the old report type was still used

The requirements needed to include puppeteer and html-to-docs

See merge request proj-wise2526-video2document/video2document!44

.gitignore

@@ -963,9 +963,23 @@ app.*.symbols
 !/packages/flutter_tools/test/data/dart_dependencies_test/**/.packages
 !/dev/ci/**/Gemfile.lock
 
-#Storage files
+# Storage files
-storage/
+storage/documents/*
+storage/transcriptionSummaries/*
+storage/transcripts/*
+storage/video/*
+# The inverse for the .gitkeep files, to make sure the folders are there but not the local files
+!storage/documents/.gitkeep
+!storage/transcriptionSummaries/.gitkeep
+!storage/transcripts/.gitkeep
+!storage/video/.gitkeep
 
 *.mp4
+*.webm
+*.mp3
+*.mov
 *.wav
 *.flac
+
+
+!testvideo.mp4

.gitlab-ci.yml

@@ -1,33 +1,27 @@
-workflow:
+image: node:latest
-  rules:
-    # Run the pipeline for merge requests or when committing to a branch
-    - if: $CI_PIPELINE_SOURCE == "merge_request_event"
-    - if: $CI_COMMIT_BRANCH
-
-image: python:3.14.0
 
 stages:
-  - setup
+  # - install
   - test
 
-setup_environment:
+# job-install:
-  stage: setup
+#   stage: install
-  script:
+#   script:
-    - pip install --upgrade pip
+#     - npm install
-    - pip install -r requirements.txt
+#   artifacts:
-    - echo "Dependencies installed successfully."
+#     untracked: false
+#     when: on_success
+#     access: all
+#     expire_in: "30 days"
+#     paths:
+#       - node_modules
 
-  only:
-    - main
-    - feature/ci-pipeline-s1-09a-1  # You can add more branches if needed
 
-test_app:
+
+job-test:
   stage: test
   script:
-    - echo "Running V2D Framework basic test..."
+    - npm install
-    - python -m unittest discover || echo "No tests found."
+    - echo "ASSEMBLYAI_API_KEY=$apikey_assembly" > .env
-
+    - echo "GOOGLE_API_KEY=$apikey_gemini" >> .env
-  only:
+    - npm test
-    - main
-    - feature/ci-pipeline-s1-09a-1
-

main.js

@@ -141,7 +141,7 @@ electron.ipcMain.on("file_submit", async (event, args) => {
 
         console.log("\n\n Running the Video to Audio Extractor");
         // This code handles the Video to Audio extraction module call
-        await mapFunctions.get("module-handler").function(args.video.module, {inputVideoPath: args.video.inputVideoPath, outputType: args.video.outputType}).then(resp => {
+        await mapFunctions.get("module-handler").function(args.video.module, {inputVideoPath: args.video.inputVideoPath, outputType: mapFunctions.get(args.transcription.module).audioformat}).then(resp => {
             console.log(resp);
             audiopath = resp
             curstep++
@@ -183,7 +183,9 @@ electron.ipcMain.on("file_submit", async (event, args) => {
         // TODO implement documentation module
         // This code handles the Text to Document processing module call
         for (let i = 0; i < args.document.styles.length; i++) {
-            await mapFunctions.get("module-handler").function(args.document.module, {prompt: args.document.styles[i].prompt, transcript: transcriptpath}).then(resp => {
+            console.log(`\n\n Running the LLM for Document Style ${i+1}`);
+
+            await mapFunctions.get("module-handler").function(args.document.module, {inputTranscriptPath: transcriptpath, documentTypePath: "./storage/documentType/standard_meeting_report.txt", language: "en"}).then(resp => {
                 console.log(resp);
                 transcriptpath = resp
                 curstep++
@@ -194,6 +196,24 @@ electron.ipcMain.on("file_submit", async (event, args) => {
             })
         }
 
+        // TODO actually implement this functionality
+        // Module to get the first few lines for each speaker to send to the frontend
+//         await mapFunctions.get("speaker-getter-idfk").function(transcriptpath).then(resp => {
+//             console.log(resp);
+//             transcriptpath = resp
+//             curstep++
+//             mainWindow.webContents.send("progress", {curstep:curstep, totalsteps:totalsteps})
+
+// // {
+// // speakerA: {source: "Pfad zur Audio File"},
+// // speakerB:.....
+// // }
+            mainWindow.webContents.send("speakers", {speakerA:"pfad1", speakerB:"pfad2"})
+//         }).catch(err => {
+//             mainWindow.webContents.send("error", err)
+//             return
+//         })
+
 
     } catch (error) {
         console.log(error);

package.json

@@ -3,12 +3,14 @@
     "@google/genai": "^1.30.0",
     "@types/axios": "^0.9.36",
     "axios": "^1.13.2",
-    "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"
+    "fluent-ffmpeg": "^2.1.3",
+    "html-to-docx": "^1.8.0",
+    "mocha": "^11.7.5",
+    "puppeteer": "^24.33.0"
   },
   "devDependencies": {
     "@types/cli-progress": "^3.11.6",
@@ -27,7 +29,7 @@
   },
   "scripts": {
     "start": "electron main.js",
-    "test": "echo \"Error: no test specified\" && exit 1"
+    "test": "mocha ./test/unit/test.js"
   },
   "repository": {
     "type": "git",

requirements.txt

@@ -1,3 +0,0 @@
-fastapi
-uvicorn
-pytest

requires.js

@@ -6,17 +6,19 @@ platform        = process.platform
 mainDir         = __dirname
 fs              = require("fs")
 readline        = require("readline")
+
+puppeteer       = require("puppeteer")
+htmltodocx    = require("html-to-docx")
 config          = require("./config/config")
 
 ffmpegPath = require('ffmpeg-static');
 ffmpeg = require('fluent-ffmpeg');
 path = require('path');
-cliProgress = require('cli-progress');
+// cliProgress = require('cli-progress');
 
 // { app, BrowserWindow, ipcMain, dialog } = require('electron');
 
 electron = require('electron');
-genai    = require("@google/genai");
 
 axios = require("axios")
 

scripts/show-models.js

@@ -0,0 +1,47 @@
+const fs = require('fs');
+const path = require('path');
+
+//node show-models.js, remember to set SAIA_API_KEY in your environment before running the script
+
+const SAIA_API_KEY = process.env.SAIA_API_KEY;
+const SAIA_MODELS_URL = "https://chat-ai.academiccloud.de/v1/models";
+
+// Script to list available models
+(async () => {
+    if (!SAIA_API_KEY) {
+        console.error("ERROR: SAIA_API_KEY environment variable is not set!");
+        process.exit(1);
+    }
+
+    console.log("Fetching available models from SAIA...\n");
+
+    try {
+        const response = await fetch(SAIA_MODELS_URL, {
+            method: "GET",
+            headers: {
+                "Authorization": `Bearer ${SAIA_API_KEY}`,
+                "Accept": "application/json"
+            }
+        });
+
+        if (!response.ok) {
+            const text = await response.text();
+            throw new Error(`SAIA API error (${response.status}): ${text}`);
+        }
+
+        const data = await response.json();
+        
+        console.log("Available models:");
+        console.log(JSON.stringify(data, null, 2));
+        
+        if (data.data && Array.isArray(data.data)) {
+            console.log("\n\nModel IDs:");
+            data.data.forEach(model => {
+                console.log(`- ${model.id}`);
+            });
+        }
+
+    } catch (error) {
+        console.error("Error fetching models:", error);
+    }
+})();

services/modules/convert/convert.js

@@ -0,0 +1,194 @@
+const fs = require('fs');
+const path = require('path');
+const puppeteer = require('puppeteer');
+const htmlToDocx = require('html-to-docx');
+const { execSync } = require('child_process');
+const os = require('os');
+
+const outputDir = path.join(__dirname, "../../../storage/documents");
+
+if (!fs.existsSync(outputDir)) {
+    fs.mkdirSync(outputDir, { recursive: true });
+}
+
+async function showSaveDialog(defaultName, format) {
+    const platform = os.platform();
+    
+    if (platform === 'darwin') {
+        // macOS
+        const applescript = `
+            set defaultName to "${defaultName}.${format}"
+            set theFile to choose file name with prompt "Dokument speichern als:" default name defaultName
+            POSIX path of theFile
+        `;
+        
+        try {
+            const result = execSync(`osascript -e '${applescript}'`, { encoding: 'utf8' });
+            return result.trim();
+        } catch (err) {
+            if (err.status === 1) return null; // User canceled
+            throw err;
+        }
+    } else if (platform === 'win32') {
+        // Windows
+        const powershell = `
+            Add-Type -AssemblyName System.Windows.Forms
+            $dialog = New-Object System.Windows.Forms.SaveFileDialog
+            $dialog.FileName = "${defaultName}.${format}"
+            $dialog.Filter = "${format.toUpperCase()} Dateien (*.${format})|*.${format}|Alle Dateien (*.*)|*.*"
+            $dialog.Title = "Dokument speichern als"
+            $result = $dialog.ShowDialog()
+            if ($result -eq 'OK') { $dialog.FileName }
+        `;
+        
+        try {
+            const result = execSync(`powershell -Command "${powershell.replace(/\n/g, '; ')}"`, { 
+                encoding: 'utf8' 
+            });
+            return result.trim() || null;
+        } catch (err) {
+            throw err;
+        }
+    } else {
+        // Linux - zenity oder kdialog
+        try {
+            const result = execSync(
+                `zenity --file-selection --save --confirm-overwrite --filename="${defaultName}.${format}"`, 
+                { encoding: 'utf8' }
+            );
+            return result.trim();
+        } catch (err) {
+            try {
+                const result = execSync(
+                    `kdialog --getsavefilename . "${defaultName}.${format}"`, 
+                    { encoding: 'utf8' }
+                );
+                return result.trim();
+            } catch (err2) {
+                // Fallback
+                return path.join(os.homedir(), 'Downloads', `${defaultName}.${format}`);
+            }
+        }
+    }
+}
+
+const module_exports = {
+    name: "htmlDocumentConverter",
+    type: "converter",
+    displayname: "HTML Document Converter",
+    description: "Converts LLM-generated HTML to PDF, DOCX, TXT, or HTML",
+
+    /**
+     * Main conversion function
+     * @param {Object} options
+     * @param {string} options.inputPath - Path to the HTML input
+     * @param {string} options.format - 'pdf' | 'docx' | 'html' | 'txt'
+     * @param {string} [options.outputName] - Optional output filename (without extension)
+     * @param {boolean} [options.showDialog] - Show save dialog (default: false in module mode, true in CLI mode)
+     */
+    async convert({ inputPath, format = 'pdf', outputName, showDialog = false }) {
+        if (!fs.existsSync(inputPath)) {
+            throw new Error(`Input file not found: ${inputPath}`);
+        }
+
+        const ext = path.extname(inputPath).toLowerCase();
+        const baseName = outputName || path.basename(inputPath, ext);
+        
+        let outputFile;
+        
+        if (showDialog) {
+            // Zeige nativen Dialog
+            outputFile = await showSaveDialog(baseName, format);
+            if (!outputFile) {
+                console.log('Speichervorgang abgebrochen.');
+                return null;
+            }
+        } else {
+            // Nutze Standard-Ausgabeverzeichnis
+            outputFile = path.join(outputDir, `${baseName}.${format.toLowerCase()}`);
+        }
+
+        let htmlContent = fs.readFileSync(inputPath, 'utf8');
+
+        // Remove <think> tags if present
+        htmlContent = htmlContent.replace(/<think>[\s\S]*?<\/think>/gi, '');
+
+        switch (format.toLowerCase()) {
+            case 'html':
+                fs.writeFileSync(outputFile, htmlContent, 'utf8');
+                break;
+            case 'pdf':
+                await this.htmlToPDF(htmlContent, outputFile);
+                break;
+            case 'docx':
+                await this.htmlToDOCX(htmlContent, outputFile);
+                break;
+            case 'txt':
+                fs.writeFileSync(outputFile, this.htmlToTXT(htmlContent), 'utf8');
+                break;
+            default:
+                throw new Error(`Unsupported format: ${format}`);
+        }
+
+        console.log(`Erfolgreich gespeichert: ${outputFile}`);
+        return outputFile;
+    },
+
+    // HTML → PDF
+    async htmlToPDF(html, outputPath) {
+        const browser = await puppeteer.launch({ 
+            headless: true,
+            args: ['--no-sandbox', '--disable-setuid-sandbox']
+        });
+        const page = await browser.newPage();
+        await page.setContent(html, { waitUntil: 'networkidle0' });
+        await page.pdf({ 
+            path: outputPath, 
+            format: 'A4', 
+            printBackground: true,
+            margin: { top: '20mm', right: '20mm', bottom: '20mm', left: '20mm' }
+        });
+        await browser.close();
+    },
+
+    // HTML → DOCX
+    async htmlToDOCX(html, outputPath) {
+        const buffer = await htmlToDocx(html);
+        fs.writeFileSync(outputPath, buffer);
+    },
+
+    // HTML → TXT (rudimentär)
+    htmlToTXT(html) {
+        return html.replace(/<[^>]*>/g, '').replace(/\s+\n/g, '\n').trim();
+    }
+};
+
+module.exports = module_exports;
+
+// CLI usage mit Dialog
+if (require.main === module) {
+    (async () => {
+        const args = process.argv.slice(2);
+        if (args.length < 1) {
+            console.log('Usage: node htmlDocumentConverter.js <input.html> [format]');
+            console.log('Formats: pdf (default), docx, html, txt');
+            console.log('');
+            console.log('Ein nativer "Speichern unter" Dialog wird automatisch geöffnet.');
+            process.exit(1);
+        }
+
+        const inputPath = args[0];
+        const format = args[1] || 'pdf';
+
+        try {
+            await module_exports.convert({ 
+                inputPath, 
+                format, 
+                showDialog: true  
+            });
+        } catch (err) {
+            console.error('Konvertierung fehlgeschlagen:', err.message);
+            process.exit(1);
+        }
+    })();
+}

services/modules/extraction/ffmpegExtractor.js

@@ -24,17 +24,11 @@ module.exports = {
             outputType: String      // Audio file output format
         }
         */
-        let progressBar = new cliProgress.SingleBar({
-            format: 'Processing |{bar}| {percentage}% | {timemark}',
-            barCompleteChar: '\u2588',
-            barIncompleteChar: '\u2591',
-            hideCursor: true
-        });
         try {
             return new Promise((resolve, reject) => {
-                this.extractAudioFromVideo(parameter.inputVideoPath, progressBar, parameter.outputType)
+                this.extractAudioFromVideo(parameter.inputVideoPath, parameter.outputType)
                     .then((resp) => resolve(resp))
-                    .catch((err) => console.error(err));
+                    .catch((err) => {reject(err)});
             })
         } catch (error) {
             console.log(parameter.outputType);
@@ -52,7 +46,7 @@ module.exports = {
      * - Shows CLI progress bar
      * - Handles errors gracefully (without errors)
      */
-    extractAudioFromVideo: async function (videoFilePath, progressBar, outputType){
+    extractAudioFromVideo: async function (videoFilePath, outputType){
         let inputVideoName = path.basename(videoFilePath, path.extname(videoFilePath));
         let outputAudioPath = path.join(outputDir, `${inputVideoName}.${outputType}`);
 
@@ -63,28 +57,17 @@ module.exports = {
                 // .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(outputAudioPath);
                 })
                 .on('error', (err) => {
-                    progressBar.stop();
+                    // console.error(`failed_audio_extraction on type ${outputType}: ${err.message}`);
-                    console.error(`failed_audio_extraction on type ${outputType}: ${err.message}`);
                     reject(err);
                 })
                 .save(outputAudioPath);
 
             } catch (error) {
-                console.log();
+                // console.log(error);
             }
         });
     }

services/modules/jsonTools/transcriptionSummarizer.js

@@ -21,30 +21,36 @@ module.exports = {
                     const raw = fs.readFileSync(args.jsonPath, "utf-8");
                     inputJson = JSON.parse(raw);
                 } catch (e) {
-                    console.error("Failed to load JSON from file:", e);
+                    // console.error("Failed to load JSON from file:", e);
-                    return { error: "Could not read JSON from file path." };
+                    reject(e)
+                    return
                 }
             }
             // JSON parsen
             if (typeof args === "string") {
                 try {
-                    await new Promise((res) => {
+                    await new Promise((res, rej) => {
                         fs.readFile(args, 'utf8', function (err, data) {
-                            if (err) throw err;
+                            if (err){
+                                rej(err)
+                                return
+                            }
                             inputJson = JSON.parse(data);
                             res()
                         });
                     })
                 } catch (e) {
-                    console.log("Invalid JSON in summarize-transcription");
+                    // console.log("Invalid JSON in summarize-transcription");
-                    console.log(e)
+                    // console.log(e)
-                    return { error: "Invalid JSON" };
+                    reject(e)
+                    return
                 }
             }
 
             const words = inputJson.words;
             if (!Array.isArray(words)) {
-                return { error: "No words Array found" };
+                    reject("No words Array found")
+                    return
             }
 
             const ENDINGS = [".", "!", "?"]; // '...' auch als Satzende ?
@@ -136,11 +142,11 @@ module.exports = {
                 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}`);
+                // console.log(`Summary successfully saved:\n- ${jsonPath}\n- ${txtPath}`);
 
                 resolve(jsonPath);
             } catch (err) {
-                console.error("Error saving Summary:", err);
+                // console.error("Error saving Summary:", err);
                 reject(err);
             }
         })

services/modules/jsonTools/transcriptionSummarizer2.js

@@ -32,29 +32,35 @@ module.exports = {
                     inputJson = JSON.parse(raw);
                 } catch (e) {
                     console.error("Failed to load JSON from file:", e);
-                    return { error: "Could not read JSON from file path." };
+                    reject("Could not read JSON from file path.")
+                    return
                 }
             }
             // JSON parsen
-                if (typeof args === "string") {
+            if (typeof args === "string") {
-                    try {
+                try {
-                        await new Promise((res) => {
+                    await new Promise((res, rej) => {
-                            fs.readFile(args, 'utf8', function (err, data) {
+                        fs.readFile(args, 'utf8', function (err, data) {
-                                if (err) throw err;
+                            if (err){
-                                inputJson = JSON.parse(data);
+                                rej(err)
-                                res()
+                                return
-                            });
+                            }
-                        })
+                            inputJson = JSON.parse(data);
-                    } catch (e) {
+                            res()
-                        console.log("Invalid JSON in summarize-transcription");
+                        });
-                        console.log(e)
+                    })
-                        return { error: "Invalid JSON" };
+                } catch (e) {
-                    }
+                    // console.log("Invalid JSON in summarize-transcription");
+                    // console.log(e)
+                    reject(e)
+                    return
                 }
+            }
 
             const words = inputJson.words;
             if (!Array.isArray(words)) {
-                return { error: "No words Array found" };
+                reject("No words Array found")
+                return;
             }
 
             const ENDINGS = [".", "!", "?"]; // '...' auch als Satzende ?
@@ -70,7 +76,7 @@ module.exports = {
                 if (!currentSpeaker) currentSpeaker = w.speaker;
                 if (startTime === null) startTime = w.start;
                 endTime = w.end;
-        
+
                 //speaker changing
                 if (currentSpeaker !== w.speaker && currentSentence) {
                     result.push({
@@ -132,10 +138,10 @@ module.exports = {
                 const txtPath = path.join(outputDir, `${filename}-${new Date().getTime()}.txt`);
                 fs.writeFileSync(txtPath, output.join("\n"), "utf-8");
 
-                console.log(`Summary successfully saved:\n- ${jsonPath}\n- ${txtPath}`);
+                // console.log(`Summary successfully saved:\n- ${jsonPath}\n- ${txtPath}`);
                 resolve(jsonPath);
             } catch (err) {
-                console.error("Error saving Summary:", err);
+                // console.error("Error saving Summary:", err);
                 reject(err);
             }
         })

services/modules/llm-chat_gpt/chatgpt.js

@@ -1,8 +1,131 @@
-module.exports = {
+const fs = require('fs');
-    name:"chatgpt", // Unique name for our function that will later be used to get the function from the map via "mapFunctions.get("example").function()"
+const path = require('path');
-    type:"llm", // value used to differentiate each module to order them in the UI
+
-    displayname:"ChatGPT", // The displayname used within the UI
+const outputDir = path.join(__dirname, "../../../storage/documents"); //  path for output directory
-    async function(parameter){
+
-        // TODO add code to actually send the transcript to ChatGPT and get a response back
+if (!fs.existsSync(outputDir)) {
+    fs.mkdirSync(outputDir, { recursive: true }); // Create output directory if it doesn't exist
+}
+
+// Ensure SAIA API key is set in environment variables: export SAIA_API_KEY="your_api_key_here"
+const SAIA_API_KEY = process.env.SAIA_API_KEY; // Ensure SAIA API key is set in environment variables
+
+const SAIA_URL = "https://chat-ai.academiccloud.de/v1/chat/completions"; //URL for the REST call, used model and action
+
+const module_exports = {
+    name: "llm-saia_openai_gpt",
+    type: "llm",
+    displayname: "GPT 120B",
+    description: "Generates documents using OpenAI GPT OSS 120B via SAIA platform",
+
+    async function(parameter) {
+        try {
+            console.log("SAIA OpenAI GPT module invoked with parameters:", parameter);
+
+            await this.createDocumentFromTranscript( //Call the function to create document with transcript, document type and language
+                parameter.inputTranscriptPath, // Path to input transcript file
+                parameter.documentTypePath, // Path to document type file which is chosen in the front end by the user
+                parameter.language // Language for the document which is chosen in the front end by the user
+            );
+
+        } catch (error) {
+            console.error("Error in SAIA OpenAI GPT module:", error);
+        }
+    },
+
+    createDocumentFromTranscript: async function(transcriptPath, documentTypePath, language = "en") { // default language is English
+        try {
+            const transcript = await fs.promises.readFile(transcriptPath, "utf-8"); //read transcript file from Path
+            const documentType = await fs.promises.readFile(documentTypePath, "utf-8"); //read document type from Path
+            const promptText = `${documentType}, in language ${language}, transcript:\n\n${transcript}`; //combine doc type, language and transcript - Change prompt here if needed
+
+            // --- REST CALL ---
+            const response = await fetch(SAIA_URL, {
+                method: "POST",
+                headers: {
+                    "Authorization": `Bearer ${SAIA_API_KEY}`,
+                    "Accept": "application/json",
+                    "Content-Type": "application/json"
+                },
+                body: JSON.stringify({
+                    model: "openai-gpt-oss-120b",
+                    messages: [
+                        { role: "system", content: "You are a helpful assistant that generates HTML documents from transcripts. Output only valid HTML content without any preamble, explanations, or markdown formatting." },
+                        { role: "user", content: promptText }
+                    ],
+                    temperature: 0
+                })
+            });
+
+            if (!response.ok) { //ok is true when a response was successful
+                const text = await response.text();
+                throw new Error(`SAIA API error (${response.status}): ${text}`);
+            }
+
+            const data = await response.json();
+
+            // Get generated text from response or default to empty string (if null)
+            // SAIA uses OpenAI-compatible structure: data.choices[x].message.content
+            const output = data.choices?.[0]?.message?.content || "";
+
+            let inputTranscriptName = path.basename(transcriptPath, path.extname(transcriptPath)); // Name for the output file 
+            console.log(inputTranscriptName);
+
+            const outPath = path.join(outputDir, `${inputTranscriptName}.html`); // Output file path & name to make naming dynamic. Pulled from input transcript name
+            fs.writeFileSync(outPath, output, "utf8"); // Write output to file
+
+            console.log("Generated document written to:", outPath);
+
+        } catch (error) {
+            console.error("Error generating SAIA content:", error);
+        }
     }
+};
+
+module.exports = module_exports;
+
+// CLI Mode: Allow direct execution
+if (require.main === module) {
+    (async () => {
+        const args = process.argv.slice(2);
+        
+        if (args.length < 2) {
+            console.error("Usage: node llm-openai-gpt.js <transcriptPath> <documentTypePath> [language]");
+            console.error("Example: node llm-openai-gpt.js ./transcript.json ./docType.txt de");
+            process.exit(1);
+        }
+
+        const [transcriptPath, documentTypePath, language] = args;
+
+        // Check if API key is set
+        if (!SAIA_API_KEY) {
+            console.error("ERROR: SAIA_API_KEY environment variable is not set!");
+            console.error("Please set it with: export SAIA_API_KEY='your_api_key_here'");
+            process.exit(1);
+        }
+
+        // Check if files exist
+        if (!fs.existsSync(transcriptPath)) {
+            console.error(`ERROR: Transcript file not found: ${transcriptPath}`);
+            process.exit(1);
+        }
+
+        if (!fs.existsSync(documentTypePath)) {
+            console.error(`ERROR: Document type file not found: ${documentTypePath}`);
+            process.exit(1);
+        }
+
+        console.log("Starting document generation...");
+        console.log(`Transcript: ${transcriptPath}`);
+        console.log(`Document Type: ${documentTypePath}`);
+        console.log(`Language: ${language || 'en (default)'}`);
+        
+        await module_exports.createDocumentFromTranscript(
+            transcriptPath,
+            documentTypePath,
+            language || 'en'
+        );
+        
+        console.log("Done!");
+    })();
 }

services/modules/llm-gemini/gemini.js

@@ -1,55 +1,134 @@
-const outputDir = path.join(__dirname, "../../../storage/documents"); //  path for output directory
+const fs = require('fs');
+const path = require('path');
+
+const outputDir = path.join(__dirname, "../../../storage/documents"); // path for output directory
 
 if (!fs.existsSync(outputDir)) {
     fs.mkdirSync(outputDir, { recursive: true }); // Create output directory if it doesn't exist
 }
 
-const ai = new genai.GoogleGenAI({ 
+// Ensure Google API key is set in environment variables: export GOOGLE_API_KEY="your_api_key_here"
-    apiKey: process.env.GOOGLE_API_KEY // Ensure Google API key is set in environment variables: export GOOGLE_API_KEY="your_api_key_here"
+const GEMINI_API_KEY = process.env.GOOGLE_API_KEY; // Ensure Google API key is set in environment variables: export GOOGLE_API_KEY="your_api_key_here"
-});
+const GEMINI_URL = "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:generateContent"; // URL for the REST call, used model and action
 
-module.exports = {
+const module_exports = {
     name: "llm-gemini",
     type: "llm",
     displayname: "Gemini LLM",
     description: "Generates documents using Google Gemini LLM",
 
     async function(parameter) {
-        try {
+        return new Promise(async (resolve, reject) => {
-            console.log("Gemini LLM module invoked with parameters:", parameter);
+            try {
+                // console.log("Gemini LLM module invoked with parameters:", parameter);
 
-            await this.createDocumentFromTranscript( //Call the function to create document with transcript, document type and language
+                resolve(await this.createDocumentFromTranscript( //Call the function to create document with transcript, document type and language
-                parameter.inputTranscriptPath, // Path to input transcript file
+                    parameter.inputTranscriptPath, // Path to input transcript file
-                parameter.documentTypePath, // Path to document type file which is chosen in the front end by the user
+                    parameter.documentTypePath, // Path to document type file which is chosen in the front end by the user
-                parameter.language // Language for the document which is chosen in the front end by the user
+                    parameter.language // Language for the document which is chosen in the front end by the user
-            );
+                ));
+
+            } catch (error) {
+                // console.error("Error in Gemini LLM module:", error);
+                reject(error)
+            }
+        })
 
-        } catch (error) {
-            console.error("Error in Gemini LLM module:", error);
-        }
     },
 
     createDocumentFromTranscript: async function(transcriptPath, documentTypePath, language = "en") { // default language is English
-        try {
+        return new Promise(async(resolve, reject) => {
-            const transcript = await fs.promises.readFile(transcriptPath, "utf-8");
+            try {
-            const documentType = await fs.promises.readFile(documentTypePath, "utf-8");
+                const transcript = await fs.promises.readFile(transcriptPath, "utf-8"); //read transcript file from Path
+                const documentType = await fs.promises.readFile(documentTypePath, "utf-8"); //read document type from Path
+                const promptText = `${documentType}, in language ${language}, transcript:\n\n${transcript}`; //combine doc type, language and transcript - Change prompt here if needed
 
-            const promptText = `${documentType}, in language ${language}, transcript:\n\n${transcript}`;
+                // --- REST CALL ---
+                const response = await fetch(`${GEMINI_URL}?key=${GEMINI_API_KEY}`, { //safe model response in variable
+                    method: "POST",
+                    headers: {
+                        "Content-Type": "application/json"
+                    },
+                    body: JSON.stringify({
+                        contents: [
+                            {
+                                parts: [
+                                    { text: promptText } // Input prompt for content generation
+                                ]
+                            }
+                        ]
+                    })
+                });
 
-            const response = await ai.models.generateContent({
+                if (!response.ok) { //ok is true when a responce was successfull
-                model: "gemini-2.5-flash", // Specify the Gemini model to use
+                    const text = await response.text();
-                contents: promptText // Input prompt for content generation
+                    throw new Error(`Gemini API error (${response.status}): ${text}`);
-            });
+                }
 
-            const output = response.text || ""; // Get generated text from response or default to empty string (if null)
+                const data = await response.json();
 
-            const outPath = path.join(outputDir, "test.md"); // Output file path & name TO BE DONE to make dynamic out of input transcript name
+                // Get generated text from response or default to empty string (if null)
-            fs.writeFileSync(outPath, output, "utf8"); // Write output to file
+                const output = data?.candidates?.[0]?.content?.parts?.[0]?.text || "";
+                let inputTranscriptName = path.basename(transcriptPath, path.extname(transcriptPath)); // Name for the output file 
+                // console.log(inputTranscriptName);
+                const outPath = path.join(outputDir, `${inputTranscriptName}.md`); // Output file path & name to make naming dynamic. Pulled from input transcript name
+                fs.writeFileSync(outPath, output, "utf8"); // Write output to file
 
-            console.log("Generated document written to:", outPath);
+                // console.log("Generated document written to:", outPath);
+                resolve(outPath)
 
-        } catch (error) {
+            } catch (error) {
-            console.error("Error generating Gemini content:", error);
+                // console.error("Error generating Gemini content:", error);
-        }
+                reject(error)
+            }
+        })
     }
-};
+};
+
+module.exports = module_exports;
+
+// CLI Mode: Allow direct execution
+if (require.main === module) {
+    (async () => {
+        const args = process.argv.slice(2);
+        
+        if (args.length < 2) {
+            console.error("Usage: node llm-gemini.js <transcriptPath> <documentTypePath> [language]");
+            console.error("Example: node llm-gemini.js ./transcript.json ./docType.txt de");
+            process.exit(1);
+        }
+
+        const [transcriptPath, documentTypePath, language] = args;
+
+        // Check if API key is set
+        if (!GEMINI_API_KEY) {
+            console.error("ERROR: GOOGLE_API_KEY environment variable is not set!");
+            console.error("Please set it with: export GOOGLE_API_KEY='your_api_key_here'");
+            process.exit(1);
+        }
+
+        // Check if files exist
+        if (!fs.existsSync(transcriptPath)) {
+            console.error(`ERROR: Transcript file not found: ${transcriptPath}`);
+            process.exit(1);
+        }
+
+        if (!fs.existsSync(documentTypePath)) {
+            console.error(`ERROR: Document type file not found: ${documentTypePath}`);
+            process.exit(1);
+        }
+
+        console.log("Starting document generation...");
+        console.log(`Transcript: ${transcriptPath}`);
+        console.log(`Document Type: ${documentTypePath}`);
+        console.log(`Language: ${language || 'en (default)'}`);
+        
+        await module_exports.createDocumentFromTranscript(
+            transcriptPath,
+            documentTypePath,
+            language || 'en'
+        );
+        
+        console.log("Done!");
+    })();
+}

services/modules/quen3/quen3.js

@@ -0,0 +1,130 @@
+const fs = require('fs');
+const path = require('path');
+
+const outputDir = path.join(__dirname, "../../../storage/documents"); // path for output directory
+
+if (!fs.existsSync(outputDir)) {
+    fs.mkdirSync(outputDir, { recursive: true }); // Create output directory if it doesn't exist
+}
+
+// Ensure SAIA API key is set in environment variables: export SAIA_API_KEY="your_api_key_here"
+const SAIA_API_KEY = process.env.SAIA_API_KEY;
+const SAIA_URL = "https://chat-ai.academiccloud.de/v1/chat/completions"; // URL for the REST call, used model and action
+
+const module_exports = {
+    name: "qwen3-235b-a22b",
+    type: "llm",
+    displayname: "QWEN 3 235B",
+    description: "Generates documents using QWEN 3 235B via SAIA platform",
+
+    async function(parameter) {
+        try {
+            console.log("SAIA QWEN 3 235B module invoked with parameters:", parameter);
+
+            await this.createDocumentFromTranscript( // Call the function to create document with transcript, document type and language
+                parameter.inputTranscriptPath, // Path to input transcript file
+                parameter.documentTypePath, // Path to document type file which is chosen in the front end by the user
+                parameter.language // Language for the document which is chosen in the front end by the user
+            );
+
+        } catch (error) {
+            console.error("Error in SAIA QWEN 3 235B module:", error);
+        }
+    },
+
+    createDocumentFromTranscript: async function(transcriptPath, documentTypePath, language = "en") { // default language is English
+        try {
+            const transcript = await fs.promises.readFile(transcriptPath, "utf-8"); // read transcript file from Path
+            const documentType = await fs.promises.readFile(documentTypePath, "utf-8"); // read document type from Path
+            const promptText = `${documentType}, in language ${language}, transcript:\n\n${transcript}`; // combine doc type, language and transcript - Change prompt here if needed
+
+            // --- REST CALL ---
+            const response = await fetch(SAIA_URL, {
+                method: "POST",
+                headers: {
+                    "Authorization": `Bearer ${SAIA_API_KEY}`,
+                    "Accept": "application/json",
+                    "Content-Type": "application/json"
+                },
+                body: JSON.stringify({
+                    model: "qwen3-235b-a22b",  
+                    messages: [
+                        { role: "system", content: "You are a helpful assistant that generates HTML documents from transcripts. Output only valid HTML content without any preamble, explanations, or markdown formatting." },
+                        { role: "user", content: promptText }
+                    ],
+                    temperature: 0
+                })
+            });
+
+            if (!response.ok) { // ok is true when a response was successful
+                const text = await response.text();
+                throw new Error(`SAIA API error (${response.status}): ${text}`);
+            }
+
+            const data = await response.json();
+
+            // Get generated text from response or default to empty string (if null)
+            // SAIA uses OpenAI-compatible structure: data.choices[x].message.content
+            const output = data.choices?.[0]?.message?.content || "";
+
+            let inputTranscriptName = path.basename(transcriptPath, path.extname(transcriptPath)); // Name for the output file
+            console.log(inputTranscriptName);
+
+            const outPath = path.join(outputDir, `${inputTranscriptName}.html`); // Output file path & name to make naming dynamic. Pulled from input transcript name
+            fs.writeFileSync(outPath, output, "utf8"); // Write output to file
+
+            console.log("Generated document written to:", outPath);
+
+        } catch (error) {
+            console.error("Error generating SAIA content:", error);
+        }
+    }
+};
+
+module.exports = module_exports;
+
+// CLI Mode: Allow direct execution
+if (require.main === module) {
+    (async () => {
+        const args = process.argv.slice(2);
+        
+        if (args.length < 2) {
+            console.error("Usage: node qwen3.js <transcriptPath> <documentTypePath> [language]");
+            console.error("Example: node qwen3.js ./transcript.json ./docType.txt de");
+            process.exit(1);
+        }
+
+        const [transcriptPath, documentTypePath, language] = args;
+
+        // Check if API key is set
+        if (!SAIA_API_KEY) {
+            console.error("ERROR: SAIA_API_KEY environment variable is not set!");
+            console.error("Please set it with: export SAIA_API_KEY='your_api_key_here'");
+            process.exit(1);
+        }
+
+        // Check if files exist
+        if (!fs.existsSync(transcriptPath)) {
+            console.error(`ERROR: Transcript file not found: ${transcriptPath}`);
+            process.exit(1);
+        }
+
+        if (!fs.existsSync(documentTypePath)) {
+            console.error(`ERROR: Document type file not found: ${documentTypePath}`);
+            process.exit(1);
+        }
+
+        console.log("Starting document generation...");
+        console.log(`Transcript: ${transcriptPath}`);
+        console.log(`Document Type: ${documentTypePath}`);
+        console.log(`Language: ${language || 'en (default)'}`);
+        
+        await module_exports.createDocumentFromTranscript(
+            transcriptPath,
+            documentTypePath,
+            language || 'en'
+        );
+        
+        console.log("Done!");
+    })();
+}

services/modules/transcription-remote/assembly.js

@@ -78,7 +78,7 @@ function saveTranscript(transcript, sessionId) {
   const outputPath = path.join(outputDir, `${sessionId}.json`);
   fs.writeFileSync(outputPath, JSON.stringify(transcript, null, 2));
 
-  console.log(`Transcript saved: ${outputPath}`);
+  // console.log(`Transcript saved: ${outputPath}`);
 
   return outputPath;
 }
@@ -89,6 +89,7 @@ module.exports = {
   name: 'assembly',
   type: 'transcription',
   displayname: 'AssemblyAI',
+  audioformat: "mp3",
 
   async function(audioFileName) {
     return new Promise(async (resolve, reject) => {
@@ -115,8 +116,9 @@ module.exports = {
         resolve(saveTranscript(transcript, sessionId));
 
       } catch (error) {
-        console.error('Transcription error:', error.message);
+        // console.error('Transcription error:', error.message);
         reject(error);
+        return
       }
     })
   }

services/modules/utility/@startup.js

@@ -1,3 +1,4 @@
+
 module.exports = {
     name:"Startup_function",
     async function(){

storage/documentType/standard_meeting_report.txt

@@ -0,0 +1,217 @@
+Generate a structured meeting report in HTML 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: Unclear:<reason>.
+
+==================== STRUCTURE & RULES ====================
+
+{
+  "FORMAT": "HTML",
+
+  "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": "HTML 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": "Unclear:<reason>"
+  },
+
+  "PROMPT_SNIPPET": "Generate meeting report in HTML using STRUCTURE and STYLE. Output only the report."
+}
+
+============================================================
+
+Insert all generated content into the following HTML TEMPLATE:
+
+# {{title}}
+
+**Date:** {{date}}  
+**Start:** {{start}}  
+**End:** {{end}}  
+**Duration:** {{duration}}  
+**Location:** {{location}}  
+**Host:** {{host}}  
+**Participants:** {{participants}}
+
+---
+
+## Table of Contents
+{{toc}}
+
+---
+Generate a structured meeting report in HTML 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": "HTML",
+
+  "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": "HTML 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 HTML using STRUCTURE and STYLE. Output only the report."
+}
+
+============================================================
+
+Insert all generated content into the following HTML 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 HTML meeting report.
+## Executive Summary
+{{exec_summary}}
+
+---
+
+## Sections
+{{sections}}
+
+---
+
+## Consolidated Decisions
+{{consolidated_decisions}}
+
+---
+
+## Consolidated Actions
+{{consolidated_actions}}
+
+---
+
+## Appendix
+{{appendix}}
+
+============================================================
+
+Final Requirement:
+Output ONLY the completed HTML meeting report.

test/test_health.py

@@ -1,8 +0,0 @@
-from fastapi.testclient import TestClient
-from app.main import app
-
-client = TestClient(app)
-
-def test_health():
-    response = client.get("/health")
-    assert response.status_code == 200

test/unit/test.js

@@ -0,0 +1,170 @@
+// DO NOT TOUCH THIS
+require("../../requires.js")
+
+mapFunctions = new Map()
+// Loading the Function Map
+var path = `${mainDir}/services/modules`
+var folders = fs.readdirSync(path).filter(function (file) {
+    return fs.statSync(path+'/'+file).isDirectory();
+});
+folders.forEach(element => {
+    var commandFiles = fs.readdirSync(`${path}/${element}`).filter(file => file.endsWith('.js') && !file.startsWith("index"));
+    for (const file of commandFiles) {
+        delete require.cache[require.resolve(`${path}/${element}/${file}`)];
+        const command = require(`${path}/${element}/${file}`);
+        mapFunctions.set(command.name, command);
+    }
+});
+// You can touch beyond this point
+
+let audiopath
+let transcriptPath
+let summarizePath
+let llmpath
+
+describe("Unit Tests", function() {
+
+    describe('Audio Extraction', function () {
+        this.slow(1000)
+        this.timeout(3000)
+        it('Extract .mp4 to .mp3', function (done) {
+
+            mapFunctions.get("extraction-video-to-audio").function({inputVideoPath: __dirname.replaceAll("\\","/")+"/testvideo.mp4", outputType: "mp3"}).then(resp => {
+                audiopath = resp
+                // console.log(resp);
+                done()
+            }).catch(err => {
+                throw err;
+            })
+        })
+        it('Extract .mp4 to .flac', function (done) {
+
+            mapFunctions.get("extraction-video-to-audio").function({inputVideoPath: __dirname.replaceAll("\\","/")+"/testvideo.mp4", outputType: "flac"}).then(resp => {
+                // console.log(resp);
+                done()
+            }).catch(err => {
+                throw err;
+            })
+        })
+        it('Extracting to a nonexistant format', function (done) {
+
+            mapFunctions.get("extraction-video-to-audio").function({inputVideoPath: __dirname.replaceAll("\\","/")+"/testvideo.mp4", outputType: "qqq"}).then(resp => {
+                // console.log(resp);
+                done("Didnt crash")
+            }).catch(err => {
+                done()
+            })
+        })
+        it('Extracting from nonexistant file', function (done) {
+
+            mapFunctions.get("extraction-video-to-audio").function({inputVideoPath: "a", outputType: "mp3"}).then(resp => {
+                // console.log(resp);
+                done("Didnt crash")
+            }).catch(err => {
+                done()
+            })
+        })
+        it('Extracting from nonexistant file to nonexistant format', function (done) {
+
+            mapFunctions.get("extraction-video-to-audio").function({inputVideoPath: "a", outputType: "qqq"}).then(resp => {
+                // console.log(resp);
+                done("Didnt crash")
+            }).catch(err => {
+                done()
+            })
+        })
+    });
+
+
+    describe("Audio Transcription", function() {
+        this.slow(20000)
+        this.timeout(120000)
+
+        it('Assembly', function (done) {
+            mapFunctions.get("assembly").function(audiopath).then(resp => {
+                // console.log(resp);
+                transcriptPath = resp
+                done()
+            }).catch(err => {
+                throw err
+            })
+        })
+
+        it('Assembly Wrong file', function (done) {
+            mapFunctions.get("assembly").function("a").then(resp => {
+                // console.log(resp);
+                // transcriptPath = resp
+                done("Didnt crash")
+            }).catch(err => {
+                // console.log(err);
+                done()
+            })
+        })
+        // TODO add more Transcription Tool tests here
+    })
+
+    describe("Transcript Summarizer", function() {
+        this.slow(100)
+        this.timeout(1000)
+
+        it("Summarizer 1", function (done){
+            mapFunctions.get("summarize-transcription").function(transcriptPath).then(resp => {
+                done()
+            }).catch(err => {
+                throw err
+            })
+        })
+
+        it("Summarizer 1 Wrong File", function (done){
+            mapFunctions.get("summarize-transcription").function("a").then(resp => {
+                done("Didnt crash")
+            }).catch(err => {
+                done()
+            })
+        })
+
+        it("Summarizer 2 (Main)", function (done){
+            mapFunctions.get("summarize-transcription2").function(transcriptPath).then(resp => {
+                summarizePath = resp
+                done()
+            }).catch(err => {
+                throw err
+            })
+        })
+
+        it("Summarizer 2 (Main) Wrong File", function (done){
+            mapFunctions.get("summarize-transcription2").function("a").then(resp => {
+                done("Didnt crash")
+            }).catch(err => {
+                done()
+            })
+        })
+    })
+
+    describe("Large Language Model", function() {
+        this.slow(30000)
+        this.timeout(120000)
+
+        // it("ChatGPT", function (done){
+        //     mapFunctions.get("chatgpt").function({inputTranscriptPath: summarizePath, documentTypePath: "./storage/documentType/meetingReport.json", language: "en"}).then(resp => {
+        //         done()
+        //     }).catch(err => {
+        //         throw err
+        //     })
+        // })
+
+        it("Gemini", function (done){
+            mapFunctions.get("llm-gemini").function({inputTranscriptPath: summarizePath, documentTypePath: "./storage/documentType/meetingReport.json", language: "en"}).then(resp => {
+                llmpath = resp
+                done()
+            }).catch(err => {
+                throw err
+            })
+        })
+    })
+    after(function() {
+        console.log(`\n\n\n${audiopath} \n${transcriptPath} \n${summarizePath} \n${llmpath}`);
+    })
+})
+
+