Compare commits

...

48 Commits

Author SHA1 Message Date
Aarthi Manivannan, Premanathan Aarthi Manivannan c00640c15d Update file meeting_report_prompt.txt 2025-12-14 11:52:21 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan d80941ca65 Delete .gitkeep 2025-12-14 11:45:47 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 0e4147e893 Delete meeting_report_template.md 2025-12-14 11:45:38 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 9ab69b4b36 Delete .gitkeep 2025-12-14 11:44:59 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan cd61d8e09b Delete transcript.txt 2025-12-14 11:44:51 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 893546d142 Delete Example_output.pdf.pdf 2025-12-14 11:44:41 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 9098dafbd5 Delete .gitkeep 2025-12-14 11:44:16 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan d68192de8a Delete meeting_report_prompt.txt 2025-12-14 11:44:06 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 5a23ec9c2f Delete .gitkeep 2025-12-14 11:43:24 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan a0237ade55 Delete meeting_report_structure.json 2025-12-14 11:39:09 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 10e3c902c5 Delete meeting_report_best_practices.md 2025-12-14 11:37:32 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan b157a90671 Upload New File 2025-11-20 12:24:46 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 6931df22e0 Add new file 2025-11-20 12:12:42 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 9eaabe80b6 Add new directory 2025-11-20 12:09:35 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan db2192dc30 Add new file 2025-11-20 12:02:32 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 76129982c3 Update file meeting_report_prompt.txt 2025-11-20 11:59:16 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan ea769d3aec Add new file 2025-11-20 11:41:14 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 21f4fe95d7 Add new file 2025-11-20 11:38:32 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 8e07bcf028 Add new file 2025-11-20 11:35:47 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan fd0798872a Add new directory 2025-11-20 11:32:46 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 0ea3fba436 Add new directory 2025-11-20 11:32:20 +01:00
Aarthi Manivannan, Premanathan Aarthi Manivannan 5615f7fd25 research and design 2025-11-20 11:29:33 +01:00
Hughes, Mike 283b4ed6af Merge branch 'develop' into 'main'
Implemented the general modular framework.

See merge request proj-wise2526-video2document/video2document!22
2025-11-15 15:14:24 +01:00
Spanier, Pit 2edc7f8351 Merge branch 'fix/transcription-module-fix' into 'develop'
New Folder structure

See merge request proj-wise2526-video2document/video2document!21
2025-11-15 15:11:35 +01:00
MikeHughes-BIN 6083773f88 New Folder structure 2025-11-15 14:45:13 +01:00
Spanier, Pit 444d408480 Merge branch 'feature/fixing-the-program' into 'develop'
fixed the program by moving the example module back to where the program can ACTUALLY load it

See merge request proj-wise2526-video2document/video2document!20
2025-11-14 14:30:16 +01:00
emily d9eacafc3a fixed the program by moving the example module back to where the program can ACTUALLY load it 2025-11-14 14:28:11 +01:00
Hughes, Mike ab737f0dc9 Merge branch 'feature/12-externe-transkription-s2-02b' into 'develop'
feat(S2-02b): Implement AssemblyAI external transcription with speaker diarization

See merge request proj-wise2526-video2document/video2document!18
2025-11-13 17:38:39 +01:00
MikeHughes-BIN 79e0c48755 Reduced Number of test paths to avoid redundancy 2025-11-13 17:35:40 +01:00
MikeHughes-BIN 9254ddc57f Changed the Folder Structure for better maintainability 2025-11-13 17:34:22 +01:00
Azeufack Noupeu Willy c021272ca4 merge: Resolve conflicts with develop 2025-11-13 15:22:33 +01:00
Azeufack Noupeu Willy e7e97a7f60 feat(S2-02b): Implement AssemblyAI external transcription with speaker diarization
- Add assembly.ts module for REST API transcription via AssemblyAI
- Implement 5-step pipeline: upload → create job → poll status → download → save
- Enable speaker_labels for diarization (Speaker A, B, C...)
- Add millisecond-precision timestamps for each utterance
- Store JSON transcripts in storage/transcripts/{session_id}.json
- Add axios, dotenv dependencies
- Add transcribeLatest.ts helper for quick testing

User Story: S2-02b - Externe Transkription per REST API
2025-11-13 13:07:18 +01:00
Spanier, Pit 80392874bb Merge branch 'feature/combination-of-all-the-things' into 'develop'
Fixed an error in checkBoxes function and uploadContainer "drop" listener...

See merge request proj-wise2526-video2document/video2document!15
2025-11-12 20:56:11 +01:00
emily 94f390f28b Implemented audio extractor into IPC call, sooooooo, frontend and backend are now officially linked successfully 2025-11-12 20:28:55 +01:00
emily 73f6fa7524 fixed the code so that it now returns the actual path of the file 2025-11-12 20:24:45 +01:00
emily 87e3368a9a fixed code 2025-11-12 20:04:48 +01:00
emily 6d9c94c685 removed some debug console outputs 2025-11-12 20:03:09 +01:00
emily fbd5368223 Implemented functionality to have the UI be able to communicate with the backend 2025-11-12 20:00:25 +01:00
emily 449255dda1 Merge commit 'c8cbd4e92a034fa108d06337a3e63d24e529432f' into feature/combination-of-all-the-things 2025-11-12 19:55:50 +01:00
eric.minning c8cbd4e92a Fixed an error in checkBoxes function and uploadContainer "drop" listener regarding file path testing 2025-11-12 18:52:28 +01:00
eric.minning 0f54edb0aa Removed the variables at the start of the script.js file 2025-11-12 18:44:27 +01:00
eric.minning 54f1f6c135 Added the check if a path is selected and the call to the transcript functio. 2025-11-12 18:16:27 +01:00
Verena Schulz a32e7e5744 Checkboxes check (with alert) 2025-11-12 17:39:02 +01:00
Verena Schulz d9e96316c0 Language selection implemented and UI reworked 2025-11-11 18:31:01 +01:00
eric.minning da3a8c7d8c Added try/catch handling and some comments 2025-11-11 16:02:07 +01:00
emily dfe91bb15a Merge commit '3a8963b49d96d0746b64244fa17458d4ae68ec45' into feature/combination-of-all-the-things 2025-11-10 17:01:30 +01:00
emily 92043440fe Combined Frontend with backend, and implemented the extractor function into the module structure 2025-11-10 16:18:57 +01:00
emily bd8ed1bf81 added mp4, flac and wav files to gitignore 2025-11-10 16:12:59 +01:00
30 changed files with 2001 additions and 208 deletions
+5 -1
View File
@@ -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

+16 -9
View File
@@ -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>
+14 -3
View File
@@ -2,6 +2,17 @@
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)}
})
} catch (error) {
console.log("Error in preload.js");
}
+27 -18
View File
@@ -1,22 +1,31 @@
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 = filePath + "";
if(holdy.endsWith(".mp4") || holdy.endsWith(".mov") || holdy.endsWith(".avi") || holdy.endsWith( ".mkv")){
console.log(filePath)
const files1 = e.dataTransfer.files;
handleFiles(files1);
}
} catch (error) {
console.log("Error in renderer.js with the listerner for the drop function");
}
})
+103 -25
View File
@@ -1,37 +1,115 @@
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');
manualBtn.addEventListener('click', () => {
fileInput.click();
//listener for the file explorer search
manualUploadBtn.addEventListener('click', () => {
try {
videoUpload.click();
} catch (error) {
console.log("Error in manualBtn EventListener 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]);
if(pathTest.endsWith(".mp4") || holdy.endsWith(".mov") || holdy.endsWith(".avi") || holdy.endsWith( ".mkv")){
window.extractor.extract({inputVideoPath: pathTest, outputType:"wav"})
}
} 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 handleFiles(files) {
if (files.length > 0) {
const file = files[0];
if (file.type.startsWith('video/')) {
fileInput.files = files;
fileName.textContent = `Chosen video: ${file.name}`;
//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: ${file.name}`;
}
}
} catch (error) {
console.log("Error in script.js handleFiles function");
}
}
//function to regulate the progress on the progressbar
function updateProgressBar(bar, value){
value = Math.round(value);
bar.querySelector(".progress_fill").style.width = `${value}%`;
bar.querySelector(".progress_text").textContent = `${value}%`;
try {
value = Math.round(value);
bar.querySelector(".progress_fill").style.width = `${value}%`;
bar.querySelector(".progress_text").textContent = `${value}%`;
} catch (error) {
console.log("Error in scripts.js updateProgressBar function");
}
}
+19 -4
View File
@@ -5,7 +5,7 @@ body {
justify-content: center;
align-items: center;
height: 100vh;
background-color: #444;
background-color: #555;
gap: 15px;
margin: 0;
}
@@ -75,6 +75,7 @@ gap: 5px;
.checkbox-group {
margin-top: 15px;
margin-bottom: 15px;
display: flex;
flex-direction: column;
gap: 10px;
@@ -92,14 +93,14 @@ gap: 5px;
}
.mitte {
background-color: #eaf0ff;
background-color: #f2f3f4;
display: flex;
flex-direction: column;
align-items: center;
padding: 5% 50px;
margin-top: 20px;
gap: 10px;
border: 2px;
border: 1px;
border-color: black;
border-style: solid;
}
@@ -112,7 +113,7 @@ h1 {
position: relative;
width: 210px;
height: 30px;
background: darkslategray;
background: rgb(42, 46, 78);
border-radius: 5px;
overflow: hidden;
}
@@ -129,4 +130,18 @@ h1 {
top: 50%;
right: 5px;
transform: translateY(-50%);
color: white;
}
.flagsBtns {
display: flex;
justify-content: flex-end;
}
.de_Btn, .eng_Btn, .in_Btn {
padding: 8px 16px;
color: white;
border: none;
border-radius: 8px;
cursor: pointer;
}
+22 -1
View File
@@ -55,4 +55,25 @@ rl.on("line", data =>{
// ----------------------------------------------------------- ELECTRON ----------------------------------------------------------- //
// TODO - Add Electron support to the project
let mainWindow;
function createWindow() {
mainWindow = new electron.BrowserWindow({
width: 800,
height: 600,
webPreferences: {
nodeIntegration: false,
contextIsolation: true,
preload: `${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)
})
+1375 -3
View File
File diff suppressed because it is too large Load Diff
+6 -5
View File
@@ -1,10 +1,12 @@
{
"type": "module",
"dependencies": {
"@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",
"express": "^5.1.0"
"fluent-ffmpeg": "^2.1.3"
},
"devDependencies": {
"@types/cli-progress": "^3.11.6",
@@ -22,6 +24,7 @@
"test": "tests"
},
"scripts": {
"start": "electron main.js",
"test": "echo \"Error: no test specified\" && exit 1"
},
"repository": {
@@ -31,5 +34,3 @@
"author": "",
"license": "ISC"
}
+109
View File
@@ -0,0 +1,109 @@
Generate a structured meeting report in MARKDOWN using STRUCTURE and STYLE.
Output ONLY the final .md document — no meta comments, no explanations.
Follow exactly the STRUCTURE defined below.
Follow exactly the STYLE rules.
Use timestamps in HH:MM:SS format.
If information is missing, use: UNKLAR:<reason>.
==================== STRUCTURE & RULES ====================
{
"FORMAT": "markdown",
"STRUCTURE": {
"titlepage": [
"title",
"date",
"start",
"end",
"duration",
"location",
"host",
"participants"
],
"toc": "[section](#anchor) — HH:MM:SS",
"section": {
"h2": "<topic> — HH:MM:SS",
"summary": "exactly 1 concise sentence",
"key_points": "maximum 5 bullet points; quotes optional",
"decisions": "list items formatted as: decision text | owner | due date",
"actions": "markdown table: id | task | owner | due | status"
},
"exec_summary": "exactly 3 short sentences",
"consolidated": [
"decisions",
"actions"
],
"appendix": "optional"
},
"STYLE": {
"tone": "neutral, concise, professional",
"ts_format": "HH:MM:SS",
"no_meta": true
},
"PROCESS": {
"timestamps": "use transcript timestamps if present; otherwise estimate minimal",
"speakers": "use names if available; else Speaker X",
"long_transcripts": "split → summarize → merge",
"unclear": "UNKLAR:<reason>"
},
"PROMPT_SNIPPET": "Generate meeting report in markdown using STRUCTURE and STYLE. Output only the report."
}
============================================================
Insert all generated content into the following MARKDOWN TEMPLATE:
# {{title}}
**Date:** {{date}}
**Start:** {{start}}
**End:** {{end}}
**Duration:** {{duration}}
**Location:** {{location}}
**Host:** {{host}}
**Participants:** {{participants}}
---
## Table of Contents
{{toc}}
---
## Executive Summary
{{exec_summary}}
---
## Sections
{{sections}}
---
## Consolidated Decisions
{{consolidated_decisions}}
---
## Consolidated Actions
{{consolidated_actions}}
---
## Appendix
{{appendix}}
============================================================
Final Requirement:
Output ONLY the completed Markdown meeting report.
+8
View File
@@ -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');
-5
View File
@@ -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
-23
View File
@@ -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: ${videoPath}`);
await extractAudioFromVideo(videoPath); // Call the extraction function (ffmpegExtractor.ts in services/modules/extraction)
console.log("Audio extraction completed successfully.");
} catch (err) {
console.error("Audio extraction failed:", err);
process.exit(1);
}
})();
-18
View File
@@ -1,18 +0,0 @@
import { whisperLocal } from "../services/modules/transcription/local/whisperLocal.ts";
const audioPath = process.argv[2];
if (!audioPath) {
console.error("Please provide an audio file path as argument.");
process.exit(1);
}
const whisper = new whisperLocal();
(async () => {
try {
const text = await whisper.transcribe(audioPath);
console.log(text);
} catch (err) {
console.error("Transcription failed:", err);
}
})();
@@ -0,0 +1,93 @@
// Ensure ffmpeg binary is available
if (!ffmpegPath) {
throw new Error('FFmpeg binary not found!');
}
ffmpeg.setFfmpegPath(ffmpegPath);
// Prepare output directory (always storage/audio under project root)
const outputDir = `${__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));
}
@@ -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 +1,133 @@
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;
}
}
};
Submodule services/modules/transcription/local/whisper.cpp deleted from 999a7e0cbf
+4
View File
@@ -5,5 +5,9 @@ module.exports = {
// We are now calling the example function from the example folder
mapFunctions.get("example").function("Startup")
// 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);
});
+14
View File
@@ -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);
});
View File