@pipedream/openai 0.7.3 → 0.9.0

package/actions/create-transcription/create-transcription.mjs

@@ -1,264 +0,0 @@
-import ffmpegInstaller from "@ffmpeg-installer/ffmpeg";
-import { ConfigurationError } from "@pipedream/platform";
-import axios from "axios";
-import Bottleneck from "bottleneck";
-import { exec } from "child_process";
-import FormData from "form-data";
-import fs from "fs";
-import {
-  extname,
-  join,
-} from "path";
-import stream from "stream";
-import { promisify } from "util";
-import openai from "../../openai.app.mjs";
-import common from "../common/common.mjs";
-import constants from "../../common/constants.mjs";
-import lang from "../common/lang.mjs";
-
-const COMMON_AUDIO_FORMATS_TEXT = "Your audio file must be in one of these formats: mp3, mp4, mpeg, mpga, m4a, wav, or webm.";
-const CHUNK_SIZE_MB = 20;
-
-const execAsync = promisify(exec);
-const pipelineAsync = promisify(stream.pipeline);
-
-export default {
-  name: "Create Transcription (Whisper)",
-  version: "0.1.16",
-  key: "openai-create-transcription",
-  description: "Transcribes audio into the input language. [See the documentation](https://platform.openai.com/docs/api-reference/audio/create).",
-  type: "action",
-  props: {
-    openai,
-    uploadType: {
-      label: "Audio Upload Type",
-      description: "Are you uploading an audio file from [your workflow's `/tmp` directory](https://pipedream.com/docs/code/nodejs/working-with-files/#the-tmp-directory), or providing a URL to the file?",
-      type: "string",
-      options: [
-        "File",
-        "URL",
-      ],
-      reloadProps: true,
-    },
-    language: {
-      label: "Language",
-      description: "**Optional**. The language of the input audio. Supplying the input language will improve accuracy and latency.",
-      type: "string",
-      optional: true,
-      options: lang.LANGUAGES.map((l) => ({
-        label: l.label,
-        value: l.value,
-      })),
-    },
-  },
-  async additionalProps() {
-    const props = {};
-    switch (this.uploadType) {
-    case "File":
-      props.path = {
-        type: "string",
-        label: "File Path",
-        description: `A path to your audio file to transcribe, e.g. \`/tmp/audio.mp3\`. ${COMMON_AUDIO_FORMATS_TEXT} Add the appropriate extension (mp3, mp4, etc.) on your filename — OpenAI uses the extension to determine the file type. [See the Pipedream docs on saving files to \`/tmp\`](https://pipedream.com/docs/code/nodejs/working-with-files/#writing-a-file-to-tmp)`,
-      };
-      break;
-    case "URL":
-      props.url = {
-        type: "string",
-        label: "URL",
-        description: `A public URL to the audio file to transcribe. This URL must point directly to the audio file, not a webpage that links to the audio file. ${COMMON_AUDIO_FORMATS_TEXT}`,
-      };
-      break;
-    default:
-      throw new ConfigurationError("Invalid upload type specified. Please provide 'File' or 'URL'.");
-    }
-    // Because we need to display the file or URL above, and not below, these optional props
-    // TODO: Will be fixed when we render optional props correctly when used with additionalProps
-    props.prompt = {
-      label: "Prompt",
-      description: "**Optional** text to guide the model's style or continue a previous audio segment. The [prompt](https://platform.openai.com/docs/guides/speech-to-text/prompting) should match the audio language.",
-      type: "string",
-      optional: true,
-    };
-    props.responseFormat = {
-      label: "Response Format",
-      description: "**Optional**. The format of the response. The default is `json`.",
-      type: "string",
-      default: "json",
-      optional: true,
-      options: constants.TRANSCRIPTION_FORMATS,
-    };
-    props.temperature = common.props.temperature;
-
-    return props;
-  },
-  methods: {
-    createForm({
-      file, outputDir,
-    }) {
-      const form = new FormData();
-      form.append("model", "whisper-1");
-      if (this.prompt) form.append("prompt", this.prompt);
-      if (this.temperature) form.append("temperature", this.temperature);
-      if (this.language) form.append("language", this.language);
-      if (this.responseFormat) form.append("response_format", this.responseFormat);
-      const readStream = fs.createReadStream(join(outputDir, file));
-      form.append("file", readStream);
-      return form;
-    },
-    async splitLargeChunks(files, outputDir) {
-      for (const file of files) {
-        if (fs.statSync(`${outputDir}/${file}`).size / (1024 * 1024) > CHUNK_SIZE_MB) {
-          await this.chunkFile({
-            file: `${outputDir}/${file}`,
-            outputDir,
-            index: file.slice(6, 9),
-          });
-          await execAsync(`rm -f "${outputDir}/${file}"`);
-        }
-      }
-    },
-    async chunkFileAndTranscribe({
-      file, $,
-    }) {
-      const outputDir = join("/tmp", "chunks");
-      await execAsync(`mkdir -p "${outputDir}"`);
-      await execAsync(`rm -f "${outputDir}/*"`);
-
-      await this.chunkFile({
-        file,
-        outputDir,
-      });
-
-      let files = await fs.promises.readdir(outputDir);
-      // ffmpeg will sometimes return chunks larger than the allowed size,
-      // so we need to identify large chunks and break them down further
-      await this.splitLargeChunks(files, outputDir);
-      files = await fs.promises.readdir(outputDir);
-
-      return this.transcribeFiles({
-        files,
-        outputDir,
-        $,
-      });
-    },
-    async chunkFile({
-      file, outputDir, index,
-    }) {
-      const ffmpegPath = ffmpegInstaller.path;
-      const ext = extname(file);
-
-      const fileSizeInMB = fs.statSync(file).size / (1024 * 1024);
-      // We're limited to 26MB per request. Because of how ffmpeg splits files,
-      // we need to be conservative in the number of chunks we create
-      const conservativeChunkSizeMB = CHUNK_SIZE_MB;
-      const numberOfChunks = !index
-        ? Math.ceil(fileSizeInMB / conservativeChunkSizeMB)
-        : 2;
-
-      if (numberOfChunks === 1) {
-        await execAsync(`cp "${file}" "${outputDir}/chunk-000${ext}"`);
-        return;
-      }
-
-      const { stdout } = await execAsync(`${ffmpegPath} -i "${file}" 2>&1 | grep "Duration"`);
-      const duration = stdout.match(/\d{2}:\d{2}:\d{2}\.\d{2}/s)[0];
-      const [
-        hours,
-        minutes,
-        seconds,
-      ] = duration.split(":").map(parseFloat);
-
-      const totalSeconds = (hours * 60 * 60) + (minutes * 60) + seconds;
-      const segmentTime = Math.ceil(totalSeconds / numberOfChunks);
-
-      const command = `${ffmpegPath} -i "${file}" -f segment -segment_time ${segmentTime} -c copy "${outputDir}/chunk-${index
-        ? `${index}-`
-        : ""}%03d${ext}"`;
-      await execAsync(command);
-    },
-    transcribeFiles({
-      files, outputDir, $,
-    }) {
-      const limiter = new Bottleneck({
-        maxConcurrent: 1,
-        minTime: 1000 / 59,
-      });
-
-      return Promise.all(files.map((file) => {
-        return limiter.schedule(() => this.transcribe({
-          file,
-          outputDir,
-          $,
-        }));
-      }));
-    },
-    transcribe({
-      file, outputDir, $,
-    }) {
-      const form = this.createForm({
-        file,
-        outputDir,
-      });
-      return this.openai.createTranscription({
-        $,
-        form,
-      });
-    },
-    getFullText(transcriptions = []) {
-      return transcriptions.map((t) => t.text || t).join(" ");
-    },
-  },
-  async run({ $ }) {
-    const {
-      url,
-      path,
-    } = this;
-
-    if (!url && !path) {
-      throw new ConfigurationError("Must specify either File URL or File Path");
-    }
-
-    let file;
-
-    if (path) {
-      if (!fs.existsSync(path)) {
-        throw new ConfigurationError(`${path} does not exist`);
-      }
-
-      file = path;
-    } else if (url) {
-      const ext = extname(url).split("?")[0];
-
-      const response = await axios({
-        method: "GET",
-        url,
-        responseType: "stream",
-        timeout: 250000,
-      });
-
-      const bufferStream = new stream.PassThrough();
-      response.data.pipe(bufferStream);
-
-      const downloadPath = join("/tmp", `audio${ext}`);
-      const writeStream = fs.createWriteStream(downloadPath);
-
-      await pipelineAsync(bufferStream, writeStream);
-
-      file = downloadPath;
-    }
-
-    const transcriptions = await this.chunkFileAndTranscribe({
-      file,
-      $,
-    });
-
-    if (transcriptions.length) {
-      $.export("$summary", "Successfully created transcription");
-    }
-
-    return {
-      transcription: this.getFullText(transcriptions),
-      transcriptions,
-    };
-  },
-};