fix: scan_full_page handles DOM-recycling virtual scroll sites (#731)

[umerkhan95]

Summary

Makes scan_full_page=True automatically detect and handle DOM-recycling virtual scroll pages without requiring VirtualScrollConfig or a container selector.

Before: scan_full_page=True on a virtual-scroll page captures only the last visible batch of recycled DOM elements.
After: All items captured — tested on skills.sh (9,707 elements) and 13 local patterns.

Fixes #731

List of files changed and why

• crawl4ai/async_crawler_strategy.py — Rewrote _handle_full_page_scan with a 5-phase pipeline: (1) setup, (2) detect recycling via fingerprint comparison + MutationObserver, (3) scroll + capture with dedup for each recycling container, (4) container-scroll pass for overflow-y/x containers, (5) fallback scroll-to-bottom for append-based pages. Supports horizontal, 2D grid, multiple containers, and nested scroll.
• crawl4ai/async_configs.py — Added max_no_change and max_captured_elements params to VirtualScrollConfig, plus from_dict forward-compat for unknown keys.
• crawl4ai/content_scraping_strategy.py — Preserve tail text when removing elements (prevents silent text loss in lxml).
• tests/test_virtual_scroll.py — 12 pytest tests for VirtualScrollConfig and scan_full_page paths.
• test_virtual_scroll_compat.py — 13 standalone tests covering all major virtualisation patterns (transform, innerHTML-wipe, append, container-scroll, variable heights, horizontal, 2D grid, multiple containers, nested, async-loaded, embedded widget, 1000-item stress, real site).

How Has This Been Tested?

• 13/13 local test patterns pass (each served from a self-contained HTML fixture via HTTPServer)
• Real-world validation on skills.sh: 9,707 elements recovered, 9,727 unique hrefs, 981 scroll steps
• Real-world validation on quotes.toscrape.com/scroll: 100/100 quotes captured
• Append-based infinite scroll regression test passes (no false positives on non-recycling pages)

Checklist

• My code follows the style guidelines of this project
• I have performed a self-review of my own code
• I have commented my code, particularly in hard-to-understand areas
• I have added/updated unit tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes