revert captions to yt auto-gen

package-lock.json

@@ -9,7 +9,7 @@
       "version": "1.0.0",
       "license": "none",
       "dependencies": {
-        "@distube/ytdl-core": "^4.13.5",
+        "@distube/ytdl-core": "^4.14.4",
         "@fastify/cookie": "^9.3.1",
         "@fastify/cors": "^8.4.2",
         "@fastify/multipart": "^8.2.0",
@@ -232,19 +232,19 @@
       }
     },
     "node_modules/@distube/ytdl-core": {
-      "version": "4.13.5",
+      "version": "4.14.4",
-      "resolved": "https://registry.npmjs.org/@distube/ytdl-core/-/ytdl-core-4.13.5.tgz",
+      "resolved": "https://registry.npmjs.org/@distube/ytdl-core/-/ytdl-core-4.14.4.tgz",
-      "integrity": "sha512-g+4UJIR/auAJbia7iB0aWvaJDbs22P53NySWa47b1NT4xMTDJYguxHFArPrvRkcJrb/AgKjv/XoSZGghpL0CJA==",
+      "integrity": "sha512-dHb4GW3qATIjRsS6VIhm3Pop7FdUcDFhsnyQlsPeXW7UhTPuNS0BmraKiTpFbpp0Ky+rxBQjJBfPRFsM+dT1fg==",
       "dependencies": {
         "http-cookie-agent": "^6.0.5",
         "m3u8stream": "^0.8.6",
         "miniget": "^4.2.3",
         "sax": "^1.4.1",
         "tough-cookie": "^4.1.4",
-        "undici": "^6.19.2"
+        "undici": "five"
       },
       "engines": {
-        "node": ">=16"
+        "node": ">=14.0"
       },
       "funding": {
         "url": "https://github.com/distubejs/ytdl-core?sponsor"

package.json

@@ -30,7 +30,7 @@
     "xo": "^0.56.0"
   },
   "dependencies": {
-    "@distube/ytdl-core": "^4.13.5",
+    "@distube/ytdl-core": "^4.14.4",
     "@fastify/cookie": "^9.3.1",
     "@fastify/cors": "^8.4.2",
     "@fastify/multipart": "^8.2.0",

src/index.js

@@ -54,7 +54,7 @@ export const main = async () => {
 
   server.register(oauth, {
     name: 'googleOAuth2',
-    scope: ['https://www.googleapis.com/auth/youtube.readonly', "https://www.googleapis.com/auth/userinfo.email", "https://www.googleapis.com/auth/userinfo.profile"],
+    scope: ['https://www.googleapis.com/auth/youtube.readonly', 'https://www.googleapis.com/auth/youtube.force-ssl', "https://www.googleapis.com/auth/userinfo.email", "https://www.googleapis.com/auth/userinfo.profile"],
     credentials: {
       client: {
         id: env.GOOGLE_CLIENT_ID,

src/routes/dashboard.js

@@ -4,7 +4,8 @@ import { and, desc, eq, getTableColumns, inArray, notInArray, sql } from "drizzl
 import { db } from "../db/index.js";
 import { articles, articles as articlesTable, signups as signupsTable, sites, users } from "../db/schemas.js";
 import { authMiddleware, authMiddlewareFn } from "../modules/middleware.js";
-import { jsonToCsv, createBlogFromCaptions, createArticleSlug, getVideoById, env, getWhisperCaptions, getVideoDetails } from "../utils/index.js";
+import { jsonToCsv, createBlogFromCaptions, createArticleSlug, getVideoById, env, getWhisperCaptions, getVideoDetails, getVideoWithCaptions } from "../utils/index.js";
+import { escape } from "querystring";
 
 const websubVerifyToken = "FQNI4Suzih";
 
@@ -231,14 +232,13 @@ export const dashboardRoutes = (fastify, _, done) => {
 
             const site = await db.select().from(sites).where(eq(sites.user_id, req.session.user_id));
 
-            const video_data = await getVideoDetails(req.body.video_id);
+            // const video_data = await getVideoDetails(req.body.video_id);
+            const video_data = await getVideoWithCaptions(req.body.video_id);
 
             reply.send({
                 success: true
             });
 
-            await delay(5000);
-
             article = (await db.insert(articlesTable).values({
                 title: video_data.title,
                 source_video_id: match[2],
@@ -247,7 +247,7 @@ export const dashboardRoutes = (fastify, _, done) => {
                 is_public: false
             }).returning({ id: articlesTable.id }))[0];
 
-            video_data.captions = await getWhisperCaptions(req.body.video_id);
+            // video_data.captions = await getWhisperCaptions(req.body.video_id);
 
             // const video_data = await getVideoById(access_token, req.body.video_id);
             await db.update(articlesTable).set({
@@ -275,13 +275,22 @@ export const dashboardRoutes = (fastify, _, done) => {
         } catch (e) {
             console.log(e);
 
+            if(e.message == "no_captions") {
+                reply.status(400).send({
+                    success: false,
+                    message: "This video does not have captions!"
+                });
+            }
+
             article ? await db.update(articlesTable).set({
                 status: "error"
             }).where(eq(articlesTable.id, article.id)) : "";
 
+            let message = "problem_creating_article";
+
             reply.status(500).send({
                 success: false,
-                message: "problem_creating_article"
+                message: message
             })
         }
     });
@@ -455,6 +464,3 @@ export const dashboardRoutes = (fastify, _, done) => {
     done();
 };
 
-function delay(time) {
-    return new Promise(resolve => setTimeout(resolve, time));
-}   

src/utils/ai.js

@@ -118,7 +118,7 @@ export async function createBlogFromCaptions(captions, {
 ## OutputFormat:
 1. Do not include a title at the top of the article.
 2. VERY IMPORTANT! Use markdown to add formatting to the article and make it easier to read.
-3. Length: The article should be roughly ${length || 700} words.
+3. Length: The article should be roughly ${length || 1200} words.
 ${faq ? "4. FAQ Section: Add a FAQ section at the end of the article to address common questions." : ""}
 ## Workflow:
 1. First, analyze the provided transcript to understand the key points and factual content.
@@ -134,8 +134,6 @@ As an SEO Content Specialist, you must follow the specified rules and communicat
         content: `# Information: 
 ## Title
 ${title}
-## Description
-${description}
 ## Transcript
 ${captions}`
     }];

src/utils/youtube.js

@@ -5,6 +5,7 @@ import { google } from 'googleapis';
 import ytdl from "@distube/ytdl-core";
 import { getWhisperCaptions } from './ai.js';
 import { articles as articlesTable } from "../db/schemas.js";
+import * as xml2js from 'xml2js';
 
 const service = google.youtube("v3");
 
@@ -127,14 +128,52 @@ export async function getVideoDetails(video_url) {
     }
 }
 
+export function parseTextFromCaptions(caption_text) {
+    let text_content = "";
+    const captionEntries = caption_text.split(/\n\n/);
+    for (const entry of captionEntries) {
+        const lines = entry.trim().split('\n');
+        if (lines.length >= 2 && !lines[1].includes('-->')) {
+            text_content += lines.slice(1).join(' ').trim() + ' ';
+        }
+    }
+
+    return text_content
+}
+
+function parseXMLCaptions(xmlString) {
+    return new Promise((resolve, reject) => {
+        const parser = new xml2js.Parser();
+
+        parser.parseString(xmlString, (err, result) => {
+            if (err) {
+                reject(err);
+            } else {
+                const textElements = result.transcript.text;
+                let captionText = textElements.map(element => element._).join('\n');
+                resolve(captionText.trim());
+            }
+        });
+    });
+}
+
+function delay(time) {
+    return new Promise(resolve => setTimeout(resolve, time));
+}   
+
 export async function getVideoWithCaptions(video_url) {
-    const info = getVideoDetails(video_url);
+    const info = await ytdl.getInfo(video_url);
-
+    const tracks = info
-    const captions = await getWhisperCaptions(video_url);
+        .player_response.captions
+        .playerCaptionsTracklistRenderer.captionTracks;
 
+    if (!tracks || tracks.length < 1) throw new Error("no_captions");
+    const track = tracks.find(t => t.languageCode == "en");
+    if(!track) throw new Error("no_captions");
+    const captions = await (await fetch(`${track.baseUrl}&fmt=srv1`)).text();
     return {
         title: info.videoDetails.title,
         description: info.videoDetails.description,
-        captions
+        captions: await parseXMLCaptions(captions)
     }
 }