The remove_empty_elements_fast() method was removing whitespace-only
span elements inside <pre> and <code> tags, causing import statements
like "import torch" to become "importtorch". Now skips elements inside
code blocks where whitespace is significant.
Major documentation restructuring to emphasize self-hosting capabilities and fully document the real-time monitoring system.
Changes:
- Renamed docker-deployment.md → self-hosting.md to better reflect the value proposition
- Updated mkdocs.yml navigation to "Self-Hosting Guide"
- Completely rewrote introduction emphasizing self-hosting benefits:
* Data privacy and ownership
* Cost control and transparency
* Performance and security advantages
* Full customization capabilities
- Expanded "Metrics & Monitoring" → "Real-time Monitoring & Operations" with:
* Monitoring Dashboard section documenting the /monitor UI
* Complete feature breakdown (system health, requests, browsers, janitor, errors)
* Monitor API Endpoints with all REST endpoints and examples
* WebSocket Streaming integration guide with Python examples
* Control Actions for manual browser management
* Production Integration patterns (Prometheus, custom dashboards, alerting)
* Key production metrics to track
- Enhanced summary section:
* What users learned checklist
* Why self-hosting matters
* Clear next steps
* Key resources with monitoring dashboard URL
The monitoring dashboard built 2-3 weeks ago is now fully documented and discoverable.
Users will understand they have complete operational visibility at http://localhost:11235/monitor
with real-time updates, browser pool management, and programmatic control via REST/WebSocket APIs.
This positions Crawl4AI as an enterprise-grade self-hosting solution with DevOps-level
monitoring capabilities, not just a Docker deployment.
execution, causing URLs to be processed sequentially instead of in parallel.
Changes:
- Added aperform_completion_with_backoff() using litellm.acompletion for async LLM calls
- Implemented arun() method in ExtractionStrategy base class with thread pool fallback
- Created async arun() and aextract() methods in LLMExtractionStrategy using asyncio.gather
- Updated AsyncWebCrawler.arun() to detect and use arun() when available
- Added comprehensive test suite to verify parallel execution
Impact:
- LLM extraction now runs truly in parallel across multiple URLs
- Significant performance improvement for multi-URL crawls with LLM strategies
- Backward compatible - existing extraction strategies continue to work
- No breaking changes to public API
Technical details:
- Uses litellm.acompletion for non-blocking LLM calls
- Leverages asyncio.gather for concurrent chunk processing
- Maintains backward compatibility via asyncio.to_thread fallback
- Works seamlessly with MemoryAdaptiveDispatcher and other dispatchers
- Add lightweight security test to verify version requirements
- Add comprehensive integration test for crawl4ai functionality
- Tests verify pyOpenSSL >= 25.3.0 and cryptography >= 45.0.7
- All tests passing: security vulnerability is resolved
Related to #1545🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Updates pyOpenSSL from >=24.3.0 to >=25.3.0
- This resolves CVE affecting cryptography package versions >=37.0.0 & <43.0.1
- pyOpenSSL 25.3.0 requires cryptography>=45.0.7, which is above the vulnerable range
- Fixes issue #1545🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
