endCall (/srv/functions/node_modules/@grpc/grpc-js/build/src/call-stream.js:63:18)
at handlingTrailers (/srv/functions/node_modules/@grpc/grpc-js/build/src/call-stream.js:152:18)
at process._tickCallback (internal/process/next_tick.js:68:7)
最佳答案
有 this code 中的任何一个吗帮助? 有一些 StackOverflow 帖子可能会有所帮助 here ,和here
我建议更详细的日志记录/调试。 例如,你能证明这两行代码符合你的预期吗?该字符串是否生成格式正确的 uri?
const bucket = storage.bucket(file.bucket);
const audio = {
uri: 'gs://${file.bucket}/${file.name}'
};
如果您找到解决方案,请告诉我们
关于node.js - 使用 Speech-to-Text API 无法解决 "Requested entity was not found"错误,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/56497171/