mirror of
https://gitlab.rlp.net/proj-wise2526-video2document/video2document.git
synced 2026-06-15 18:01:52 +02:00
Compare commits
31 Commits
| Author | SHA1 | Date | |
|---|---|---|---|
| 2f5efee9c7 | |||
| 465fe8bd41 | |||
| b87bfd444d | |||
| 97b571b7f9 | |||
| 455147a41b | |||
| 9441699561 | |||
| 2edc7f8351 | |||
| 6083773f88 | |||
| 444d408480 | |||
| d9eacafc3a | |||
| ab737f0dc9 | |||
| 79e0c48755 | |||
| 9254ddc57f | |||
| c021272ca4 | |||
| e7e97a7f60 | |||
| 80392874bb | |||
| 94f390f28b | |||
| 73f6fa7524 | |||
| 87e3368a9a | |||
| 6d9c94c685 | |||
| fbd5368223 | |||
| 449255dda1 | |||
| c8cbd4e92a | |||
| 0f54edb0aa | |||
| 54f1f6c135 | |||
| a32e7e5744 | |||
| d9e96316c0 | |||
| da3a8c7d8c | |||
| dfe91bb15a | |||
| 92043440fe | |||
| bd8ed1bf81 |
+5
-1
@@ -964,4 +964,8 @@ app.*.symbols
|
||||
!/dev/ci/**/Gemfile.lock
|
||||
|
||||
#Storage files
|
||||
storage/
|
||||
storage/
|
||||
|
||||
*.mp4
|
||||
*.wav
|
||||
*.flac
|
||||
|
||||
Binary file not shown.
|
After Width: | Height: | Size: 2.6 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 18 KiB |
Binary file not shown.
|
After Width: | Height: | Size: 35 KiB |
@@ -3,15 +3,22 @@
|
||||
<head>
|
||||
<meta charset="UTF-8">
|
||||
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
||||
<title>Video Upload Drag and Drop + Button</title>
|
||||
<title id="title">Video to document</title>
|
||||
<link rel="stylesheet" href="style.css">
|
||||
</head>
|
||||
<body>
|
||||
|
||||
<div class="mitte" id="mitte">
|
||||
<h1>Video to document</h1>
|
||||
<div class="flagsBtns" id="flagsBtns">
|
||||
<button class="de_Btn" id="de_Btn" onclick="changeLanguage('de')"><img src="flags/germany-flag-png-large.jpg" width="25px" height="20px"/></button>
|
||||
<button class="eng_Btn" id="eng_Btn" onclick="changeLanguage('en')"><img src="flags/united-kingdom-flag-png-large.jpg" width="25px" height="20px"/></button>
|
||||
<button class="in_Btn" id="in_Btn" onclick="changeLanguage('in')"><img src="flags/india-flag-png-large.png" width="25px" height="20px"/></button>
|
||||
</div>
|
||||
|
||||
<h1 id="h1">Video to document</h1>
|
||||
|
||||
<div class="upload-container" id="uploadContainer">
|
||||
<p>Drag and drop video file</p>
|
||||
<p id="p1">Drag and drop video file</p>
|
||||
<div class="file-name" id="fileName">No video chosen</div>
|
||||
</div>
|
||||
|
||||
@@ -19,19 +26,19 @@
|
||||
<input type="file" id="videoUpload" accept="video/*">
|
||||
|
||||
<div class="checkbox-group">
|
||||
<label for="checkbox-group">Choose prefered document style:</label>
|
||||
<label id="checkbox_group" for="checkbox-group">Choose prefered document style:</label>
|
||||
<div class="checkbox-container">
|
||||
<input type="checkbox" id="docFormat">
|
||||
<label for "docFormat">Meeting report</label>
|
||||
<input type="checkbox" name ="docFormat" id="docFormat">
|
||||
<label id="label_format" for="docFormat">Meeting report</label>
|
||||
</div>
|
||||
|
||||
<div class="checkbox-container">
|
||||
<input type="checkbox" id="docFormatCustom">
|
||||
<label for "docFormatCustom">Summary with timestamps</label>
|
||||
<input type="checkbox" name="docFormat" id="docFormatSummary">
|
||||
<label id="label_summary" for="docFormatSummary">Summary with timestamps</label>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<button class="submit-btn" id="submitButton">Submit Video</button>
|
||||
<button class="submit-btn" id="submitButton" onclick="checkBoxes()">Submit</button>
|
||||
|
||||
<div class="progressbar">
|
||||
<div class="progress_fill"></div>
|
||||
|
||||
@@ -2,6 +2,20 @@
|
||||
|
||||
const { contextBridge, ipcRenderer, webUtils } = require('electron')
|
||||
|
||||
contextBridge.exposeInMainWorld("explorer", {
|
||||
onFileDrop: (file) => webUtils.getPathForFile(file)
|
||||
})
|
||||
try {
|
||||
contextBridge.exposeInMainWorld("explorer", {
|
||||
onFileDrop: (file) => webUtils.getPathForFile(file)
|
||||
})
|
||||
contextBridge.exposeInMainWorld("extractor", {
|
||||
extract: (file) => ipcRenderer.send("extract", file)
|
||||
})
|
||||
contextBridge.exposeInMainWorld("electronAPI", {
|
||||
getFilePath: (file) => {return webUtils.getPathForFile(file)}
|
||||
})
|
||||
contextBridge.exposeInMainWorld("summarizer", {
|
||||
runFile: (file) => ipcRenderer.send("summarize-transcription", file)
|
||||
});
|
||||
} catch (error) {
|
||||
console.log("Error in preload.js");
|
||||
}
|
||||
|
||||
|
||||
+33
-18
@@ -1,22 +1,37 @@
|
||||
|
||||
const dropzone = document.getElementById("uploadContainer");
|
||||
|
||||
dropzone.addEventListener("dragover", (e) =>{
|
||||
e.stopPropagation();
|
||||
e.preventDefault();
|
||||
});
|
||||
|
||||
dropzone.addEventListener("drop", (e) => {
|
||||
e.stopPropagation()
|
||||
e.preventDefault()
|
||||
const files = e.dataTransfer.files
|
||||
const filePath = window.explorer.onFileDrop(files[0])
|
||||
var holdy = filePath + "";
|
||||
if(holdy.endsWith(".mp4")){
|
||||
console.log(filePath)
|
||||
|
||||
const files1 = e.dataTransfer.files;
|
||||
handleFiles(files1);
|
||||
uploadContainer.addEventListener("dragover", (e) =>{
|
||||
try {
|
||||
e.stopPropagation();
|
||||
e.preventDefault();
|
||||
} catch (error) {
|
||||
console.log("Error in renderer.js dragover listener function")
|
||||
}
|
||||
|
||||
});
|
||||
|
||||
//listener for when a file get dropped on the drag&drop field
|
||||
uploadContainer.addEventListener("drop", (e) => {
|
||||
try {
|
||||
e.stopPropagation()
|
||||
e.preventDefault()
|
||||
const files = e.dataTransfer.files
|
||||
const filePath = window.explorer.onFileDrop(files[0])
|
||||
|
||||
var holdy = String(filePath);
|
||||
const lower = holdy.toLowerCase();
|
||||
const validExt = [".mp4", ".mov", ".avi", ".mkv"];
|
||||
|
||||
if(validExt.some(ext => lower.endsWith(ext))){
|
||||
console.log(filePath);
|
||||
const files1 = e.dataTransfer.files;
|
||||
handleFiles(files1);
|
||||
}else{
|
||||
console.log('Video format invalid!');
|
||||
}
|
||||
} catch (error) {
|
||||
console.log("Error in renderer.js with the listerner for the drop function");
|
||||
console.log(error);
|
||||
}
|
||||
|
||||
|
||||
})
|
||||
+110
-26
@@ -1,37 +1,121 @@
|
||||
const uploadContainer = document.getElementById('uploadContainer');
|
||||
const fileInput = document.getElementById('videoUpload');
|
||||
const fileName = document.getElementById('fileName');
|
||||
const manualBtn = document.getElementById('manualUploadBtn');
|
||||
const videoPreview = document.getElementById('videoPreview');
|
||||
|
||||
//listener for the file explorer search
|
||||
manualUploadBtn.addEventListener('click', () => {
|
||||
try {
|
||||
videoUpload.click();
|
||||
} catch (error) {
|
||||
console.log("Error in manualBtn EventListener click");
|
||||
}
|
||||
|
||||
|
||||
|
||||
|
||||
manualBtn.addEventListener('click', () => {
|
||||
fileInput.click();
|
||||
});
|
||||
|
||||
//function to check if one checkbox is at least klicked
|
||||
function checkBoxes() {
|
||||
try {
|
||||
const checkboxes = document.querySelectorAll('input[name="docFormat"]');
|
||||
let isChecked = false;
|
||||
|
||||
fileInput.addEventListener('change', () => {
|
||||
handleFiles(fileInput.files);
|
||||
});
|
||||
checkboxes.forEach(function (checkbox) {
|
||||
if (checkbox.checked) {
|
||||
isChecked = true;
|
||||
}
|
||||
});
|
||||
|
||||
if (isChecked) {
|
||||
//Code to submit the video
|
||||
var pathTest = window.electronAPI.getFilePath(videoUpload.files[0]);
|
||||
|
||||
const lower = pathTest.toLowerCase();
|
||||
const validExt = [".mp4", ".mov", ".avi", ".mkv"];
|
||||
|
||||
if(validExt.some(ext => lower.endsWith(ext))){
|
||||
window.extractor.extract({ inputVideoPath: pathTest, outputType: "wav" });
|
||||
}
|
||||
|
||||
|
||||
|
||||
|
||||
function handleFiles(files) {
|
||||
if (files.length > 0) {
|
||||
const file = files[0];
|
||||
if (file.type.startsWith('video/')) {
|
||||
fileInput.files = files;
|
||||
fileName.textContent = `Chosen video: ${file.name}`;
|
||||
} else {
|
||||
//language only english at the moment
|
||||
alert('Please select at least one document type.');
|
||||
}
|
||||
}
|
||||
} catch (error) {
|
||||
console.log(error)
|
||||
}
|
||||
|
||||
// mapFunctions.get("extraction-video-to-audio").function({inputVideoPath:"./a.mp4", outputType:"wav"})
|
||||
}
|
||||
|
||||
function updateProgressBar(bar, value){
|
||||
value = Math.round(value);
|
||||
bar.querySelector(".progress_fill").style.width = `${value}%`;
|
||||
bar.querySelector(".progress_text").textContent = `${value}%`;
|
||||
//language changing feature
|
||||
function changeLanguage(language) {
|
||||
if (language === 'en') {
|
||||
document.getElementById('title').textContent = 'Video to document';
|
||||
document.getElementById('h1').textContent = 'Video to document';
|
||||
document.getElementById('p1').textContent = 'Drag and drop video file';
|
||||
document.getElementById('fileName').textContent = 'No video chosen';
|
||||
document.getElementById('manualUploadBtn').textContent = 'Search video';
|
||||
document.getElementById('checkbox_group').textContent = 'Choose prefered document style:';
|
||||
document.getElementById('label_format').textContent = 'Meeting report';
|
||||
document.getElementById('label_summary').textContent = 'Summary with timestamps';
|
||||
document.getElementById('submitButton').textContent = 'Submit';
|
||||
} else if (language === 'de') {
|
||||
document.getElementById('title').textContent = 'Video zu Dokument';
|
||||
document.getElementById('h1').textContent = 'Video zu Dokument';
|
||||
document.getElementById('p1').textContent = 'Video per Drag & Drop ablegen';
|
||||
document.getElementById('fileName').textContent = 'Kein Video ausgewaehlt';
|
||||
document.getElementById('manualUploadBtn').textContent = 'Video suchen';
|
||||
document.getElementById('checkbox_group').textContent = 'Bevorzugte Dokumentvarianten:';
|
||||
document.getElementById('label_format').textContent = 'Meeting Bericht';
|
||||
document.getElementById('label_summary').textContent = 'Zusammenfassung mit Zeitstempeln';
|
||||
document.getElementById('submitButton').textContent = 'Absenden';
|
||||
} else if (language == "in") {
|
||||
document.getElementById('title').textContent = 'दस्तावेज़ के लिए वीडियो';
|
||||
document.getElementById('h1').textContent = 'दस्तावेज़ के लिए वीडियो';
|
||||
document.getElementById('p1').textContent = 'वीडियो फ़ाइल खींचें और छोड़ें';
|
||||
document.getElementById('fileName').textContent = 'कोई वीडियो नहीं चुना गया';
|
||||
document.getElementById('manualUploadBtn').textContent = 'वीडियो खोजें';
|
||||
document.getElementById('checkbox_group').textContent = 'पसंदीदा दस्तावेज़ शैली चुनें:';
|
||||
document.getElementById('label_format').textContent = 'बैठक रिपोर्ट';
|
||||
document.getElementById('label_summary').textContent = 'टाइमस्टैम्प के साथ सारांश';
|
||||
document.getElementById('submitButton').textContent = 'जमा करना';
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
//listener for the file explorer search when something got selected
|
||||
videoUpload.addEventListener('change', () => {
|
||||
try {
|
||||
handleFiles(videoUpload.files);
|
||||
} catch (error) {
|
||||
console.log("Error in manualBtn EventListener change");
|
||||
}
|
||||
|
||||
});
|
||||
|
||||
|
||||
|
||||
//function to display the file path in the drop down box
|
||||
function handleFiles(files) {
|
||||
try {
|
||||
if (files.length > 0) {
|
||||
const file = files[0];
|
||||
if (file.type.startsWith('video/')) {
|
||||
videoUpload.files = files;
|
||||
fileName.textContent = `Chosen video: ${file.name}`;
|
||||
}
|
||||
}
|
||||
} catch (error) {
|
||||
console.log("Error in script.js handleFiles function");
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
//function to regulate the progress on the progressbar
|
||||
function updateProgressBar(bar, value) {
|
||||
try {
|
||||
value = Math.round(value);
|
||||
bar.querySelector(".progress_fill").style.width = `${value}%`;
|
||||
bar.querySelector(".progress_text").textContent = `${value}%`;
|
||||
} catch (error) {
|
||||
console.log("Error in scripts.js updateProgressBar function");
|
||||
}
|
||||
|
||||
}
|
||||
+19
-4
@@ -5,7 +5,7 @@ body {
|
||||
justify-content: center;
|
||||
align-items: center;
|
||||
height: 100vh;
|
||||
background-color: #444;
|
||||
background-color: #555;
|
||||
gap: 15px;
|
||||
margin: 0;
|
||||
}
|
||||
@@ -75,6 +75,7 @@ gap: 5px;
|
||||
|
||||
.checkbox-group {
|
||||
margin-top: 15px;
|
||||
margin-bottom: 15px;
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
gap: 10px;
|
||||
@@ -92,14 +93,14 @@ gap: 5px;
|
||||
}
|
||||
|
||||
.mitte {
|
||||
background-color: #eaf0ff;
|
||||
background-color: #f2f3f4;
|
||||
display: flex;
|
||||
flex-direction: column;
|
||||
align-items: center;
|
||||
padding: 5% 50px;
|
||||
margin-top: 20px;
|
||||
gap: 10px;
|
||||
border: 2px;
|
||||
border: 1px;
|
||||
border-color: black;
|
||||
border-style: solid;
|
||||
}
|
||||
@@ -112,7 +113,7 @@ h1 {
|
||||
position: relative;
|
||||
width: 210px;
|
||||
height: 30px;
|
||||
background: darkslategray;
|
||||
background: rgb(42, 46, 78);
|
||||
border-radius: 5px;
|
||||
overflow: hidden;
|
||||
}
|
||||
@@ -129,4 +130,18 @@ h1 {
|
||||
top: 50%;
|
||||
right: 5px;
|
||||
transform: translateY(-50%);
|
||||
color: white;
|
||||
}
|
||||
|
||||
.flagsBtns {
|
||||
display: flex;
|
||||
justify-content: flex-end;
|
||||
}
|
||||
|
||||
.de_Btn, .eng_Btn, .in_Btn {
|
||||
padding: 8px 16px;
|
||||
color: white;
|
||||
border: none;
|
||||
border-radius: 8px;
|
||||
cursor: pointer;
|
||||
}
|
||||
@@ -55,4 +55,30 @@ rl.on("line", data =>{
|
||||
|
||||
// ----------------------------------------------------------- ELECTRON ----------------------------------------------------------- //
|
||||
|
||||
// TODO - Add Electron support to the project
|
||||
let mainWindow;
|
||||
|
||||
function createWindow() {
|
||||
mainWindow = new electron.BrowserWindow({
|
||||
width: 800,
|
||||
height: 600,
|
||||
webPreferences: {
|
||||
nodeIntegration: false,
|
||||
contextIsolation: true,
|
||||
preload: `${mainDir}/electron/main/preload.js`
|
||||
}
|
||||
});
|
||||
|
||||
mainWindow.loadFile('./electron/main/index.html');
|
||||
}
|
||||
|
||||
electron.app.whenReady().then(createWindow);
|
||||
|
||||
|
||||
electron.ipcMain.on("extract", (event, args) => {
|
||||
mapFunctions.get("extraction-video-to-audio").function(args)
|
||||
})
|
||||
|
||||
electron.ipcMain.on("summarize-transcription", (event, args) => {
|
||||
mapFunctions.get("summarize-transcription").function(args);
|
||||
});
|
||||
|
||||
|
||||
Generated
+599
-759
File diff suppressed because it is too large
Load Diff
+4
-2
@@ -1,8 +1,9 @@
|
||||
{
|
||||
"type": "module",
|
||||
"dependencies": {
|
||||
"@google/genai": "^1.29.0",
|
||||
"@types/axios": "^0.9.36",
|
||||
"cli-progress": "^3.12.0",
|
||||
"dotenv": "^17.2.3",
|
||||
"electron": "^39.1.1",
|
||||
"express": "^5.1.0",
|
||||
"ffmpeg-static": "^5.2.0",
|
||||
"fluent-ffmpeg": "^2.1.3"
|
||||
@@ -23,6 +24,7 @@
|
||||
"test": "tests"
|
||||
},
|
||||
"scripts": {
|
||||
"start": "electron main.js",
|
||||
"test": "echo \"Error: no test specified\" && exit 1"
|
||||
},
|
||||
"repository": {
|
||||
|
||||
@@ -8,3 +8,11 @@ fs = require("fs")
|
||||
readline = require("readline")
|
||||
config = require("./config/config")
|
||||
|
||||
ffmpegPath = require('ffmpeg-static');
|
||||
ffmpeg = require('fluent-ffmpeg');
|
||||
path = require('path');
|
||||
cliProgress = require('cli-progress');
|
||||
|
||||
// { app, BrowserWindow, ipcMain, dialog } = require('electron');
|
||||
|
||||
electron = require('electron');
|
||||
|
||||
@@ -1,5 +0,0 @@
|
||||
npx ts-node ./extract.ts /Users/mikehughes/Downloads/Testvideo/Kurzgesagt.mov
|
||||
npx ts-node ./transcribe.ts ../storage/audio/Kurzgesagt.wav
|
||||
|
||||
npx ts-node ./extract.ts /Users/mikehughes/Downloads/Testvideo/GitLabMeeting.mov
|
||||
npx ts-node ./transcribe.ts ../storage/audio/GitLabMeeting.wav
|
||||
@@ -1,23 +0,0 @@
|
||||
#!/usr/bin/env ts-node
|
||||
|
||||
import { extractAudioFromVideo } from "../services/modules/extraction/ffmpegExtractor.ts";
|
||||
|
||||
const videoPath = process.argv[2];
|
||||
|
||||
if (!videoPath) {
|
||||
console.error("Usage: ts-node extractAudio.ts <videoPath>");
|
||||
process.exit(1);
|
||||
}
|
||||
|
||||
(async () => {
|
||||
try {
|
||||
console.log(`Extracting audio from: ${videoPath}`);
|
||||
|
||||
await extractAudioFromVideo(videoPath); // Call the extraction function (ffmpegExtractor.ts in services/modules/extraction)
|
||||
|
||||
console.log("Audio extraction completed successfully.");
|
||||
} catch (err) {
|
||||
console.error("Audio extraction failed:", err);
|
||||
process.exit(1);
|
||||
}
|
||||
})();
|
||||
@@ -1,18 +0,0 @@
|
||||
import { whisperLocal } from "../services/modules/transcription/local/whisperLocal.ts";
|
||||
|
||||
const audioPath = process.argv[2];
|
||||
if (!audioPath) {
|
||||
console.error("Please provide an audio file path as argument.");
|
||||
process.exit(1);
|
||||
}
|
||||
|
||||
const whisper = new whisperLocal();
|
||||
|
||||
(async () => {
|
||||
try {
|
||||
const text = await whisper.transcribe(audioPath);
|
||||
console.log(text);
|
||||
} catch (err) {
|
||||
console.error("Transcription failed:", err);
|
||||
}
|
||||
})();
|
||||
@@ -1,31 +0,0 @@
|
||||
import { GoogleGenAI } from "@google/genai";
|
||||
import fs from "fs";
|
||||
|
||||
// Mike Hughes 10/11/25 - Code inspired from
|
||||
// https://www.geeksforgeeks.org/javascript/javascript-program-to-read-text-file/ for reading text files in JS
|
||||
// and from https://ai.google.dev/gemini-api/docs/quickstart?hl=de for using Gemini API
|
||||
// Before using, make sure to export the Google Gemini API key as an environment variable:
|
||||
// export GOOGLE_API_KEY="your_api_key_here"
|
||||
// Keys can be obtained from https://aistudio.google.com/app/api-keys
|
||||
// Also make sure to install the package via sudo npm install @google/genai
|
||||
|
||||
const ai = new GoogleGenAI({});
|
||||
let transcript = "";
|
||||
|
||||
async function main() {
|
||||
try {
|
||||
const transcript = await fs.promises.readFile('../../../../storage/transcripts/Kurzgesagt.txt', 'utf-8');
|
||||
|
||||
const response = await ai.models.generateContent({
|
||||
model: "gemini-2.5-flash",
|
||||
contents: `Create a short documentation about the content of the following transcript:\n\n${transcript}`,
|
||||
});
|
||||
|
||||
console.log(response.text);
|
||||
fs.writeFileSync("../../../../storage/documents/Kurzgesagt_documentation.md", response.text, "utf8");
|
||||
} catch (error) {
|
||||
console.error('Error generating content:', error);
|
||||
}
|
||||
}
|
||||
|
||||
main();
|
||||
@@ -0,0 +1,93 @@
|
||||
|
||||
// Ensure ffmpeg binary is available
|
||||
if (!ffmpegPath) {
|
||||
throw new Error('FFmpeg binary not found!');
|
||||
}
|
||||
ffmpeg.setFfmpegPath(ffmpegPath);
|
||||
|
||||
// Prepare output directory (always storage/audio under project root)
|
||||
const outputDir = `${__dirname}/../../../storage/audio`;
|
||||
if (!fs.existsSync(outputDir)) {
|
||||
fs.mkdirSync(outputDir, { recursive: true });
|
||||
}
|
||||
|
||||
|
||||
|
||||
module.exports = {
|
||||
name:"extraction-video-to-audio", // Unique name for our function that will later be used to get the function from the map via "mapFunctions.get("example").function()"
|
||||
type:"extractor", // value used to differentiate each module to order them in the UI
|
||||
displayname:"Default extractor", // The displayname used within the UI
|
||||
async function(parameter){
|
||||
/*
|
||||
parameter structure:
|
||||
{
|
||||
inputVideoPath: String, // Path to the file
|
||||
outputType: String // Audio file output format
|
||||
}
|
||||
*/
|
||||
let progressBar = new cliProgress.SingleBar({
|
||||
format: 'Processing |{bar}| {percentage}% | {timemark}',
|
||||
barCompleteChar: '\u2588',
|
||||
barIncompleteChar: '\u2591',
|
||||
hideCursor: true
|
||||
});
|
||||
try {
|
||||
// if (meta.url === `file://${process.argv[1]}`) {
|
||||
this.extractAudioFromVideo(parameter.inputVideoPath, progressBar, parameter.outputType)
|
||||
.then(() => console.log('Audio extraction successful.'))
|
||||
.catch((err) => console.error(err));
|
||||
// }
|
||||
} catch (error) {
|
||||
console.log(parameter.outputType);
|
||||
|
||||
}
|
||||
|
||||
},
|
||||
// Derive input and output paths
|
||||
// const inputVideoPath = process.argv[2];
|
||||
// console.log(process.argv);
|
||||
|
||||
/**
|
||||
* Extracts audio from a video using ffmpeg.
|
||||
* - Converts video to WAV (16 kHz, Mono, PCM optional if needed)
|
||||
* - Shows CLI progress bar
|
||||
* - Handles errors gracefully (without errors)
|
||||
*/
|
||||
extractAudioFromVideo: async function (videoFilePath, progressBar, outputType){
|
||||
let inputVideoName = path.basename(videoFilePath, path.extname(videoFilePath));
|
||||
let outputAudioPath = path.join(outputDir, `${inputVideoName}.${outputType}`);
|
||||
|
||||
return new Promise((resolve, reject) => {
|
||||
try {
|
||||
ffmpeg(videoFilePath)
|
||||
.outputFormat(outputType)
|
||||
// .audioCodec('pcm_s16le')
|
||||
.audioChannels(1)
|
||||
.audioFrequency(16000)
|
||||
// .setFfmpegPath("./ffmpeg.exe")
|
||||
.on('progress', (progress) => {
|
||||
if (!progressBar.isActive) progressBar.start(100, 0, { timemark: '00:00:00' });
|
||||
if (progress.percent) {
|
||||
progressBar.update(progress.percent, { timemark: progress.timemark });
|
||||
}
|
||||
})
|
||||
.on('end', () => {
|
||||
progressBar.update(100, { timemark: 'done' });
|
||||
progressBar.stop();
|
||||
console.log(`Extraction completed: ${outputAudioPath}`);
|
||||
resolve();
|
||||
})
|
||||
.on('error', (err) => {
|
||||
progressBar.stop();
|
||||
console.error(`failed_audio_extraction on type ${outputType}: ${err.message}`);
|
||||
reject(err);
|
||||
})
|
||||
.save(outputAudioPath);
|
||||
|
||||
} catch (error) {
|
||||
console.log();
|
||||
}
|
||||
});
|
||||
}
|
||||
|
||||
}
|
||||
@@ -1,90 +0,0 @@
|
||||
import ffmpegPath from 'ffmpeg-static';
|
||||
import ffmpeg from 'fluent-ffmpeg';
|
||||
import path from 'path';
|
||||
import fs from 'fs';
|
||||
import cliProgress from 'cli-progress';
|
||||
import { fileURLToPath } from 'url';
|
||||
|
||||
// Base code reference: https://docs.yemreak.com/arsiv/programming/extract-audio-from-video-with-typescript-and-ffmpeg
|
||||
// Test command: npx ts-node ./extract.ts /path/to/video.mp4
|
||||
|
||||
/**
|
||||
* Extracts audio from a video file and saves it as WAV.
|
||||
* @param videoFilePath Path to the input video file.
|
||||
*/
|
||||
|
||||
// Ensure ffmpeg binary is available
|
||||
if (!ffmpegPath) {
|
||||
throw new Error('FFmpeg binary not found!');
|
||||
}
|
||||
ffmpeg.setFfmpegPath(ffmpegPath);
|
||||
|
||||
// Ensure an input video path is provided via CLI
|
||||
if (process.argv.length < 3) {
|
||||
console.error('Usage: ts-node ./extract.ts <input-video-path>');
|
||||
process.exit(1);
|
||||
}
|
||||
|
||||
// Resolve __dirname equivalent in ESM
|
||||
const __filename = fileURLToPath(import.meta.url);
|
||||
const __dirname = path.dirname(__filename);
|
||||
|
||||
// Prepare output directory (always storage/audio under project root)
|
||||
const outputDir = path.join(__dirname, '..', '..', '..', 'storage', 'audio');
|
||||
if (!fs.existsSync(outputDir)) {
|
||||
fs.mkdirSync(outputDir, { recursive: true });
|
||||
}
|
||||
|
||||
// Derive input and output paths
|
||||
const inputVideoPath = process.argv[2];
|
||||
const inputVideoName = path.basename(inputVideoPath, path.extname(inputVideoPath));
|
||||
const outputAudioPath = path.join(outputDir, `${inputVideoName}.wav`);
|
||||
|
||||
// Initialize CLI progress bar
|
||||
const progressBar = new cliProgress.SingleBar({
|
||||
format: 'Processing |{bar}| {percentage}% | {timemark}',
|
||||
barCompleteChar: '\u2588',
|
||||
barIncompleteChar: '\u2591',
|
||||
hideCursor: true
|
||||
});
|
||||
|
||||
/**
|
||||
* Extracts audio from a video using ffmpeg.
|
||||
* - Converts video to WAV (16 kHz, Mono, PCM optional if needed)
|
||||
* - Shows CLI progress bar
|
||||
* - Handles errors gracefully (without errors)
|
||||
*/
|
||||
export function extractAudioFromVideo(videoFilePath: string): Promise<void> {
|
||||
return new Promise((resolve, reject) => {
|
||||
ffmpeg(videoFilePath)
|
||||
.outputFormat('wav')
|
||||
.audioCodec('pcm_s16le')
|
||||
.audioChannels(1)
|
||||
.audioFrequency(16000)
|
||||
.on('progress', (progress) => {
|
||||
if (!progressBar.isActive) progressBar.start(100, 0, { timemark: '00:00:00' });
|
||||
if (progress.percent) {
|
||||
progressBar.update(progress.percent, { timemark: progress.timemark });
|
||||
}
|
||||
})
|
||||
.on('end', () => {
|
||||
progressBar.update(100, { timemark: 'done' });
|
||||
progressBar.stop();
|
||||
console.log(`Extraction completed: ${outputAudioPath}`);
|
||||
resolve();
|
||||
})
|
||||
.on('error', (err) => {
|
||||
progressBar.stop();
|
||||
console.error(`failed_audio_extraction: ${err.message}`);
|
||||
reject(err);
|
||||
})
|
||||
.save(outputAudioPath);
|
||||
});
|
||||
}
|
||||
|
||||
// Run extraction if executed directly from CLI
|
||||
if (import.meta.url === `file://${process.argv[1]}`) {
|
||||
extractAudioFromVideo(inputVideoPath)
|
||||
.then(() => console.log('Audio extraction successful.'))
|
||||
.catch((err) => console.error(err));
|
||||
}
|
||||
@@ -0,0 +1,138 @@
|
||||
const fs = require("fs");
|
||||
const path = require("path");
|
||||
|
||||
// Prepare output directory (always storage/transcriptionSummaries under project root)
|
||||
const outputDir = `${__dirname}/../../../storage/transcriptionSummaries`;
|
||||
if (!fs.existsSync(outputDir)) {
|
||||
fs.mkdirSync(outputDir, { recursive: true });
|
||||
}
|
||||
|
||||
//Speaker, ALL-Sentences, Start, End
|
||||
|
||||
module.exports = {
|
||||
name: "summarize-transcription", // Unique name for our function that will later be used to get the function from the map via "mapFunctions.get("example").function()"
|
||||
type: "summarizer", // value used to differentiate each module to order them in the UI
|
||||
displayname: "Summarizer", // The displayname used within the UI
|
||||
async function(args) {
|
||||
let inputJson = args.json;
|
||||
|
||||
//JSON Path
|
||||
if (args.jsonPath) {
|
||||
try {
|
||||
const raw = fs.readFileSync(args.jsonPath, "utf-8");
|
||||
inputJson = JSON.parse(raw);
|
||||
} catch (e) {
|
||||
console.error("Failed to load JSON from file:", e);
|
||||
return { error: "Could not read JSON from file path." };
|
||||
}
|
||||
}
|
||||
// JSON parsen
|
||||
if (typeof inputJson === "string") {
|
||||
try {
|
||||
inputJson = JSON.parse(inputJson);
|
||||
} catch (e) {
|
||||
console.log("Invalid JSON in summarize-transcription");
|
||||
return { error: "Invalid JSON" };
|
||||
}
|
||||
}
|
||||
|
||||
const words = inputJson.words;
|
||||
if (!Array.isArray(words)) {
|
||||
return { error: "No words Array found" };
|
||||
}
|
||||
|
||||
const ENDINGS = [".", "!", "?"]; // '...' auch als Satzende ?
|
||||
const ABBREVIATIONS = new Set(["z.B.", "bzw.", "u.a.", "Dr.", "Mr.", "Mrs.", "Prof.", "etc."]); //TODO weitere Ergaenzen
|
||||
|
||||
const result = [];
|
||||
let currentSentence = "";
|
||||
let currentSpeaker = null;
|
||||
let startTime = null;
|
||||
let endTime = null;
|
||||
|
||||
for (const w of words) {
|
||||
if (!currentSpeaker) currentSpeaker = w.speaker;
|
||||
if (startTime === null) startTime = w.start;
|
||||
endTime = w.end;
|
||||
|
||||
//speaker changing
|
||||
if (currentSpeaker !== w.speaker && currentSentence) {
|
||||
const lastEntry = result[result.length - 1];
|
||||
if (lastEntry && lastEntry.speaker === currentSpeaker) {
|
||||
lastEntry.sentence += " " + currentSentence;
|
||||
lastEntry.end = endTime;
|
||||
} else {
|
||||
result.push({
|
||||
speaker: currentSpeaker,
|
||||
sentence: currentSentence,
|
||||
start: startTime,
|
||||
end: endTime
|
||||
});
|
||||
}
|
||||
currentSentence = "";
|
||||
startTime = w.start;
|
||||
}
|
||||
currentSpeaker = w.speaker;
|
||||
currentSentence += (currentSentence ? " " : "") + w.text; //sentence beginning or not
|
||||
const lastWord = w.text.trim();
|
||||
const lastChar = lastWord.slice(-1);
|
||||
const isAbbreviation = ABBREVIATIONS.has(lastWord);
|
||||
|
||||
//sentence ending
|
||||
if (ENDINGS.includes(lastChar) && !isAbbreviation) {
|
||||
const lastEntry = result[result.length - 1];
|
||||
if (lastEntry && lastEntry.speaker === currentSpeaker) {
|
||||
lastEntry.sentence += " " + currentSentence;
|
||||
lastEntry.end = endTime;
|
||||
} else {
|
||||
result.push({
|
||||
speaker: currentSpeaker,
|
||||
sentence: currentSentence,
|
||||
start: startTime,
|
||||
end: endTime
|
||||
});
|
||||
}
|
||||
currentSentence = "";
|
||||
startTime = null;
|
||||
endTime = null;
|
||||
currentSpeaker = null;
|
||||
}
|
||||
}
|
||||
|
||||
// safe last sentence
|
||||
if (currentSentence) {
|
||||
const lastEntry = result[result.length - 1];
|
||||
if (lastEntry && lastEntry.speaker === currentSpeaker) {
|
||||
lastEntry.sentence += " " + currentSentence;
|
||||
lastEntry.end = endTime;
|
||||
} else {
|
||||
result.push({
|
||||
speaker: currentSpeaker,
|
||||
sentence: currentSentence,
|
||||
start: startTime,
|
||||
end: endTime
|
||||
});
|
||||
}
|
||||
}
|
||||
|
||||
// Output as Text
|
||||
const output = result.map(r =>
|
||||
`Sprecher ${r.speaker} [${r.start.toFixed(2)} - ${r.end.toFixed(2)}]: ${r.sentence}`
|
||||
);
|
||||
|
||||
// Output on cosole
|
||||
//console.log("\n------------\nMerged Transcription Result:\n", output, "\n------------\n");
|
||||
|
||||
try {
|
||||
const jsonPath = path.join(outputDir, "transcription_result.json");
|
||||
fs.writeFileSync(jsonPath, JSON.stringify(result, null, 2), "utf-8");
|
||||
|
||||
const txtPath = path.join(outputDir, "transcription_result.txt");
|
||||
fs.writeFileSync(txtPath, output.join("\n"), "utf-8");
|
||||
|
||||
console.log(`Summary successfully saved:\n- ${jsonPath}\n- ${txtPath}`);
|
||||
} catch (err) {
|
||||
console.error("Error saving Summary:", err);
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,121 @@
|
||||
const fs = require("fs");
|
||||
const path = require("path");
|
||||
|
||||
// Prepare output directory (always storage/transcriptionSummaries under project root)
|
||||
const outputDir = `${__dirname}/../../../storage/transcriptionSummaries`;
|
||||
if (!fs.existsSync(outputDir)) {
|
||||
fs.mkdirSync(outputDir, { recursive: true });
|
||||
}
|
||||
|
||||
//Speaker, Sentence, Start, End
|
||||
|
||||
module.exports = {
|
||||
name: "summarize-transcription2", // Unique name for our function that will later be used to get the function from the map via "mapFunctions.get("example").function()"
|
||||
type: "summarizer", // value used to differentiate each module to order them in the UI
|
||||
displayname: "Summarizer", // The displayname used within the UI
|
||||
async function(args) {
|
||||
let inputJson = args.json;
|
||||
|
||||
//JSON Path
|
||||
if (args.jsonPath) {
|
||||
try {
|
||||
const raw = fs.readFileSync(args.jsonPath, "utf-8");
|
||||
inputJson = JSON.parse(raw);
|
||||
} catch (e) {
|
||||
console.error("Failed to load JSON from file:", e);
|
||||
return { error: "Could not read JSON from file path." };
|
||||
}
|
||||
}
|
||||
// JSON parsen
|
||||
if (typeof inputJson === "string") {
|
||||
try {
|
||||
inputJson = JSON.parse(inputJson);
|
||||
} catch (e) {
|
||||
console.log("Invalid JSON in summarize-transcription");
|
||||
return { error: "Invalid JSON" };
|
||||
}
|
||||
}
|
||||
|
||||
const words = inputJson.words;
|
||||
if (!Array.isArray(words)) {
|
||||
return { error: "No words Array found" };
|
||||
}
|
||||
|
||||
const ENDINGS = [".", "!", "?"]; // '...' auch als Satzende ?
|
||||
const ABBREVIATIONS = new Set(["z.B.", "bzw.", "u.a.", "Dr.", "Mr.", "Mrs.", "Prof.", "etc."]); //TODO weitere Ergaenzen
|
||||
|
||||
const result = [];
|
||||
let currentSentence = "";
|
||||
let currentSpeaker = null;
|
||||
let startTime = null;
|
||||
let endTime = null;
|
||||
|
||||
for (const w of words) {
|
||||
if (!currentSpeaker) currentSpeaker = w.speaker;
|
||||
if (startTime === null) startTime = w.start;
|
||||
endTime = w.end;
|
||||
|
||||
//speaker changing
|
||||
if (currentSpeaker !== w.speaker && currentSentence) {
|
||||
result.push({
|
||||
speaker: currentSpeaker,
|
||||
sentence: currentSentence,
|
||||
start: startTime,
|
||||
end: endTime
|
||||
});
|
||||
currentSentence = "";
|
||||
startTime = w.start;
|
||||
}
|
||||
currentSpeaker = w.speaker;
|
||||
currentSentence += (currentSentence ? " " : "") + w.text; //sentence beginning or not
|
||||
const lastWord = w.text.trim();
|
||||
const lastChar = lastWord.slice(-1);
|
||||
const isAbbreviation = ABBREVIATIONS.has(lastWord);
|
||||
|
||||
//sentence ending
|
||||
if (ENDINGS.includes(lastChar) && !isAbbreviation) {
|
||||
result.push({
|
||||
speaker: currentSpeaker,
|
||||
sentence: currentSentence,
|
||||
start: startTime,
|
||||
end: endTime
|
||||
});
|
||||
currentSentence = "";
|
||||
startTime = null;
|
||||
endTime = null;
|
||||
currentSpeaker = null;
|
||||
}
|
||||
}
|
||||
|
||||
// safe last sentence
|
||||
if (currentSentence) {
|
||||
result.push({
|
||||
speaker: currentSpeaker,
|
||||
sentence: currentSentence,
|
||||
start: startTime,
|
||||
end: endTime
|
||||
});
|
||||
}
|
||||
|
||||
|
||||
// Output as Text
|
||||
const output = result.map(r =>
|
||||
`Sprecher ${r.speaker} [${r.start.toFixed(2)} - ${r.end.toFixed(2)}]: ${r.sentence}`
|
||||
);
|
||||
|
||||
// Output on cosole
|
||||
//console.log("\n------------\nMerged Transcription Result:\n", output, "\n------------\n");
|
||||
|
||||
try {
|
||||
const jsonPath = path.join(outputDir, "transcription_result.json");
|
||||
fs.writeFileSync(jsonPath, JSON.stringify(result, null, 2), "utf-8");
|
||||
|
||||
const txtPath = path.join(outputDir, "transcription_result.txt");
|
||||
fs.writeFileSync(txtPath, output.join("\n"), "utf-8");
|
||||
|
||||
console.log(`Summary successfully saved:\n- ${jsonPath}\n- ${txtPath}`);
|
||||
} catch (err) {
|
||||
console.error("Error saving Summary:", err);
|
||||
}
|
||||
}
|
||||
}
|
||||
+1
-2
@@ -5,7 +5,7 @@ import { fileURLToPath } from "url"; // To handle __dirname in ES modules
|
||||
|
||||
const __filename = fileURLToPath(import.meta.url); // Get current file path
|
||||
const __dirname = path.dirname(__filename); // Get current directory path
|
||||
const transcriptsDir = path.resolve(__dirname, "../../storage/transcriptions");
|
||||
const transcriptsDir = path.resolve(__dirname, "../../../storage/transcriptions");
|
||||
|
||||
|
||||
export class whisperLocal { // is called by transcribe.ts
|
||||
@@ -26,7 +26,6 @@ export class whisperLocal { // is called by transcribe.ts
|
||||
|
||||
async transcribe(audioPath: string): Promise<string> { //asyncronous function to transcribe audio
|
||||
return new Promise((resolve, reject) => {
|
||||
const transcriptsDir = path.resolve(__dirname, "../../../../storage/transcripts"); //storage directory for transcripts
|
||||
|
||||
if (!fs.existsSync(transcriptsDir)) { //if transcripts directory does not exist, create it
|
||||
fs.mkdirSync(transcriptsDir, { recursive: true });
|
||||
@@ -0,0 +1,123 @@
|
||||
require('dotenv').config();
|
||||
|
||||
const API_KEY = process.env.API_KEY;
|
||||
const BASE_URL = 'https://api.assemblyai.com/v2';
|
||||
|
||||
//---------------------------------------------------Upload audio---------------------------------------------------
|
||||
|
||||
async function uploadAudio(audioPath) {
|
||||
const audioData = fs.readFileSync(audioPath);
|
||||
|
||||
const response = await axios.post(`${BASE_URL}/upload`, audioData, {
|
||||
headers: {
|
||||
authorization: API_KEY,
|
||||
'content-type': 'application/octet-stream'
|
||||
}
|
||||
});
|
||||
|
||||
return response.data.upload_url;
|
||||
}
|
||||
|
||||
////---------------------------------------------------Extract session id---------------------------------------------------
|
||||
|
||||
function getSessionId(inputPath) {
|
||||
try {
|
||||
const parsed = new URL(inputPath);
|
||||
const base = path.basename(parsed.pathname);
|
||||
return base.replace(/\.[^.]+$/, '');
|
||||
} catch {
|
||||
return path.basename(inputPath, path.extname(inputPath));
|
||||
}
|
||||
}
|
||||
|
||||
//---------------------------------------------------Create transcript---------------------------------------------------
|
||||
|
||||
async function createTranscript(audioUrl) {
|
||||
const response = await axios.post(
|
||||
`${BASE_URL}/transcript`,
|
||||
{
|
||||
audio_url: audioUrl,
|
||||
speaker_labels: true,
|
||||
language_detection: true
|
||||
},
|
||||
{
|
||||
headers: {
|
||||
authorization: API_KEY,
|
||||
'content-type': 'application/json'
|
||||
}
|
||||
}
|
||||
);
|
||||
|
||||
return response.data.id;
|
||||
}
|
||||
|
||||
//---------------------------------------------------Poll transcript---------------------------------------------------
|
||||
|
||||
async function pollTranscript(transcriptId) {
|
||||
while (true) {
|
||||
const response = await axios.get(`${BASE_URL}/transcript/${transcriptId}`, {
|
||||
headers: { authorization: API_KEY }
|
||||
});
|
||||
|
||||
const status = response.data.status;
|
||||
|
||||
if (status === 'completed') return response.data;
|
||||
if (status === 'error') throw new Error(`Transcription failed: ${response.data.error}`);
|
||||
|
||||
await new Promise(res => setTimeout(res, 3000));
|
||||
}
|
||||
}
|
||||
|
||||
//---------------------------------------------------Save transcript---------------------------------------------------
|
||||
|
||||
function saveTranscript(transcript, sessionId) {
|
||||
const outputDir = path.join(__dirname, '../../../storage/transcripts');
|
||||
|
||||
if (!fs.existsSync(outputDir)) {
|
||||
fs.mkdirSync(outputDir, { recursive: true });
|
||||
}
|
||||
|
||||
const outputPath = path.join(outputDir, `${sessionId}.json`);
|
||||
fs.writeFileSync(outputPath, JSON.stringify(transcript, null, 2));
|
||||
|
||||
console.log(`Transcript saved: ${outputPath}`);
|
||||
}
|
||||
|
||||
//---------------------------------------------------Modul---------------------------------------------------
|
||||
|
||||
module.exports = {
|
||||
name: 'assembly',
|
||||
type: 'transcription',
|
||||
displayname: 'AssemblyAI',
|
||||
|
||||
async function(audioFileName) {
|
||||
try {
|
||||
// audioFileName ist nur "datei.mp3"
|
||||
const audioPath = path.join(
|
||||
__dirname,
|
||||
'../../../storage/audio',
|
||||
audioFileName
|
||||
);
|
||||
|
||||
let audioUrl;
|
||||
|
||||
if (/^https?:\/\//i.test(audioFileName)) {
|
||||
audioUrl = audioFileName;
|
||||
} else {
|
||||
if (!fs.existsSync(audioPath)) {
|
||||
throw new Error(`Audio file not found: ${audioPath}`);
|
||||
}
|
||||
audioUrl = await uploadAudio(audioPath);
|
||||
}
|
||||
|
||||
const transcriptId = await createTranscript(audioUrl);
|
||||
const transcript = await pollTranscript(transcriptId);
|
||||
|
||||
const sessionId = getSessionId(audioFileName);
|
||||
saveTranscript(transcript, sessionId);
|
||||
|
||||
} catch (error) {
|
||||
console.error('Transcription error:', error.message);
|
||||
}
|
||||
}
|
||||
};
|
||||
@@ -0,0 +1,133 @@
|
||||
import 'dotenv/config';
|
||||
import axios from 'axios';
|
||||
import fs from 'fs';
|
||||
import path from 'path';
|
||||
import { fileURLToPath } from 'url';
|
||||
|
||||
const __filename = fileURLToPath(import.meta.url);
|
||||
const __dirname = path.dirname(__filename);
|
||||
|
||||
const API_KEY = process.env.ASSEMBLYAI_API_KEY;
|
||||
const BASE_URL = 'https://api.assemblyai.com/v2';
|
||||
|
||||
/**
|
||||
* Uploads audio file to AssemblyAI
|
||||
*/
|
||||
async function uploadAudio(audioPath: string): Promise<string> {
|
||||
const audioData = fs.readFileSync(audioPath);
|
||||
|
||||
const response = await axios.post<{ upload_url: string }>(`${BASE_URL}/upload`, audioData, {
|
||||
headers: {
|
||||
'authorization': API_KEY,
|
||||
'content-type': 'application/octet-stream'
|
||||
}
|
||||
});
|
||||
|
||||
return response.data.upload_url;
|
||||
}
|
||||
|
||||
/**
|
||||
* Extract a session id (basename without extension) from a local path or a URL
|
||||
*/
|
||||
function getSessionId(inputPath: string): string {
|
||||
try {
|
||||
const parsed = new URL(inputPath);
|
||||
const base = path.basename(parsed.pathname);
|
||||
return base.replace(/\.[^.]+$/, '');
|
||||
} catch (err) {
|
||||
// not a URL, treat as local path
|
||||
return path.basename(inputPath, path.extname(inputPath));
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Creates transcription job with speaker diarization
|
||||
*/
|
||||
async function createTranscript(audioUrl: string): Promise<string> {
|
||||
const response = await axios.post<{ id: string }>(`${BASE_URL}/transcript`, {
|
||||
audio_url: audioUrl,
|
||||
speaker_labels: true,
|
||||
language_detection: true
|
||||
}, {
|
||||
headers: {
|
||||
'authorization': API_KEY,
|
||||
'content-type': 'application/json'
|
||||
}
|
||||
});
|
||||
|
||||
return response.data.id;
|
||||
}
|
||||
|
||||
/**
|
||||
* Polls transcript status until completed
|
||||
*/
|
||||
async function pollTranscript(transcriptId: string): Promise<any> {
|
||||
while (true) {
|
||||
const response = await axios.get<any>(`${BASE_URL}/transcript/${transcriptId}`, {
|
||||
headers: { 'authorization': API_KEY }
|
||||
});
|
||||
|
||||
const status = response.data.status;
|
||||
|
||||
if (status === 'completed') {
|
||||
return response.data;
|
||||
} else if (status === 'error') {
|
||||
throw new Error(`Transcription failed: ${response.data.error}`);
|
||||
}
|
||||
|
||||
// Wait 3 seconds before next poll
|
||||
await new Promise(resolve => setTimeout(resolve, 3000));
|
||||
}
|
||||
}
|
||||
|
||||
/**
|
||||
* Saves transcript to storage
|
||||
*/
|
||||
function saveTranscript(transcript: any, sessionId: string): void {
|
||||
const outputDir = path.join(__dirname, '..', '..', '..', 'storage', 'transcripts');
|
||||
|
||||
if (!fs.existsSync(outputDir)) {
|
||||
fs.mkdirSync(outputDir, { recursive: true });
|
||||
}
|
||||
|
||||
const outputPath = path.join(outputDir, `${sessionId}.json`);
|
||||
fs.writeFileSync(outputPath, JSON.stringify(transcript, null, 2));
|
||||
|
||||
console.log(`✅ Transcript saved: ${outputPath}`);
|
||||
}
|
||||
|
||||
export default {
|
||||
name: "assembly",
|
||||
type: "transcription",
|
||||
displayname: "AssemblyAI",
|
||||
run: async (audioPath: string) => {
|
||||
try {
|
||||
// Determine if audioPath is an external URL or a local file
|
||||
let audioUrl: string;
|
||||
if (/^https?:\/\//i.test(audioPath)) {
|
||||
console.log('🔗 Using external audio URL...');
|
||||
audioUrl = audioPath;
|
||||
} else {
|
||||
console.log('🔄 Uploading local audio...');
|
||||
if (!fs.existsSync(audioPath)) {
|
||||
throw new Error(`Audio file not found: ${audioPath}`);
|
||||
}
|
||||
audioUrl = await uploadAudio(audioPath);
|
||||
}
|
||||
|
||||
console.log('🔄 Creating transcript job...');
|
||||
const transcriptId = await createTranscript(audioUrl);
|
||||
|
||||
console.log('⏳ Waiting for transcription...');
|
||||
const transcript = await pollTranscript(transcriptId);
|
||||
|
||||
const sessionId = getSessionId(audioPath);
|
||||
saveTranscript(transcript, sessionId);
|
||||
|
||||
return transcript;
|
||||
} catch (error: any) {
|
||||
console.error('❌ Transcription error:', error.message);
|
||||
throw error;
|
||||
}
|
||||
}
|
||||
};
|
||||
@@ -1,8 +0,0 @@
|
||||
module.exports = {
|
||||
name:"assembly", // Unique name for our function that will later be used to get the function from the map via "mapFunctions.get("example").function()"
|
||||
type:"transcription", // value used to differentiate each module to order them in the UI
|
||||
displayname:"Assembly", // The displayname used within the UI
|
||||
async function(parameter){
|
||||
// TODO add code to actually process the audio file
|
||||
}
|
||||
}
|
||||
Submodule services/modules/transcription/local/whisper.cpp deleted from 999a7e0cbf
@@ -5,5 +5,16 @@ module.exports = {
|
||||
|
||||
// We are now calling the example function from the example folder
|
||||
mapFunctions.get("example").function("Startup")
|
||||
|
||||
let transcript = await mapFunctions.get("assembly").function('../../storage/audio/IMG_2978.wav');
|
||||
|
||||
let summary = await mapFunctions.get("summarize-transcription").function({jsonPath:'/Users/santa/Proj25/video2document/storage/transcripts/IMG_2978.json'});
|
||||
|
||||
|
||||
|
||||
|
||||
// mapFunctions.get("extraction-video-to-audio").function({inputVideoPath:"./a.mp4", outputType:"wav"})
|
||||
// mapFunctions.get("extraction-video-to-audio").function({inputVideoPath:"./b.mp4", outputType:"wav"})
|
||||
// mapFunctions.get("extraction-video-to-audio").function({inputVideoPath:"./b.mp4", outputType:"flac"})
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,52 @@
|
||||
// services/pipeline/jobs/transcribeLatest.ts
|
||||
import path from 'path';
|
||||
import fs from 'fs';
|
||||
import assembly from '../../modules/transcription/assembly';
|
||||
|
||||
/**
|
||||
* Finds the most recently modified .wav file in storage/audio/
|
||||
*/
|
||||
function getLatestWav(): string {
|
||||
const audioDir = path.join(process.cwd(), 'storage', 'audio');
|
||||
const files = fs.readdirSync(audioDir).filter(f => f.toLowerCase().endsWith('.wav'));
|
||||
if (files.length === 0) throw new Error('⚠️ No .wav file found in storage/audio');
|
||||
|
||||
const newest = files
|
||||
.map(f => ({ f, t: fs.statSync(path.join(audioDir, f)).mtimeMs }))
|
||||
.sort((a, b) => b.t - a.t)[0].f;
|
||||
|
||||
return path.join(audioDir, newest);
|
||||
}
|
||||
|
||||
/**
|
||||
* Full transcription pipeline according to the defined workflow:
|
||||
* 1. Audio Upload → AssemblyAI
|
||||
* 2. Job Creation (transcript_id)
|
||||
* 3. Polling Status (queued → processing → completed)
|
||||
* 4. Download Transcript JSON
|
||||
* 5. Storage: /transcripts/{session_id}.json
|
||||
*/
|
||||
async function main() {
|
||||
const audioPath = getLatestWav();
|
||||
|
||||
console.log('1️⃣ Audio Upload → AssemblyAI');
|
||||
console.log(' Source:', audioPath);
|
||||
|
||||
console.log('2️⃣ Job Creation (transcript_id)');
|
||||
console.log('3️⃣ Polling Status (queued → processing → completed)');
|
||||
console.log('4️⃣ Download Transcript JSON');
|
||||
console.log('5️⃣ Storage: /transcripts/{session_id}.json');
|
||||
|
||||
// Execute the transcription process via the AssemblyAI module
|
||||
const result = await assembly.run(audioPath);
|
||||
|
||||
console.log('✅ Transcription completed successfully');
|
||||
console.log('🆔 Transcript ID:', result.id);
|
||||
console.log('📁 Transcript file saved under: storage/transcripts/');
|
||||
}
|
||||
|
||||
// Entry point
|
||||
main().catch((err) => {
|
||||
console.error('❌ Transcription pipeline failed:', err.message || err);
|
||||
process.exit(1);
|
||||
});
|
||||
@@ -0,0 +1,14 @@
|
||||
import 'dotenv/config';
|
||||
import assemblyModule from '../../services/modules/transcription-remote/assembly.ts';
|
||||
|
||||
// Test: URL passed as argument OR local file ./storage/audio/test.wav
|
||||
const audioPath = process.argv[2] || './storage/audio/test.wav';
|
||||
|
||||
assemblyModule.run(audioPath)
|
||||
.then(result => {
|
||||
console.log('✅ Success!');
|
||||
console.log('Transcript ID:', result.id);
|
||||
})
|
||||
.catch(error => {
|
||||
console.error('❌ Error:', error?.message || error);
|
||||
});
|
||||
@@ -0,0 +1,18 @@
|
||||
require('dotenv').config();
|
||||
const path = require('path');
|
||||
const assemblyModule = require('../../services/modules/transcription-remote/assembly.js');
|
||||
|
||||
// Audio-Datei oder URL aus Kommandozeile, Standard: test.wav
|
||||
const audioPath = process.argv[2] || './storage/audio/IMG_2978.wav';
|
||||
|
||||
(async () => {
|
||||
try {
|
||||
const transcript = await assemblyModule.run(audioPath);
|
||||
|
||||
console.log('Transcription succesful');
|
||||
console.log('Transcript ID:', transcript?.id);
|
||||
console.log('Speaker labels:', transcript?.utterances?.length || 0);
|
||||
} catch (error) {
|
||||
console.error('Error in Transcription:', error?.message || error);
|
||||
}
|
||||
})();
|
||||
File diff suppressed because it is too large
Load Diff
@@ -0,0 +1,12 @@
|
||||
const fs = require('fs');
|
||||
const transSummarizer = require("../../services/modules/jsonTools/transcriptionSummarizer.js");
|
||||
|
||||
// JSON-Datei laden
|
||||
const inputJson = JSON.parse(fs.readFileSync("./testFile.json", "utf8"));
|
||||
|
||||
// Übergabe an den Summarizer
|
||||
transSummarizer.function({
|
||||
json: inputJson
|
||||
});
|
||||
|
||||
|
||||
Reference in New Issue
Block a user