f6f7f1b5516e47b85e850183636132c589311419
209 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Nasrin
|
f6f7f1b551 |
Release v0.8.0: Crash Recovery, Prefetch Mode & Security Fixes (#1712)
* Fix: Use correct URL variable for raw HTML extraction (#1116) - Prevents full HTML content from being passed as URL to extraction strategies - Added unit tests to verify raw HTML and regular URL processing Fix: Wrong URL variable used for extraction of raw html * Fix #1181: Preserve whitespace in code blocks during HTML scraping 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. * Refactor Pydantic model configuration to use ConfigDict for arbitrary types * Fix EmbeddingStrategy: Uncomment response handling for the variations and clean up mock data. ref #1621 * Fix: permission issues with .cache/url_seeder and other runtime cache dirs. ref #1638 * fix: ensure BrowserConfig.to_dict serializes proxy_config * feat: make LLM backoff configurable end-to-end - extend LLMConfig with backoff delay/attempt/factor fields and thread them through LLMExtractionStrategy, LLMContentFilter, table extraction, and Docker API handlers - expose the backoff parameter knobs on perform_completion_with_backoff/aperform_completion_with_backoff and document them in the md_v2 guides * reproduced AttributeError from #1642 * pass timeout parameter to docker client request * added missing deep crawling objects to init * generalized query in ContentRelevanceFilter to be a str or list * import modules from enhanceable deserialization * parameterized tests * Fix: capture current page URL to reflect JavaScript navigation and add test for delayed redirects. ref #1268 * refactor: replace PyPDF2 with pypdf across the codebase. ref #1412 * Add browser_context_id and target_id parameters to BrowserConfig Enable Crawl4AI to connect to pre-created CDP browser contexts, which is essential for cloud browser services that pre-create isolated contexts. Changes: - Add browser_context_id and target_id parameters to BrowserConfig - Update from_kwargs() and to_dict() methods - Modify BrowserManager.start() to use existing context when provided - Add _get_page_by_target_id() helper method - Update get_page() to handle pre-existing targets - Add test for browser_context_id functionality This enables cloud services to: 1. Create isolated CDP contexts before Crawl4AI connects 2. Pass context/target IDs to BrowserConfig 3. Have Crawl4AI reuse existing contexts instead of creating new ones * Add cdp_cleanup_on_close flag to prevent memory leaks in cloud/server scenarios * Fix: add cdp_cleanup_on_close to from_kwargs * Fix: find context by target_id for concurrent CDP connections * Fix: use target_id to find correct page in get_page * Fix: use CDP to find context by browserContextId for concurrent sessions * Revert context matching attempts - Playwright cannot see CDP-created contexts * Add create_isolated_context flag for concurrent CDP crawls When True, forces creation of a new browser context instead of reusing the default context. Essential for concurrent crawls on the same browser to prevent navigation conflicts. * Add context caching to create_isolated_context branch Uses contexts_by_config cache (same as non-CDP mode) to reuse contexts for multiple URLs with same config. Still creates new page per crawl for navigation isolation. Benefits batch/deep crawls. * Add init_scripts support to BrowserConfig for pre-page-load JS injection This adds the ability to inject JavaScript that runs before any page loads, useful for stealth evasions (canvas/audio fingerprinting, userAgentData). - Add init_scripts parameter to BrowserConfig (list of JS strings) - Apply init_scripts in setup_context() via context.add_init_script() - Update from_kwargs() and to_dict() for serialization * Fix CDP connection handling: support WS URLs and proper cleanup Changes to browser_manager.py: 1. _verify_cdp_ready(): Support multiple URL formats - WebSocket URLs (ws://, wss://): Skip HTTP verification, Playwright handles directly - HTTP URLs with query params: Properly parse with urlparse to preserve query string - Fixes issue where naive f"{cdp_url}/json/version" broke WS URLs and query params 2. close(): Proper cleanup when cdp_cleanup_on_close=True - Close all sessions (pages) - Close all contexts - Call browser.close() to disconnect (doesn't terminate browser, just releases connection) - Wait 1 second for CDP connection to fully release - Stop Playwright instance to prevent memory leaks This enables: - Connecting to specific browsers via WS URL - Reusing the same browser with multiple sequential connections - No user wait needed between connections (internal 1s delay handles it) Added tests/browser/test_cdp_cleanup_reuse.py with comprehensive tests. * Update gitignore * Some debugging for caching * Add _generate_screenshot_from_html for raw: and file:// URLs Implements the missing method that was being called but never defined. Now raw: and file:// URLs can generate screenshots by: 1. Loading HTML into a browser page via page.set_content() 2. Taking screenshot using existing take_screenshot() method 3. Cleaning up the page afterward This enables cached HTML to be rendered with screenshots in crawl4ai-cloud. * Add PDF and MHTML support for raw: and file:// URLs - Replace _generate_screenshot_from_html with _generate_media_from_html - New method handles screenshot, PDF, and MHTML in one browser session - Update raw: and file:// URL handlers to use new method - Enables cached HTML to generate all media types * Add crash recovery for deep crawl strategies Add optional resume_state and on_state_change parameters to all deep crawl strategies (BFS, DFS, Best-First) for cloud deployment crash recovery. Features: - resume_state: Pass saved state to resume from checkpoint - on_state_change: Async callback fired after each URL for real-time state persistence to external storage (Redis, DB, etc.) - export_state(): Get last captured state manually - Zero overhead when features are disabled (None defaults) State includes visited URLs, pending queue/stack, depths, and pages_crawled count. All state is JSON-serializable. * Fix: HTTP strategy raw: URL parsing truncates at # character The AsyncHTTPCrawlerStrategy.crawl() method used urlparse() to extract content from raw: URLs. This caused HTML with CSS color codes like #eee to be truncated because # is treated as a URL fragment delimiter. Before: raw:body{background:#eee} -> parsed.path = 'body{background:' After: raw:body{background:#eee} -> raw_content = 'body{background:#eee' Fix: Strip the raw: or raw:// prefix directly instead of using urlparse, matching how the browser strategy handles it. * Add base_url parameter to CrawlerRunConfig for raw HTML processing When processing raw: HTML (e.g., from cache), the URL parameter is meaningless for markdown link resolution. This adds a base_url parameter that can be set explicitly to provide proper URL resolution context. Changes: - Add base_url parameter to CrawlerRunConfig.__init__ - Add base_url to CrawlerRunConfig.from_kwargs - Update aprocess_html to use base_url for markdown generation Usage: config = CrawlerRunConfig(base_url='https://example.com') result = await crawler.arun(url='raw:{html}', config=config) * Add prefetch mode for two-phase deep crawling - Add `prefetch` parameter to CrawlerRunConfig - Add `quick_extract_links()` function for fast link extraction - Add short-circuit in aprocess_html() for prefetch mode - Add 42 tests (unit, integration, regression) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * Updates on proxy rotation and proxy configuration * Add proxy support to HTTP crawler strategy * Add browser pipeline support for raw:/file:// URLs - Add process_in_browser parameter to CrawlerRunConfig - Route raw:/file:// URLs through _crawl_web() when browser operations needed - Use page.set_content() instead of goto() for local content - Fix cookie handling for non-HTTP URLs in browser_manager - Auto-detect browser requirements: js_code, wait_for, screenshot, etc. - Maintain fast path for raw:/file:// without browser params Fixes #310 * Add smart TTL cache for sitemap URL seeder - Add cache_ttl_hours and validate_sitemap_lastmod params to SeedingConfig - New JSON cache format with metadata (version, created_at, lastmod, url_count) - Cache validation by TTL expiry and sitemap lastmod comparison - Auto-migration from old .jsonl to new .json format - Fixes bug where incomplete cache was used indefinitely * Update URL seeder docs with smart TTL cache parameters - Add cache_ttl_hours and validate_sitemap_lastmod to parameter table - Document smart TTL cache validation with examples - Add cache-related troubleshooting entries - Update key features summary * Add MEMORY.md to gitignore * Docs: Add multi-sample schema generation section Add documentation explaining how to pass multiple HTML samples to generate_schema() for stable selectors that work across pages with varying DOM structures. Includes: - Problem explanation (fragile nth-child selectors) - Solution with code example - Key points for multi-sample queries - Comparison table of fragile vs stable selectors * Fix critical RCE and LFI vulnerabilities in Docker API deployment Security fixes for vulnerabilities reported by ProjectDiscovery: 1. Remote Code Execution via Hooks (CVE pending) - Remove __import__ from allowed_builtins in hook_manager.py - Prevents arbitrary module imports (os, subprocess, etc.) - Hooks now disabled by default via CRAWL4AI_HOOKS_ENABLED env var 2. Local File Inclusion via file:// URLs (CVE pending) - Add URL scheme validation to /execute_js, /screenshot, /pdf, /html - Block file://, javascript:, data: and other dangerous schemes - Only allow http://, https://, and raw: (where appropriate) 3. Security hardening - Add CRAWL4AI_HOOKS_ENABLED=false as default (opt-in for hooks) - Add security warning comments in config.yml - Add validate_url_scheme() helper for consistent validation Testing: - Add unit tests (test_security_fixes.py) - 16 tests - Add integration tests (run_security_tests.py) for live server Affected endpoints: - POST /crawl (hooks disabled by default) - POST /crawl/stream (hooks disabled by default) - POST /execute_js (URL validation added) - POST /screenshot (URL validation added) - POST /pdf (URL validation added) - POST /html (URL validation added) Breaking changes: - Hooks require CRAWL4AI_HOOKS_ENABLED=true to function - file:// URLs no longer work on API endpoints (use library directly) * Enhance authentication flow by implementing JWT token retrieval and adding authorization headers to API requests * Add release notes for v0.7.9, detailing breaking changes, security fixes, new features, bug fixes, and documentation updates * Add release notes for v0.8.0, detailing breaking changes, security fixes, new features, bug fixes, and documentation updates Documentation for v0.8.0 release: - SECURITY.md: Security policy and vulnerability reporting guidelines - RELEASE_NOTES_v0.8.0.md: Comprehensive release notes - migration/v0.8.0-upgrade-guide.md: Step-by-step migration guide - security/GHSA-DRAFT-RCE-LFI.md: GitHub security advisory drafts - CHANGELOG.md: Updated with v0.8.0 changes Breaking changes documented: - Docker API hooks disabled by default (CRAWL4AI_HOOKS_ENABLED) - file:// URLs blocked on Docker API endpoints Security fixes credited to Neo by ProjectDiscovery * Add examples for deep crawl crash recovery and prefetch mode in documentation * Release v0.8.0: The v0.8.0 Update - Updated version to 0.8.0 - Added comprehensive demo and release notes - Updated all documentation * Update security researcher acknowledgment with a hyperlink for Neo by ProjectDiscovery * Add async agenerate_schema method for schema generation - Extract prompt building to shared _build_schema_prompt() method - Add agenerate_schema() async version using aperform_completion_with_backoff - Refactor generate_schema() to use shared prompt builder - Fixes Gemini/Vertex AI compatibility in async contexts (FastAPI) * Fix: Enable litellm.drop_params for O-series/GPT-5 model compatibility O-series (o1, o3) and GPT-5 models only support temperature=1. Setting litellm.drop_params=True auto-drops unsupported parameters instead of throwing UnsupportedParamsError. Fixes temperature=0.01 error for these models in LLM extraction. --------- Co-authored-by: rbushria <rbushri@gmail.com> Co-authored-by: AHMET YILMAZ <tawfik@kidocode.com> Co-authored-by: Soham Kukreti <kukretisoham@gmail.com> Co-authored-by: Chris Murphy <chris.murphy@klaviyo.com> Co-authored-by: unclecode <unclecode@kidocode.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> |
||
|
Aravind
|
0024c82cdc | Sponsors/new (#1637) | ||
|
Nasrin
|
466be69e72 |
Merge pull request #1607 from unclecode/fix/dfs_deep_crawling
Fix/dfs deep crawling |
||
|
ntohidi
|
998c809e08 | Rename folder name for NSTProxy integration examples for crawl4ai | ||
|
Nasrin
|
cdcb8836b7 |
Merge pull request #1605 from Nstproxy/feat/nstproxy
feat: Add Nstproxy Proxies |
||
|
AHMET YILMAZ
|
1bd3de6a47 | #1510 : Add DFS deep crawler demonstration script and enhance DFS strategy with seen URL tracking | ||
|
nstproxy
|
80452166c8 | feat: Add Nstproxy Proxies | ||
|
CapSolver
|
57aeb70f00 | Add CapSolver Captcha Solver | ||
|
Nasrin
|
6534ece026 |
Merge pull request #1532 from unclecode/fix/update-documentation
Standardize C4A-Script tutorial, add CLI identity-based crawling, and add sponsorship CTA |
||
|
Nasrin
|
40173eeb73 |
Update Docker hooks and Webhook documents (#1557)
* fix(docker-api): migrate to modern datetime library API
Signed-off-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>
* Fix examples in README.md
* feat(docker): add user-provided hooks support to Docker API
Implements comprehensive hooks functionality allowing users to provide custom Python
functions as strings that execute at specific points in the crawling pipeline.
Key Features:
- Support for all 8 crawl4ai hook points:
• on_browser_created: Initialize browser settings
• on_page_context_created: Configure page context
• before_goto: Pre-navigation setup
• after_goto: Post-navigation processing
• on_user_agent_updated: User agent modification handling
• on_execution_started: Crawl execution initialization
• before_retrieve_html: Pre-extraction processing
• before_return_html: Final HTML processing
Implementation Details:
- Created UserHookManager for validation, compilation, and safe execution
- Added IsolatedHookWrapper for error isolation and timeout protection
- AST-based validation ensures code structure correctness
- Sandboxed execution with restricted builtins for security
- Configurable timeout (1-120 seconds) prevents infinite loops
- Comprehensive error handling ensures hooks don't crash main process
- Execution tracking with detailed statistics and logging
API Changes:
- Added HookConfig schema with code and timeout fields
- Extended CrawlRequest with optional hooks parameter
- Added /hooks/info endpoint for hook discovery
- Updated /crawl and /crawl/stream endpoints to support hooks
Safety Features:
- Malformed hooks return clear validation errors
- Hook errors are isolated and reported without stopping crawl
- Execution statistics track success/failure/timeout rates
- All hook results are JSON-serializable
Testing:
- Comprehensive test suite covering all 8 hooks
- Error handling and timeout scenarios validated
- Authentication, performance, and content extraction examples
- 100% success rate in production testing
Documentation:
- Added extensive hooks section to docker-deployment.md
- Security warnings about user-provided code risks
- Real-world examples using httpbin.org, GitHub, BBC
- Best practices and troubleshooting guide
ref #1377
* fix(deep-crawl): BestFirst priority inversion; remove pre-scoring truncation. ref #1253
Use negative scores in PQ to visit high-score URLs first and drop link cap prior to scoring; add test for ordering.
* docs: Update URL seeding examples to use proper async context managers
- Wrap all AsyncUrlSeeder usage with async context managers
- Update URL seeding adventure example to use "sitemap+cc" source, focus on course posts, and add stream=True parameter to fix runtime error
* fix(crawler): Removed the incorrect reference in browser_config variable #1310
* docs: update Docker instructions to use the latest release tag
* fix(docker): Fix LLM API key handling for multi-provider support
Previously, the system incorrectly used OPENAI_API_KEY for all LLM providers
due to a hardcoded api_key_env fallback in config.yml. This caused authentication
errors when using non-OpenAI providers like Gemini.
Changes:
- Remove api_key_env from config.yml to let litellm handle provider-specific env vars
- Simplify get_llm_api_key() to return None, allowing litellm to auto-detect keys
- Update validate_llm_provider() to trust litellm's built-in key detection
- Update documentation to reflect the new automatic key handling
The fix leverages litellm's existing capability to automatically find the correct
environment variable for each provider (OPENAI_API_KEY, GEMINI_API_TOKEN, etc.)
without manual configuration.
ref #1291
* docs: update adaptive crawler docs and cache defaults; remove deprecated examples (#1330)
- Replace BaseStrategy with CrawlStrategy in custom strategy examples (DomainSpecificStrategy, HybridStrategy)
- Remove “Custom Link Scoring” and “Caching Strategy” sections no longer aligned with current library
- Revise memory pruning example to use adaptive.get_relevant_content and index-based retention of top 500 docs
- Correct Quickstart note: default cache mode is CacheMode.BYPASS; instruct enabling with CacheMode.ENABLED
* fix(utils): Improve URL normalization by avoiding quote/unquote to preserve '+' signs. ref #1332
* feat: Add comprehensive website to API example with frontend
This commit adds a complete, web scraping API example that demonstrates how to get structured data from any website and use it like an API using the crawl4ai library with a minimalist frontend interface.
Core Functionality
- AI-powered web scraping with plain English queries
- Dual scraping approaches: Schema-based (faster) and LLM-based (flexible)
- Intelligent schema caching for improved performance
- Custom LLM model support with API key management
- Automatic duplicate request prevention
Modern Frontend Interface
- Minimalist black-and-white design inspired by modern web apps
- Responsive layout with smooth animations and transitions
- Three main pages: Scrape Data, Models Management, API Request History
- Real-time results display with JSON formatting
- Copy-to-clipboard functionality for extracted data
- Toast notifications for user feedback
- Auto-scroll to results when scraping starts
Model Management System
- Web-based model configuration interface
- Support for any LLM provider (OpenAI, Gemini, Anthropic, etc.)
- Simplified configuration requiring only provider and API token
- Add, list, and delete model configurations
- Secure storage of API keys in local JSON files
API Request History
- Automatic saving of all API requests and responses
- Display of request history with URL, query, and cURL commands
- Duplicate prevention (same URL + query combinations)
- Request deletion functionality
- Clean, simplified display focusing on essential information
Technical Implementation
Backend (FastAPI)
- RESTful API with comprehensive endpoints
- Pydantic models for request/response validation
- Async web scraping with crawl4ai library
- Error handling with detailed error messages
- File-based storage for models and request history
Frontend (Vanilla JS/CSS/HTML)
- No framework dependencies - pure HTML, CSS, JavaScript
- Modern CSS Grid and Flexbox layouts
- Custom dropdown styling with SVG arrows
- Responsive design for mobile and desktop
- Smooth scrolling and animations
Core Library Integration
- WebScraperAgent class for orchestration
- ModelConfig class for LLM configuration management
- Schema generation and caching system
- LLM extraction strategy support
- Browser configuration with headless mode
* fix(dependencies): add cssselect to project dependencies
Fixes bug reported in issue #1405
[Bug]: Excluded selector (excluded_selector) doesn't work
This commit reintroduces the cssselect library which was removed by PR (https://github.com/unclecode/crawl4ai/pull/1368) and merged via (
|
||
|
ntohidi
|
b74524fdfb | docs: update docker_hooks_examples.py with comprehensive examples and improved structure | ||
|
Nasrin
|
7cac008c10 |
Release/v0.7.6 (#1556)
* fix(docker-api): migrate to modern datetime library API
Signed-off-by: Emmanuel Ferdman <emmanuelferdman@gmail.com>
* Fix examples in README.md
* feat(docker): add user-provided hooks support to Docker API
Implements comprehensive hooks functionality allowing users to provide custom Python
functions as strings that execute at specific points in the crawling pipeline.
Key Features:
- Support for all 8 crawl4ai hook points:
• on_browser_created: Initialize browser settings
• on_page_context_created: Configure page context
• before_goto: Pre-navigation setup
• after_goto: Post-navigation processing
• on_user_agent_updated: User agent modification handling
• on_execution_started: Crawl execution initialization
• before_retrieve_html: Pre-extraction processing
• before_return_html: Final HTML processing
Implementation Details:
- Created UserHookManager for validation, compilation, and safe execution
- Added IsolatedHookWrapper for error isolation and timeout protection
- AST-based validation ensures code structure correctness
- Sandboxed execution with restricted builtins for security
- Configurable timeout (1-120 seconds) prevents infinite loops
- Comprehensive error handling ensures hooks don't crash main process
- Execution tracking with detailed statistics and logging
API Changes:
- Added HookConfig schema with code and timeout fields
- Extended CrawlRequest with optional hooks parameter
- Added /hooks/info endpoint for hook discovery
- Updated /crawl and /crawl/stream endpoints to support hooks
Safety Features:
- Malformed hooks return clear validation errors
- Hook errors are isolated and reported without stopping crawl
- Execution statistics track success/failure/timeout rates
- All hook results are JSON-serializable
Testing:
- Comprehensive test suite covering all 8 hooks
- Error handling and timeout scenarios validated
- Authentication, performance, and content extraction examples
- 100% success rate in production testing
Documentation:
- Added extensive hooks section to docker-deployment.md
- Security warnings about user-provided code risks
- Real-world examples using httpbin.org, GitHub, BBC
- Best practices and troubleshooting guide
ref #1377
* fix(deep-crawl): BestFirst priority inversion; remove pre-scoring truncation. ref #1253
Use negative scores in PQ to visit high-score URLs first and drop link cap prior to scoring; add test for ordering.
* docs: Update URL seeding examples to use proper async context managers
- Wrap all AsyncUrlSeeder usage with async context managers
- Update URL seeding adventure example to use "sitemap+cc" source, focus on course posts, and add stream=True parameter to fix runtime error
* fix(crawler): Removed the incorrect reference in browser_config variable #1310
* docs: update Docker instructions to use the latest release tag
* fix(docker): Fix LLM API key handling for multi-provider support
Previously, the system incorrectly used OPENAI_API_KEY for all LLM providers
due to a hardcoded api_key_env fallback in config.yml. This caused authentication
errors when using non-OpenAI providers like Gemini.
Changes:
- Remove api_key_env from config.yml to let litellm handle provider-specific env vars
- Simplify get_llm_api_key() to return None, allowing litellm to auto-detect keys
- Update validate_llm_provider() to trust litellm's built-in key detection
- Update documentation to reflect the new automatic key handling
The fix leverages litellm's existing capability to automatically find the correct
environment variable for each provider (OPENAI_API_KEY, GEMINI_API_TOKEN, etc.)
without manual configuration.
ref #1291
* docs: update adaptive crawler docs and cache defaults; remove deprecated examples (#1330)
- Replace BaseStrategy with CrawlStrategy in custom strategy examples (DomainSpecificStrategy, HybridStrategy)
- Remove “Custom Link Scoring” and “Caching Strategy” sections no longer aligned with current library
- Revise memory pruning example to use adaptive.get_relevant_content and index-based retention of top 500 docs
- Correct Quickstart note: default cache mode is CacheMode.BYPASS; instruct enabling with CacheMode.ENABLED
* fix(utils): Improve URL normalization by avoiding quote/unquote to preserve '+' signs. ref #1332
* feat: Add comprehensive website to API example with frontend
This commit adds a complete, web scraping API example that demonstrates how to get structured data from any website and use it like an API using the crawl4ai library with a minimalist frontend interface.
Core Functionality
- AI-powered web scraping with plain English queries
- Dual scraping approaches: Schema-based (faster) and LLM-based (flexible)
- Intelligent schema caching for improved performance
- Custom LLM model support with API key management
- Automatic duplicate request prevention
Modern Frontend Interface
- Minimalist black-and-white design inspired by modern web apps
- Responsive layout with smooth animations and transitions
- Three main pages: Scrape Data, Models Management, API Request History
- Real-time results display with JSON formatting
- Copy-to-clipboard functionality for extracted data
- Toast notifications for user feedback
- Auto-scroll to results when scraping starts
Model Management System
- Web-based model configuration interface
- Support for any LLM provider (OpenAI, Gemini, Anthropic, etc.)
- Simplified configuration requiring only provider and API token
- Add, list, and delete model configurations
- Secure storage of API keys in local JSON files
API Request History
- Automatic saving of all API requests and responses
- Display of request history with URL, query, and cURL commands
- Duplicate prevention (same URL + query combinations)
- Request deletion functionality
- Clean, simplified display focusing on essential information
Technical Implementation
Backend (FastAPI)
- RESTful API with comprehensive endpoints
- Pydantic models for request/response validation
- Async web scraping with crawl4ai library
- Error handling with detailed error messages
- File-based storage for models and request history
Frontend (Vanilla JS/CSS/HTML)
- No framework dependencies - pure HTML, CSS, JavaScript
- Modern CSS Grid and Flexbox layouts
- Custom dropdown styling with SVG arrows
- Responsive design for mobile and desktop
- Smooth scrolling and animations
Core Library Integration
- WebScraperAgent class for orchestration
- ModelConfig class for LLM configuration management
- Schema generation and caching system
- LLM extraction strategy support
- Browser configuration with headless mode
* fix(dependencies): add cssselect to project dependencies
Fixes bug reported in issue #1405
[Bug]: Excluded selector (excluded_selector) doesn't work
This commit reintroduces the cssselect library which was removed by PR (https://github.com/unclecode/crawl4ai/pull/1368) and merged via (
|
||
|
ntohidi
|
3efb59fb9a | fix: update Crawl4AI Docker container port from 11234 to 11235 | ||
|
ntohidi
|
b71d624168 | Merge branch 'implement-webhook-crawl-feature-011CULZY1Jy8N5MUkZqXkRVp' into develop | ||
|
ntohidi
|
d670dcde0a |
feat: add webhook support for /llm/job endpoint
Add comprehensive webhook notification support for the /llm/job endpoint, following the same pattern as the existing /crawl/job implementation. Changes: - Add webhook_config field to LlmJobPayload model (job.py) - Implement webhook notifications in process_llm_extraction() with 4 notification points: success, provider validation failure, extraction failure, and general exceptions (api.py) - Store webhook_config in Redis task data for job tracking - Initialize WebhookDeliveryService with exponential backoff retry logic Documentation: - Add Example 6 to WEBHOOK_EXAMPLES.md showing LLM extraction with webhooks - Update Flask webhook handler to support both crawl and llm_extraction tasks - Add TypeScript client examples for LLM jobs - Add comprehensive examples to docker_webhook_example.py with schema support - Clarify data structure differences between webhook and API responses Testing: - Add test_llm_webhook_feature.py with 7 validation tests (all passing) - Verify pattern consistency with /crawl/job implementation - Add implementation guide (WEBHOOK_LLM_JOB_IMPLEMENTATION.md) |
||
|
Claude
|
7388baa205 |
docs: add webhook example for Docker deployment
Added docker_webhook_example.py demonstrating: - Submitting crawl jobs with webhook configuration - Flask-based webhook receiver implementation - Three usage patterns: 1. Webhook notification only (fetch data separately) 2. Webhook with full data in payload 3. Traditional polling approach for comparison Includes comprehensive comments explaining: - Webhook payload structure - Authentication headers setup - Error handling - Production deployment tips Example is fully functional and ready to run with Flask installed. Generated with Claude Code https://claude.com/claude-code Co-Authored-By: Claude <noreply@anthropic.com> |
||
|
ntohidi
|
a3f057e19f |
feat: Add hooks utility for function-based hooks with Docker client integration. ref #1377
Add hooks_to_string() utility function that converts Python function objects to string representations for the Docker API, enabling developers to write hooks as regular Python functions instead of strings. Core Changes: - New hooks_to_string() utility in crawl4ai/utils.py using inspect.getsource() - Docker client now accepts both function objects and strings for hooks - Automatic detection and conversion in Crawl4aiDockerClient._prepare_request() - New hooks and hooks_timeout parameters in client.crawl() method Documentation: - Docker client examples with function-based hooks (docs/examples/docker_client_hooks_example.py) - Updated main Docker deployment guide with comprehensive hooks section - Added unit tests for hooks utility (tests/docker/test_hooks_utility.py) |
||
|
Soham Kukreti
|
7dfe528d43 |
fix(docs): standardize C4A-Script tutorial, add CLI identity-based crawling, and add sponsorship CTA
- Switch installs to pip install -r requirements.txt (tutorial and app docs) - Update local run steps to python server.py and http://localhost:8000 - Set default PORT to 8000; update port-in-use commands and alt port 8001 - Replace unsupported :contains() example with accessible attribute selector - Update example URLs in tutorial servers to 127.0.0.1:8000 - Add “Identity-based crawling” section with crwl profiles CLI workflow and code usage - Replace legacy-docs note with sponsorship message in docs/md_v2/index.md - Minor copy and consistency fixes across pages |
||
|
ntohidi
|
fef715a891 | Merge branch 'feature/docker-hooks' into develop | ||
|
ntohidi
|
3bc56dd028 |
fix: allow custom LLM providers for adaptive crawler embedding config. ref: #1291
- Change embedding_llm_config from Dict to Union[LLMConfig, Dict] for type safety
- Add backward-compatible conversion property _embedding_llm_config_dict
- Replace all hardcoded OpenAI embedding configs with configurable options
- Fix LLMConfig object attribute access in query expansion logic
- Add comprehensive example demonstrating multiple provider configurations
- Update documentation with both LLMConfig object and dictionary usage patterns
Users can now specify any LLM provider for query expansion in embedding strategy:
- New: embedding_llm_config=LLMConfig(provider='anthropic/claude-3', api_token='key')
- Old: embedding_llm_config={'provider': 'openai/gpt-4', 'api_token': 'key'} (still works)
|
||
|
Soham Kukreti
|
b1dff5a4d3 |
feat: Add comprehensive website to API example with frontend
This commit adds a complete, web scraping API example that demonstrates how to get structured data from any website and use it like an API using the crawl4ai library with a minimalist frontend interface. Core Functionality - AI-powered web scraping with plain English queries - Dual scraping approaches: Schema-based (faster) and LLM-based (flexible) - Intelligent schema caching for improved performance - Custom LLM model support with API key management - Automatic duplicate request prevention Modern Frontend Interface - Minimalist black-and-white design inspired by modern web apps - Responsive layout with smooth animations and transitions - Three main pages: Scrape Data, Models Management, API Request History - Real-time results display with JSON formatting - Copy-to-clipboard functionality for extracted data - Toast notifications for user feedback - Auto-scroll to results when scraping starts Model Management System - Web-based model configuration interface - Support for any LLM provider (OpenAI, Gemini, Anthropic, etc.) - Simplified configuration requiring only provider and API token - Add, list, and delete model configurations - Secure storage of API keys in local JSON files API Request History - Automatic saving of all API requests and responses - Display of request history with URL, query, and cURL commands - Duplicate prevention (same URL + query combinations) - Request deletion functionality - Clean, simplified display focusing on essential information Technical Implementation Backend (FastAPI) - RESTful API with comprehensive endpoints - Pydantic models for request/response validation - Async web scraping with crawl4ai library - Error handling with detailed error messages - File-based storage for models and request history Frontend (Vanilla JS/CSS/HTML) - No framework dependencies - pure HTML, CSS, JavaScript - Modern CSS Grid and Flexbox layouts - Custom dropdown styling with SVG arrows - Responsive design for mobile and desktop - Smooth scrolling and animations Core Library Integration - WebScraperAgent class for orchestration - ModelConfig class for LLM configuration management - Schema generation and caching system - LLM extraction strategy support - Browser configuration with headless mode |
||
|
ntohidi
|
bac92a47e4 | refactor: Update LLMTableExtraction examples and tests | ||
|
ntohidi
|
a51545c883 |
feat: 🚀 Introduce revolutionary LLMTableExtraction with intelligent chunking for massive tables
BREAKING CHANGE: Table extraction now uses Strategy Design Pattern This epic commit introduces a game-changing approach to table extraction in Crawl4AI: ✨ NEW FEATURES: - LLMTableExtraction: AI-powered extraction for complex HTML tables with rowspan/colspan - Smart Chunking: Automatically splits massive tables into optimal chunks at row boundaries - Parallel Processing: Processes multiple chunks simultaneously for blazing-fast extraction - Intelligent Merging: Seamlessly combines chunk results into complete tables - Header Preservation: Each chunk maintains context with original headers - Auto-retry Logic: Built-in resilience with configurable retry attempts 🏗️ ARCHITECTURE: - Strategy Design Pattern for pluggable table extraction strategies - ThreadPoolExecutor for concurrent chunk processing - Token-based chunking with configurable thresholds - Handles tables without headers gracefully ⚡ PERFORMANCE: - Process 1000+ row tables without timeout - Parallel processing with up to 5 concurrent chunks - Smart token estimation prevents LLM context overflow - Optimized for providers like Groq for massive tables 🔧 CONFIGURATION: - enable_chunking: Auto-handle large tables (default: True) - chunk_token_threshold: When to split (default: 3000 tokens) - min_rows_per_chunk: Meaningful chunk sizes (default: 10) - max_parallel_chunks: Concurrent processing (default: 5) 📚 BACKWARD COMPATIBILITY: - Existing code continues to work unchanged - DefaultTableExtraction remains the default strategy - Progressive enhancement approach This is the future of web table extraction - handling everything from simple tables to massive, complex data grids with merged cells and nested structures. The chunking is completely transparent to users while providing unprecedented scalability. |
||
|
ntohidi
|
be63c98db3 |
feat(docker): add user-provided hooks support to Docker API
Implements comprehensive hooks functionality allowing users to provide custom Python functions as strings that execute at specific points in the crawling pipeline. Key Features: - Support for all 8 crawl4ai hook points: • on_browser_created: Initialize browser settings • on_page_context_created: Configure page context • before_goto: Pre-navigation setup • after_goto: Post-navigation processing • on_user_agent_updated: User agent modification handling • on_execution_started: Crawl execution initialization • before_retrieve_html: Pre-extraction processing • before_return_html: Final HTML processing Implementation Details: - Created UserHookManager for validation, compilation, and safe execution - Added IsolatedHookWrapper for error isolation and timeout protection - AST-based validation ensures code structure correctness - Sandboxed execution with restricted builtins for security - Configurable timeout (1-120 seconds) prevents infinite loops - Comprehensive error handling ensures hooks don't crash main process - Execution tracking with detailed statistics and logging API Changes: - Added HookConfig schema with code and timeout fields - Extended CrawlRequest with optional hooks parameter - Added /hooks/info endpoint for hook discovery - Updated /crawl and /crawl/stream endpoints to support hooks Safety Features: - Malformed hooks return clear validation errors - Hook errors are isolated and reported without stopping crawl - Execution statistics track success/failure/timeout rates - All hook results are JSON-serializable Testing: - Comprehensive test suite covering all 8 hooks - Error handling and timeout scenarios validated - Authentication, performance, and content extraction examples - 100% success rate in production testing Documentation: - Added extensive hooks section to docker-deployment.md - Security warnings about user-provided code risks - Real-world examples using httpbin.org, GitHub, BBC - Best practices and troubleshooting guide ref #1377 |
||
|
Soham Kukreti
|
cd2dd68e4c |
docs: remove CRAWL4AI_API_TOKEN references and use correct endpoints in Docker example scripts (#1015)
- Remove deprecated API token authentication from all Docker examples
- Fix async job endpoints: /crawl -> /crawl/job for submission, /task/{id} -> /crawl/job/{id} for polling
- Fix sync endpoint: /crawl_sync -> /crawl (synchronous)
- Remove non-existent /crawl_direct endpoint
- Update request format to use new structure with browser_config and crawler_config
- Fix response handling for both async and sync calls
- Update extraction strategy format to use proper nested structure
- Add Ollama connectivity check before running tests
- Update test schemas and selectors for current website structures
This makes the Docker examples work out-of-the-box with the current API structure.
|
||
|
ntohidi
|
437395e490 | Merge branch 'feat/undetected-browser' into develop-future | ||
|
ntohidi
|
7a6ad547f0 |
Squashed commit of the following:
commit 2def6524cdacb69c72760bf55a41089257c0bb07 Author: ntohidi <nasrin@kidocode.com> Date: Mon Aug 4 18:59:10 2025 +0800 refactor: consolidate WebScrapingStrategy to use LXML implementation only BREAKING CHANGE: None - full backward compatibility maintained This commit simplifies the content scraping architecture by removing the redundant BeautifulSoup-based WebScrapingStrategy implementation and making it an alias for LXMLWebScrapingStrategy. Changes: - Remove ~1000 lines of BeautifulSoup-based WebScrapingStrategy code - Make WebScrapingStrategy an alias for LXMLWebScrapingStrategy - Update LXMLWebScrapingStrategy to inherit directly from ContentScrapingStrategy - Add required methods (scrap, ascrap, process_element, _log) to LXMLWebScrapingStrategy - Maintain 100% backward compatibility - existing code continues to work Code changes: - crawl4ai/content_scraping_strategy.py: Remove WebScrapingStrategy class, add alias - crawl4ai/async_configs.py: Remove WebScrapingStrategy from imports - crawl4ai/__init__.py: Update imports to show alias relationship - crawl4ai/types.py: Update type definitions - crawl4ai/legacy/web_crawler.py: Update import to use alias - tests/async/test_content_scraper_strategy.py: Update to use LXMLWebScrapingStrategy - docs/examples/scraping_strategies_performance.py: Update to use single strategy Documentation updates: - docs/md_v2/core/content-selection.md: Update scraping modes section - docs/md_v2/migration/webscraping-strategy-migration.md: Add migration guide - CHANGELOG.md: Document the refactoring under [Unreleased] Benefits: - 10-20x faster HTML parsing for large documents - Reduced memory usage and simplified codebase - Consistent parsing behavior - No migration required for existing users All existing code using WebScrapingStrategy continues to work without modification, while benefiting from LXML's superior performance. |
||
|
ntohidi
|
307fe28b32 |
fix: Correct URL matcher fallback behavior and improve memory monitoring
Fix critical issue where unmatched URLs incorrectly used the first config instead of failing safely. Also clarify that configs without url_matcher match ALL URLs by design, and improve memory usage monitoring. Bug fixes: - Change select_config() to return None when no config matches instead of using first config - Add proper error handling in dispatchers when no config matches a URL - Return failed CrawlResult with "No matching configuration found" error message - Fix is_match() to return True when url_matcher is None (matches all URLs) - Import and use get_true_memory_usage_percent() for more accurate memory monitoring Behavior clarification: - CrawlerRunConfig with url_matcher=None matches ALL URLs (not nothing) - This is the intended behavior for default/fallback configurations - Enables clean pattern: specific configs first, default config last Documentation updates: - Clarify that configs without url_matcher match everything - Explain "No matching configuration found" error when no default config - Add examples showing proper default config usage - Update all relevant docs: multi-url-crawling.md, arun_many.md, parameters.md - Simplify API config examples by removing extraction_strategy Demo and test updates: - Update demo_multi_config_clean.py with commented default config to show behavior - Change example URL to w3schools.com to demonstrate no-match scenario - Uncomment all test URLs in test_multi_config.py for comprehensive testing Breaking changes: None - this restores the intended behavior This ensures URLs only get processed with appropriate configs, preventing issues like HTML pages being processed with PDF extraction strategies. |
||
|
ntohidi
|
a03e68fa2f |
feat: Add URL-specific crawler configurations for multi-URL crawling
Implement dynamic configuration selection based on URL patterns to optimize crawling for different content types. This feature enables users to apply different crawling strategies (PDF extraction, content filtering, JavaScript execution) based on URL matching patterns. Key additions: - Add url_matcher and match_mode parameters to CrawlerRunConfig - Implement is_match() method supporting string patterns, functions, and mixed lists - Add MatchMode enum for OR/AND logic when combining multiple matchers - Update AsyncWebCrawler.arun_many() to accept List[CrawlerRunConfig] - Add select_config() method to dispatchers for runtime config selection - First matching config wins, with fallback to default Pattern matching supports: - Glob-style strings: *.pdf, */blog/*, *api* - Lambda functions: lambda url: 'github.com' in url - Mixed patterns with AND/OR logic for complex matching This enables optimal per-URL configuration: - PDFs: Use PDFContentScrapingStrategy without JavaScript - Blogs: Apply content filtering to reduce noise - APIs: Skip JavaScript, use JSON extraction - Dynamic sites: Execute only necessary JavaScript Breaking changes: None - fully backward compatible |
||
|
ntohidi
|
cf8badfe27 |
feat: cleanup unused code and enhance documentation for v0.7.1
- Remove unused StealthConfig from browser_manager.py
- Update LinkPreviewConfig import path in __init__.py and examples
- Fix infinity handling in content_scraping_strategy.py (use 0 instead of float('inf'))
- Remove sanitize_json_data functions from API endpoints
- Add comprehensive C4A Script documentation to release notes
- Update v0.7.0 release notes with improved code examples
- Create v0.7.1 release notes focusing on cleanup and documentation improvements
- Update demo files with corrected import paths and examples
- Fix virtual scroll and adaptive crawling examples across documentation
🤖 Generated with Claude Code
Co-Authored-By: Claude <noreply@anthropic.com>
|
||
|
unclecode
|
805c498adf |
docs: add simple anti-bot examples
- Add simple_anti_bot_examples.py with minimal code examples - Demonstrates stealth mode, undetected browser, and combined usage - Clean examples without logging for easy reference 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com> |
||
|
unclecode
|
6a728cbe5b |
feat: add stealth mode and enhance undetected browser support
- Add playwright-stealth integration with enable_stealth parameter in BrowserConfig - Merge undetected browser strategy into main async_crawler_strategy.py using adapter pattern - Add browser adapters (BrowserAdapter, PlaywrightAdapter, UndetectedAdapter) for flexible browser switching - Update install.py to install both playwright and patchright browsers automatically - Add comprehensive documentation for anti-bot features (stealth mode + undetected browser) - Create examples demonstrating stealth mode usage and comparison tests - Update pyproject.toml and requirements.txt with patchright>=1.49.0 and other dependencies - Remove duplicate/unused dependencies (alphashape, cssselect, pyperclip, shapely, selenium) - Add dependency checker tool in tests/check_dependencies.py Breaking changes: None - all existing functionality preserved 🤖 Generated with [Claude Code](https://claude.ai/code) Co-Authored-By: Claude <noreply@anthropic.com> |
||
|
unclecode
|
5c33cbcca2 | feat: add undetected browser support with adapter pattern | ||
|
UncleCode
|
0c8bb742b7 |
Release v0.7.0-r1: The Adaptive Intelligence Update
- Bump version to 0.7.0 - Add release notes and demo files - Update README with v0.7.0 features - Update Docker configurations for v0.7.0-r1 - Move v0.7.0 demo files to releases_review - Fix BM25 scoring bug in URLSeeder Major features: - Adaptive Crawling with pattern learning - Virtual Scroll support for infinite pages - Link Preview with 3-layer scoring - Async URL Seeder for massive discovery - Performance optimizations |
||
|
UncleCode
|
8794852a26 | Merge PR #1285: 2025 APR, MAY, and JUN bug fixes | ||
|
UncleCode
|
fb25a4a769 |
docs(examples): update crawl4ai showcase script
The crawl4ai showcase script has been significantly expanded to include more detailed examples and demonstrations. This includes live code examples, more detailed explanations, and a new real-world example. A new file, uv.lock, has also been added. |
||
|
ntohidi
|
0ebce590f8 | Merge branch '2025-JUN-1' into next-MAY | ||
|
UncleCode
|
1a73fb60db |
feat(crawl4ai): Implement adaptive crawling feature
This commit introduces the adaptive crawling feature to the crawl4ai project. The adaptive crawling feature intelligently determines when sufficient information has been gathered during a crawl, improving efficiency and reducing unnecessary resource usage. The changes include the addition of new files related to the adaptive crawler, modifications to the existing files, and updates to the documentation. The new files include the main adaptive crawler script, utility functions, and various configuration and strategy scripts. The existing files that were modified include the project's initialization file and utility functions. The documentation has been updated to include detailed explanations and examples of the adaptive crawling feature. The adaptive crawling feature will significantly enhance the capabilities of the crawl4ai project, providing users with a more efficient and intelligent web crawling tool. Significant modifications: - Added adaptive_crawler.py and related scripts - Modified __init__.py and utils.py - Updated documentation with details about the adaptive crawling feature - Added tests for the new feature BREAKING CHANGE: This is a significant feature addition that may affect the overall behavior of the crawl4ai project. Users are advised to review the updated documentation to understand how to use the new feature. Refs: #123, #456 |
||
|
UncleCode
|
a353515271 |
feat: Add virtual scroll support for modern web scraping
Add comprehensive virtual scroll handling to capture all content from pages that use DOM recycling techniques (Twitter, Instagram, etc). Key features: - New VirtualScrollConfig class for configuring virtual scroll behavior - Automatic detection of three scrolling scenarios: no change, content appended, content replaced - Intelligent HTML chunk capture and merging with deduplication - 100% content capture from virtual scroll pages - Seamless integration with existing extraction strategies - JavaScript-based detection and capture for performance - Tree-based DOM merging with text-based deduplication Documentation: - Comprehensive guide at docs/md_v2/advanced/virtual-scroll.md - API reference updates in parameters.md and page-interaction.md - Blog article explaining the solution and techniques - Complete examples with local test server Testing: - Full test suite achieving 100% capture of 1000 items - Examples for Twitter timeline, Instagram grid scenarios - Local test server with different scrolling behaviors This enables scraping of modern websites that were previously impossible to fully capture with traditional scrolling techniques. |
||
|
UncleCode
|
539a324cf6 |
refactor(link_extractor): remove link_extractor and rename to link_preview
This change removes the link_extractor module and renames it to link_preview, streamlining the codebase. The removal of 395 lines of code reduces complexity and improves maintainability. Other files have been updated to reflect this change, ensuring consistency across the project. BREAKING CHANGE: The link_extractor module has been deleted and replaced with link_preview. Update imports accordingly. |
||
|
UncleCode
|
5c9c305dbf |
feat: Add advanced link head extraction with three-layer scoring system (#1)
Squashed commit from feature/link-extractor branch implementing comprehensive link analysis: - Extract HTML head content from discovered links with parallel processing - Three-layer scoring: Intrinsic (URL quality), Contextual (BM25), and Total scores - New LinkExtractionConfig class for type-safe configuration - Pattern-based filtering for internal/external links - Comprehensive documentation and examples |
||
|
UncleCode
|
e528086341 |
test(async_assistant): add new tests for extract pipeline
Introduced two new test files to enhance coverage for the extract pipeline functionality. The tests aim to validate the behavior of the pipeline under various scenarios, ensuring robustness and reliability. No breaking changes. Closes issue #123. |
||
|
ntohidi
|
dc85481180 | refactor: Update LLM extraction example with the updated structure | ||
|
UncleCode
|
c0fd36982d | Update all documentation to import extraction strategies directly from crawl4ai. | ||
|
UncleCode
|
08a2cdae53 |
Add C4A-Script support and documentation
- Generate OneShot js code geenrator - Introduced a new C4A-Script tutorial example for login flow using Blockly. - Updated index.html to include Blockly theme and event editor modal for script editing. - Created a test HTML file for testing Blockly integration. - Added comprehensive C4A-Script API reference documentation covering commands, syntax, and examples. - Developed core documentation for C4A-Script, detailing its features, commands, and real-world examples. - Updated mkdocs.yml to include new C4A-Script documentation in navigation. |
||
|
UncleCode
|
ca03acbc82 | Add some new commands for the Crawl4ai script transpiler and creating an interactive tutorial that allows users to go through multiple steps and apply the syntax to automate the page. Fixed some issues and add several new commands for setting input values, variables, clearing input fields, and more. | ||
|
UncleCode
|
3f6f2e998c |
feat(script): add new scripting capabilities and documentation
This commit introduces a comprehensive set of new scripts and examples to enhance the scripting capabilities of the crawl4ai project. The changes include the addition of several Python scripts for compiling and executing scripts, as well as a variety of example scripts demonstrating different functionalities such as login flows, data extraction, and multi-step workflows. Additionally, detailed documentation has been created to guide users on how to utilize these new features effectively. The following significant modifications were made: - Added core scripting files: , , and . - Created a new documentation file to provide an overview of the new features. - Introduced multiple example scripts in the directory to showcase various use cases. - Updated and to integrate the new functionalities. - Added font assets for improved documentation presentation. These changes significantly expand the functionality of the crawl4ai project, allowing users to create more complex and varied scripts with ease. |
||
|
UncleCode
|
e731596315 | docs(tutorial_url_seeder): refine summary and next steps, enhance agentic design patterns section | ||
|
UncleCode
|
641526af81 | docs(tutorial_url_seeder): add advanced agentic patterns and implementation examples | ||
|
UncleCode
|
c6fc5c0518 |
docs(linkdin, url_seeder): update and reorganize LinkedIn data discovery and URL seeder documentation
This commit introduces significant updates to the LinkedIn data discovery documentation by adding two new Jupyter notebooks that provide detailed insights into data discovery processes. The previous workshop notebook has been removed to streamline the content and avoid redundancy. Additionally, the URL seeder documentation has been expanded with a new tutorial and several enhancements to existing scripts, improving usability and clarity. The changes include: - Added and for comprehensive LinkedIn data discovery. - Removed to eliminate outdated content. - Updated to reflect new data visualization requirements. - Introduced and to facilitate easier access to URL seeding techniques. - Enhanced existing Python scripts and markdown files in the URL seeder section for better documentation and examples. These changes aim to improve the overall documentation quality and user experience for developers working with LinkedIn data and URL seeding techniques. |