Youtube_transcript_playlist

index.mjs

import { YoutubeTranscript } from 'youtube-transcript';
import puppeteer from 'puppeteer';
import fs from 'fs/promises';
import readline from 'readline/promises';
import * as path from 'path';

const MAX_CHARS_PER_FILE = 500000;

async function main() {
    const rl = readline.createInterface({
        input: process.stdin,
        output: process.stdout
    });

    try {
        const playlistUrl = await rl.question('Enter the URL of the YouTube playlist: ');
        const browser = await puppeteer.launch({ args: ['--no-sandbox'] });
        const page = await browser.newPage();

        await page.goto(playlistUrl, { waitUntil: 'networkidle2' });

        try {
            await page.waitForSelector('ytd-playlist-video-renderer, #title h1 yt-formatted-string', { timeout: 60000 });
        } catch (error) {
            console.error('Error waiting for playlist items or title. Invalid playlist?', error);
            return;
        }

        let playlistTitle = await page.evaluate(() => {
            const titleElement = document.querySelector('#title h1 yt-formatted-string');
            return titleElement ? titleElement.textContent.replaceAll(/\s+/g, "_").trim() : 'Untitled_Playlist';
        });

        // Sanitize playlistTitle for filenames
        let safeTitle = playlistTitle.replace(/[/\\?%*:|"<>]/g, '-');
        if (!safeTitle || !safeTitle.trim()) {
            safeTitle = "Untitled_Playlist";
        }

        const videos = await page.evaluate(() => {
            const videoData = [];
            const playlistItems = document.querySelectorAll('ytd-playlist-video-renderer');
            playlistItems.forEach(item => {
                const link = item.querySelector('a#video-title');
                if (link) {
                    const url = link.href;
                    const title = link.textContent.trim();
                    videoData.push({ url, title });
                }
            });
            return videoData;
        });

        await browser.close();

        const finalDoc = [];
        let currentFileIndex = 1;
        let currentFileContent = [];
        let currentCharCount = 0;

        for (const element of videos) {
            try {
                const transcript = await YoutubeTranscript.fetchTranscript(element.url, {
                    lang: 'en',
                    enableAutoGeneratedCaptions: true
                });
                const allText = transcript.map(item => item.text).join(' ');
                const videoObj = { url: element.url, title: element.title, transcription: allText };
                const videoJson = JSON.stringify(videoObj);

                if (currentCharCount + videoJson.length > MAX_CHARS_PER_FILE) {
                    finalDoc.push({
                        filename: `${safeTitle}_part${currentFileIndex}.txt`,
                        data: currentFileContent
                    });
                    currentFileIndex++;
                    currentFileContent = [];
                    currentCharCount = 0;
                }

                currentFileContent.push(videoObj);
                currentCharCount += videoJson.length;
                console.log(`Transcription fetched for ${element.title}`);

            } catch (error) {
                console.error(`Error fetching transcript for ${element.url}:`, error);
                currentFileContent.push({ 
                    url: element.url, 
                    title: element.title, 
                    transcription: 'Error fetching transcript' 
                });
            }
        }

        // Save the last file
        if (currentFileContent.length > 0) {
            finalDoc.push({
                filename: `${safeTitle}_part${currentFileIndex}.txt`,
                data: currentFileContent
            });
        }

        const outputDir = await rl.question('Enter output directory (leave blank for current directory): ');
        const safeOutputDir = outputDir.trim() || '.';

        for (const fileData of finalDoc) {
            if (!fileData || !fileData.filename) {
                console.error('Invalid file data:', fileData);
                continue;
            }

            const filePath = path.join(safeOutputDir, fileData.filename);
            const data = JSON.stringify(fileData.data, null, 2);

            try {
                await fs.mkdir(path.dirname(filePath), { recursive: true });
                await fs.writeFile(filePath, data);
                console.log(`File ${filePath} created successfully`);
            } catch (err) {
                console.error(`Error writing file ${filePath}:`, err);
            }
        }

    } catch (error) {
        console.error('An unexpected error occurred:', error);
    } finally {
        rl.close();
    }
}

main();