rss-feedmonitor/scripts/human-behavior.js

/**
 * Human-like behavior utilities for Playwright to avoid bot detection
 * Includes realistic mouse movements, scrolling, and timing variations
 */

/**
 * Generate a random number between min and max (inclusive)
 */
function randomInt(min, max) {
  return Math.floor(Math.random() * (max - min + 1)) + min;
}

/**
 * Generate a random float between min and max
 */
function randomFloat(min, max) {
  return Math.random() * (max - min) + min;
}

/**
 * Sleep for a random duration within a range
 */
export async function randomDelay(minMs = 100, maxMs = 500) {
  const delay = randomInt(minMs, maxMs);
  await new Promise(resolve => setTimeout(resolve, delay));
}

/**
 * Generate bezier curve points for smooth mouse movement
 * Uses cubic bezier with random control points for natural curves
 */
function generateBezierPath(start, end, steps = 25) {
  const points = [];
  
  // Add some randomness to control points
  const cp1x = start.x + (end.x - start.x) * randomFloat(0.25, 0.4);
  const cp1y = start.y + (end.y - start.y) * randomFloat(-0.2, 0.2);
  const cp2x = start.x + (end.x - start.x) * randomFloat(0.6, 0.75);
  const cp2y = start.y + (end.y - start.y) * randomFloat(-0.2, 0.2);
  
  for (let i = 0; i <= steps; i++) {
    const t = i / steps;
    const t2 = t * t;
    const t3 = t2 * t;
    const mt = 1 - t;
    const mt2 = mt * mt;
    const mt3 = mt2 * mt;
    
    const x = mt3 * start.x + 
              3 * mt2 * t * cp1x + 
              3 * mt * t2 * cp2x + 
              t3 * end.x;
    const y = mt3 * start.y + 
              3 * mt2 * t * cp1y + 
              3 * mt * t2 * cp2y + 
              t3 * end.y;
    
    points.push({ x: Math.round(x), y: Math.round(y) });
  }
  
  return points;
}

/**
 * Move mouse in a realistic, smooth path with occasional overshooting
 * @param {Page} page - Playwright page object
 * @param {Object} target - Target coordinates {x, y}
 * @param {Object} options - Movement options
 */
export async function humanMouseMove(page, target, options = {}) {
  const {
    overshootChance = 0.15, // 15% chance to overshoot
    overshootDistance = 20,  // pixels to overshoot by
    steps = 25,              // number of steps in the path
    stepDelay = 10           // ms between steps
  } = options;
  
  // Get current mouse position (or start from a random position)
  const viewport = page.viewportSize();
  const start = {
    x: randomInt(viewport.width * 0.3, viewport.width * 0.7),
    y: randomInt(viewport.height * 0.3, viewport.height * 0.7)
  };
  
  // Decide if we should overshoot
  const shouldOvershoot = Math.random() < overshootChance;
  
  let finalTarget = target;
  if (shouldOvershoot) {
    // Calculate overshoot position (slightly past the target)
    const angle = Math.atan2(target.y - start.y, target.x - start.x);
    const overshoot = {
      x: target.x + Math.cos(angle) * randomInt(5, overshootDistance),
      y: target.y + Math.sin(angle) * randomInt(5, overshootDistance)
    };
    
    // Move to overshoot position first
    const overshootPath = generateBezierPath(start, overshoot, steps);
    for (const point of overshootPath) {
      await page.mouse.move(point.x, point.y);
      await new Promise(resolve => setTimeout(resolve, stepDelay));
    }
    
    // Then correct back to target
    const correctionPath = generateBezierPath(overshoot, target, Math.floor(steps * 0.3));
    for (const point of correctionPath) {
      await page.mouse.move(point.x, point.y);
      await new Promise(resolve => setTimeout(resolve, stepDelay));
    }
  } else {
    // Normal smooth movement
    const path = generateBezierPath(start, target, steps);
    for (const point of path) {
      await page.mouse.move(point.x, point.y);
      await new Promise(resolve => setTimeout(resolve, stepDelay));
    }
  }
  
  // Add a tiny random pause after reaching target
  await randomDelay(50, 150);
}

/**
 * Perform random mouse movements to simulate human reading/scanning
 */
export async function randomMouseMovements(page, count = 3) {
  const viewport = page.viewportSize();
  
  for (let i = 0; i < count; i++) {
    const target = {
      x: randomInt(100, viewport.width - 100),
      y: randomInt(100, viewport.height - 100)
    };
    
    await humanMouseMove(page, target, {
      overshootChance: 0.1,
      steps: randomInt(15, 30)
    });
    
    await randomDelay(200, 800);
  }
}

/**
 * Scroll page in a human-like manner with random intervals and amounts
 * @param {Page} page - Playwright page object
 * @param {Object} options - Scrolling options
 */
export async function humanScroll(page, options = {}) {
  const {
    direction = 'down',      // 'down' or 'up'
    scrollCount = 3,         // number of scroll actions
    minScroll = 100,         // minimum pixels per scroll
    maxScroll = 400,         // maximum pixels per scroll
    minDelay = 500,          // minimum delay between scrolls
    maxDelay = 2000,         // maximum delay between scrolls
    randomDirection = false  // occasionally scroll in opposite direction
  } = options;
  
  for (let i = 0; i < scrollCount; i++) {
    // Determine scroll direction
    let scrollDir = direction;
    if (randomDirection && Math.random() < 0.15) {
      scrollDir = direction === 'down' ? 'up' : 'down';
    }
    
    // Random scroll amount
    const scrollAmount = randomInt(minScroll, maxScroll);
    const scrollValue = scrollDir === 'down' ? scrollAmount : -scrollAmount;
    
    // Perform scroll in small increments for smoothness
    const increments = randomInt(5, 12);
    const incrementValue = scrollValue / increments;
    
    for (let j = 0; j < increments; j++) {
      await page.evaluate((delta) => {
        window.scrollBy(0, delta);
      }, incrementValue);
      await new Promise(resolve => setTimeout(resolve, randomInt(20, 50)));
    }
    
    // Random pause between scrolls (simulating reading)
    await randomDelay(minDelay, maxDelay);
  }
}

/**
 * Scroll to a specific element in a human-like way
 */
export async function scrollToElement(page, selector, options = {}) {
  const element = await page.locator(selector).first();
  
  // Get element position
  const box = await element.boundingBox();
  if (!box) {
    console.warn(`Element ${selector} not found or not visible`);
    return;
  }
  
  // Get current scroll position
  const currentScroll = await page.evaluate(() => window.scrollY);
  const viewportHeight = page.viewportSize().height;
  
  // Calculate target scroll position (element near middle of viewport)
  const targetScroll = box.y + currentScroll - (viewportHeight / 2);
  const scrollDistance = targetScroll - currentScroll;
  
  // Scroll in chunks
  const chunks = Math.max(3, Math.abs(Math.floor(scrollDistance / 200)));
  const chunkSize = scrollDistance / chunks;
  
  for (let i = 0; i < chunks; i++) {
    await page.evaluate((delta) => {
      window.scrollBy(0, delta);
    }, chunkSize);
    await randomDelay(50, 150);
  }
  
  await randomDelay(300, 700);
}

/**
 * Click an element with human-like behavior
 */
export async function humanClick(page, selector, options = {}) {
  const {
    moveToElement = true,
    doubleClickChance = 0.02  // 2% chance of accidental double-click
  } = options;
  
  const element = await page.locator(selector).first();
  const box = await element.boundingBox();
  
  if (!box) {
    throw new Error(`Element ${selector} not found or not visible`);
  }
  
  // Calculate click position (slightly random within element bounds)
  const target = {
    x: box.x + randomInt(box.width * 0.3, box.width * 0.7),
    y: box.y + randomInt(box.height * 0.3, box.height * 0.7)
  };
  
  if (moveToElement) {
    await humanMouseMove(page, target);
  }
  
  // Random pre-click pause
  await randomDelay(100, 300);
  
  // Click
  await page.mouse.click(target.x, target.y);
  
  // Occasional accidental double-click
  if (Math.random() < doubleClickChance) {
    await randomDelay(50, 150);
    await page.mouse.click(target.x, target.y);
  }
  
  await randomDelay(200, 500);
}

/**
 * Type text with human-like timing variations
 */
export async function humanType(page, selector, text, options = {}) {
  const {
    minDelay = 50,
    maxDelay = 150,
    mistakes = 0.02  // 2% chance of typo
  } = options;
  
  await page.click(selector);
  await randomDelay(200, 400);
  
  const chars = text.split('');
  let typedText = '';
  
  for (let i = 0; i < chars.length; i++) {
    const char = chars[i];
    
    // Occasional typo
    if (Math.random() < mistakes && i < chars.length - 1) {
      // Type wrong char
      const wrongChar = String.fromCharCode(char.charCodeAt(0) + randomInt(-2, 2));
      await page.keyboard.type(wrongChar);
      await randomDelay(minDelay, maxDelay);
      
      // Pause (realize mistake)
      await randomDelay(200, 500);
      
      // Backspace
      await page.keyboard.press('Backspace');
      await randomDelay(100, 200);
    }
    
    // Type correct char
    await page.keyboard.type(char);
    
    // Variable delay based on character type
    let delay;
    if (char === ' ') {
      delay = randomInt(maxDelay * 1.5, maxDelay * 2);
    } else if (char.match(/[.!?,]/)) {
      delay = randomInt(maxDelay * 1.2, maxDelay * 2);
    } else {
      delay = randomInt(minDelay, maxDelay);
    }
    
    await new Promise(resolve => setTimeout(resolve, delay));
  }
  
  await randomDelay(300, 600);
}

/**
 * Wait for page load with random human-like observation time
 */
export async function humanWaitForLoad(page, options = {}) {
  const {
    minWait = 1000,
    maxWait = 3000
  } = options;
  
  // Wait for network to be idle
  await page.waitForLoadState('networkidle', { timeout: 30000 });
  
  // Additional random observation time (simulating reading/scanning)
  await randomDelay(minWait, maxWait);
}

/**
 * Simulate reading behavior - random scrolls and mouse movements
 */
export async function simulateReading(page, duration = 5000) {
  const endTime = Date.now() + duration;
  
  while (Date.now() < endTime) {
    const action = Math.random();
    
    if (action < 0.4) {
      // Scroll a bit
      await humanScroll(page, {
        scrollCount: 1,
        minScroll: 50,
        maxScroll: 200,
        minDelay: 800,
        maxDelay: 1500
      });
    } else if (action < 0.7) {
      // Move mouse randomly
      await randomMouseMovements(page, 1);
    } else {
      // Just wait (reading)
      await randomDelay(1000, 2000);
    }
  }
}

/**
 * Configure browser context with realistic human-like settings
 */
export async function getHumanizedContext(browser, options = {}) {
  const {
    locale = 'en-CA',
    timezone = 'America/Toronto',
    viewport = null
  } = options;
  
  // Random but realistic viewport sizes
  const viewports = [
    { width: 1920, height: 1080 },
    { width: 1366, height: 768 },
    { width: 1536, height: 864 },
    { width: 1440, height: 900 },
    { width: 2560, height: 1440 }
  ];
  
  const selectedViewport = viewport || viewports[randomInt(0, viewports.length - 1)];
  
  // Realistic user agents (updated to current versions)
  const userAgents = [
    'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36',
    'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36',
    'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/130.0.0.0 Safari/537.36',
    'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/130.0.0.0 Safari/537.36',
    'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36'
  ];
  
  const context = await browser.newContext({
    viewport: selectedViewport,
    userAgent: userAgents[randomInt(0, userAgents.length - 1)],
    locale,
    timezoneId: timezone,
    permissions: [],
    geolocation: { latitude: 43.6532, longitude: -79.3832 }, // Toronto
    colorScheme: 'light',  // Always light for consistency
    deviceScaleFactor: 1,  // Standard scaling
    hasTouch: false,
    isMobile: false,
    javaScriptEnabled: true,
    // Add realistic headers
    extraHTTPHeaders: {
      'Accept-Language': 'en-CA,en-US;q=0.9,en;q=0.8',
      'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8',
      'Sec-Fetch-Site': 'none',
      'Sec-Fetch-Mode': 'navigate',
      'Sec-Fetch-User': '?1',
      'Sec-Fetch-Dest': 'document',
      'Sec-Ch-Ua': '"Google Chrome";v="131", "Chromium";v="131", "Not_A Brand";v="24"',
      'Sec-Ch-Ua-Mobile': '?0',
      'Sec-Ch-Ua-Platform': '"macOS"',
      'Upgrade-Insecure-Requests': '1'
    }
  });
  
  // Inject additional fingerprint randomization and anti-detection
  await context.addInitScript(() => {
    // Remove webdriver property
    Object.defineProperty(navigator, 'webdriver', {
      get: () => undefined
    });
    
    // Override permissions
    const originalQuery = window.navigator.permissions.query;
    window.navigator.permissions.query = (parameters) => (
      parameters.name === 'notifications' ?
        Promise.resolve({ state: Notification.permission }) :
        originalQuery(parameters)
    );
    
    // Add chrome property
    window.chrome = {
      runtime: {}
    };
    
    // Override plugins
    Object.defineProperty(navigator, 'plugins', {
      get: () => [
        {
          0: { type: 'application/x-google-chrome-pdf', suffixes: 'pdf', description: 'Portable Document Format' },
          description: 'Portable Document Format',
          filename: 'internal-pdf-viewer',
          length: 1,
          name: 'Chrome PDF Plugin'
        },
        {
          0: { type: 'application/pdf', suffixes: 'pdf', description: '' },
          description: '',
          filename: 'mhjfbmdgcfjbbpaeojofohoefgiehjai',
          length: 1,
          name: 'Chrome PDF Viewer'
        }
      ]
    });
  });
  
  return context;
}

export default {
  randomDelay,
  humanMouseMove,
  randomMouseMovements,
  humanScroll,
  scrollToElement,
  humanClick,
  humanType,
  humanWaitForLoad,
  simulateReading,
  getHumanizedContext
};