ytdlp-segment-downloader/src/cli/downloader.ts

import { spawn } from "child_process";
import { writeFileSync, mkdirSync, existsSync } from "fs";
import { join } from "path";

export interface DownloadOptions {
  url: string;
  outputDir: string;
  format: string;
}

interface HeatmapSegment {
  start_seconds: number;
  end_seconds: number;
  intensity: number;
}

interface VideoInfo {
  title: string;
  duration: number;
  heatmap?: HeatmapSegment[];
}

async function getVideoInfo(url: string): Promise<VideoInfo> {
  return new Promise((resolve, reject) => {
    const ytDlp = spawn("yt-dlp", [
      "--dump-json",
      "--no-download",
      "--compat-option",
      "no-youtube-channel-redirect",
      url,
    ]);

    let stdout = "";
    let stderr = "";

    ytDlp.stdout.on("data", (data) => {
      stdout += data.toString();
    });

    ytDlp.stderr.on("data", (data) => {
      stderr += data.toString();
    });

    ytDlp.on("close", (code) => {
      if (code !== 0) {
        reject(new Error(`yt-dlp failed: ${stderr}`));
        return;
      }

      try {
        const info = JSON.parse(stdout);

        // Extract heatmap data from YouTube's internal API
        // The heatmap shows what segments were re-watched the most
        const heatmapData = info.heatmap;

        resolve({
          title: info.title || "video",
          duration: info.duration || 0,
          heatmap: heatmapData,
        });
      } catch (parseError) {
        reject(new Error(`Failed to parse video info: ${parseError}`));
      }
    });

    ytDlp.on("error", (err) => {
      reject(new Error(`Failed to run yt-dlp: ${err.message}`));
    });
  });
}

async function downloadSegment(
  url: string,
  outputPath: string,
  startTime: number,
  endTime: number,
  format: string
): Promise<void> {
  const section = `*${startTime.toFixed(3)}-${endTime.toFixed(3)}`;

  return new Promise((resolve, reject) => {
    const ytDlp = spawn("yt-dlp", [
      "-f", format,
      "--download-sections", section,
      "-o", outputPath,
      url,
    ]);

    let stderr = "";

    ytDlp.stderr.on("data", (data) => {
      stderr += data.toString();
    });

    ytDlp.on("close", (code) => {
      if (code !== 0) {
        reject(new Error(`yt-dlp failed: ${stderr}`));
        return;
      }
      resolve();
    });

    ytDlp.on("error", (err) => {
      reject(new Error(`Failed to run yt-dlp: ${err.message}`));
    });
  });
}

function sanitizeFilename(filename: string): string {
  return filename
    .replace(/[^a-zA-Z0-9\s\-_]/g, "")
    .replace(/\s+/g, "_")
    .substring(0, 100);
}

function formatTime(seconds: number): string {
  const mins = Math.floor(seconds / 60);
  const secs = Math.floor(seconds % 60);
  return `${mins}:${secs.toString().padStart(2, "0")}`;
}

export async function downloadMostWatchedSegment(options: DownloadOptions): Promise<void> {
  const { url, outputDir, format } = options;

  // Create output directory if it doesn't exist
  if (!existsSync(outputDir)) {
    mkdirSync(outputDir, { recursive: true });
  }

  // Get video info with heatmap data from YouTube
  console.log("Fetching video information from YouTube...");
  const info = await getVideoInfo(url);
  const safeTitle = sanitizeFilename(info.title);

  console.log(`Video: ${info.title}`);
  console.log(`Duration: ${formatTime(info.duration)}`);

  // Check for heatmap data - this shows what was re-watched the most
  if (!info.heatmap || info.heatmap.length === 0) {
    console.log("\nNo heatmap data available for this video.");
    console.log("The video may not have enough view data to determine most watched segments.");
    console.log("Downloading full video instead...");
    
    const outputPath = join(outputDir, `${safeTitle}.%(ext)s`);
    await downloadSegment(url, outputPath, 0, info.duration, format);
    return;
  }

  // Find the most watched segment (highest intensity)
  const mostWatched = info.heatmap.reduce((max, current) => {
    return current.intensity > max.intensity ? current : max;
  });

  console.log(`\nHeatmap data found: ${info.heatmap.length} segments`);
  console.log(`Most watched segment intensity: ${(mostWatched.intensity * 100).toFixed(1)}%`);
  console.log(`Segment: ${formatTime(mostWatched.start_seconds)} - ${formatTime(mostWatched.end_seconds)}`);

  // Download the most watched segment
  const outputPath = join(outputDir, `${safeTitle}_most_watched.%(ext)s`);

  console.log(`\nDownloading most watched segment...`);
  await downloadSegment(
    url,
    outputPath,
    mostWatched.start_seconds,
    mostWatched.end_seconds,
    format
  );

  // Save segment info
  const segmentInfoPath = join(outputDir, `${safeTitle}_segment_info.txt`);
  const segmentInfo = `# ${info.title}\n\n` +
    `Most watched segment (from YouTube heatmap):\n` +
    `  Start: ${formatTime(mostWatched.start_seconds)} (${mostWatched.start_seconds}s)\n` +
    `  End: ${formatTime(mostWatched.end_seconds)} (${mostWatched.end_seconds}s)\n` +
    `  Duration: ${formatTime(mostWatched.end_seconds - mostWatched.start_seconds)}\n` +
    `  Intensity: ${(mostWatched.intensity * 100).toFixed(1)}%\n\n` +
    `Note: This segment had the highest re-watch rate according to YouTube's analytics.\n`;

  writeFileSync(segmentInfoPath, segmentInfo);
  console.log(`\nSegment info saved to: ${segmentInfoPath}`);
  console.log("Download complete!");
}