Simulate realistic mouse movement; remove click events to avoid accid… #1548
Conversation
WalkthroughReplaces straight-line mouse moves in the async crawler strategy with randomized quadratic Bezier trajectories: viewport-aware start/end points, a randomized control point, stepped movement along the curve, and small random delays between steps. Overall user-simulation flow is preserved. Changes
Sequence Diagram(s)sequenceDiagram
participant Crawler
participant Strategy as AsyncCrawlerStrategy
participant Browser
participant Mouse
rect #F0F9FF
Crawler->>Strategy: trigger user-simulation
Strategy->>Browser: get viewport size (width,height)
note right of Strategy #F6FFF0: choose random start/end\nwithin viewport and control offsets
Strategy->>Strategy: compute quadratic Bezier control point
Strategy->>Mouse: for each step -> move to next (x,y)
Mouse-->>Browser: dispatch move events (with tiny random delays)
end
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Poem
Pre-merge checks and finishing touches❌ Failed checks (1 warning)
✅ Passed checks (2 passed)
✨ Finishing touches
🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
crawl4ai/async_crawler_strategy.py (1)
928-930: Consider slower delays for more human-like movement.The current delay range of 0.001-0.003 seconds (1-3ms) between steps results in very fast mouse movements (15-75ms total for 15-25 steps). Typical human cursor movements take 300-1000ms. While fast movements might be acceptable for anti-detection purposes, consider whether a slightly longer delay range (e.g., 0.005-0.015 seconds) would better simulate natural human behavior without compromising the goal of avoiding detection.
If you'd like to adjust the timing, consider:
await page.mouse.move(x, y) # Random small delay between movements to simulate human behavior - await asyncio.sleep(random.uniform(0.001, 0.003)) + await asyncio.sleep(random.uniform(0.005, 0.015))
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
crawl4ai/async_crawler_strategy.py(2 hunks)
🧰 Additional context used
🪛 Ruff (0.14.0)
crawl4ai/async_crawler_strategy.py
908-908: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
909-909: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
910-910: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
911-911: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
915-915: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
916-916: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
919-919: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
930-930: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
🔇 Additional comments (1)
crawl4ai/async_crawler_strategy.py (1)
6-6: LGTM! Import is appropriate for the use case.The
randommodule is correctly used here for simulating human-like mouse movements. The static analysis warnings (S311) about cryptographic suitability are false positives in this context, as the randomness is for UI simulation, not security purposes.
There was a problem hiding this comment.
Actionable comments posted: 1
♻️ Duplicate comments (1)
crawl4ai/async_crawler_strategy.py (1)
906-914: Good fix: viewport_size None guard added.Prevents TypeError and resolves the earlier review concern. Looks solid.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
crawl4ai/async_crawler_strategy.py(2 hunks)
🧰 Additional context used
🧬 Code graph analysis (1)
crawl4ai/async_crawler_strategy.py (1)
tests/async/test_dispatchers.py (1)
browser_config(17-18)
🪛 Ruff (0.14.0)
crawl4ai/async_crawler_strategy.py
918-918: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
919-919: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
920-920: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
921-921: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
925-925: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
926-926: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
929-929: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
940-940: Standard pseudo-random generators are not suitable for cryptographic purposes
(S311)
1 participant
…ental navigation
Summary
when set simulate_user=True
mouse click at 100,100 may trigger page navigation which may raise exception if process_iframes=Ture
(at : page.query_selector_all("iframe"))
Summary by CodeRabbit