added json to send to frontend for the speaker naming functionality

REST API & Auto Naming

See merge request proj-wise2526-video2document/video2document!31

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++
@@ -194,6 +194,24 @@ 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-lock.json

@@ -299,7 +299,6 @@
       "resolved": "https://registry.npmjs.org/@types/node/-/node-24.9.2.tgz",
       "integrity": "sha512-uWN8YqxXxqFMX2RqGOrumsKeti4LlmIMIyV0lgut4jx7KQBcBiW6vkDtIBvHnHIquwNfJhk8v2OtmO8zXWHfPA==",
       "license": "MIT",
-      "peer": true,
       "dependencies": {
         "undici-types": "~7.16.0"
       }
@@ -2535,7 +2534,6 @@
       "integrity": "sha512-jl1vZzPDinLr9eUt3J/t7V6FgNEw9QjvBPdysz9KfQDD41fQrC2Y4vKQdiaUpFT4bXlb1RHhLpp8wtm6M5TgSw==",
       "dev": true,
       "license": "Apache-2.0",
-      "peer": true,
       "bin": {
         "tsc": "bin/tsc",
         "tsserver": "bin/tsserver"

services/modules/llm-gemini/gemini.js

@@ -4,9 +4,10 @@ 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 = {
     name: "llm-gemini",
@@ -31,19 +32,39 @@ module.exports = {
 
     createDocumentFromTranscript: async function(transcriptPath, documentTypePath, language = "en") { // default language is English
         try {
-            const transcript = await fs.promises.readFile(transcriptPath, "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");
+            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
-            const response = await ai.models.generateContent({
+                method: "POST",
-                model: "gemini-2.5-flash", // Specify the Gemini model to use
+                headers: {
-                contents: promptText // Input prompt for content generation
+                    "Content-Type": "application/json"
+                },
+                body: JSON.stringify({
+                    contents: [
+                        {
+                            parts: [
+                                { text: promptText } // Input prompt for content generation
+                            ]
+                        }
+                    ]
+                })
             });
 
-            const output = response.text || ""; // Get generated text from response or default to empty string (if null)
+            if (!response.ok) { //ok is true when a responce was successfull
+                const text = await response.text();
+                throw new Error(`Gemini API error (${response.status}): ${text}`);
+            }
 
-            const outPath = path.join(outputDir, "test.md"); // Output file path & name TO BE DONE to make dynamic out of input transcript name
+            const data = await response.json();
+
+            // Get generated text from response or default to empty string (if null)
+            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);

services/modules/transcription-remote/assembly.js

@@ -89,6 +89,7 @@ module.exports = {
   name: 'assembly',
   type: 'transcription',
   displayname: 'AssemblyAI',
+  audioformat: "mp3",
 
   async function(audioFileName) {
     return new Promise(async (resolve, reject) => {