Spaces:
Sleeping
Sleeping
Commit History
-
feat: Migrate to google-genai SDK + fix remaining low-res images
Nikhil Pravin Pise## Google SDK Migration (Future-proofing) - Migrated from deprecated google.generativeai to google.genai SDK - Updated all files: knowledge.py, embeddings.py, vision.py, multimedia.py - Updated server.py and app.py with new Client pattern - Updated requirements.txt to use google-genai package - Now uses gemini-2.0-flash-exp model for better performance ## Image Resolution Fix (Complete) - Enhanced upgrade_medium_image_url() to handle ALL Medium CDN formats: - miro.medium.com (v2/resize, max/, freeze/) - cdn-images-1.medium.com - cdn-images-2.medium.com - cdn.medium.com - Added 7 pattern matchers for comprehensive URL upgrading - All images now upgrade to 1400px width automatically This fully addresses the blurry image issue across all edge cases.545358b -
fix: Upgrade Medium images to high resolution (1400px) across entire app
Nikhil Pravin Pise- Added centralized upgrade_medium_image_url() and get_medium_image_url() to utils.py - Set MEDIUM_IMAGE_DEFAULT_WIDTH = 1400 as single source of truth - Updated parser.py to automatically upgrade search/tag result images - Updated paragraph_parser.py get_image_url() to use 1400px default - Updated html_renderer.py image sizes (700->1400px, 320->800px) - Updated app.py to use centralized image upgrade from utils.py Fixes low-resolution image thumbnails in search results and article previews60742a2
-
fix: restore block_resources function in utils.py
Nikhil Pravin Piseaa69bc4 -
feat: implement comprehensive improvement plan (Phases 1-5)
Nikhil Pravin PisePhase 1: Foundation - pyproject.toml with ruff, mypy, pytest config - .pre-commit-config.yaml with hooks - .github/workflows/ci.yml (7 CI jobs) - src/types.py, constants.py, http_pool.py Phase 2: Testing - 13 test files (unit, integration, e2e) - tests/conftest.py with fixtures Phase 3: Decomposition - ui/styles/ (6 CSS modules) - mcp/tools/ (scrape, search, export) - src/tiers/ (tier chain pattern) Phase 4: Hardening - src/validation.py, sanitization.py - src/exceptions.py, security.py Phase 5: Polish - src/logging_config.py (structured logging) - src/utils.py, metrics.py - docs/API.md, ARCHITECTURE.md Deployment - deploy/ folder for HF Spacese98cc10
-
feat: Switch to Groq as primary LLM provider
Nikhil Pravin Pise40cdc42
-
Clean up: remove old files, update docs with Hero tab and improvements
Nikhil Pravin Pisebbc5345
-
fix: add missing COOKIES_PATH to Config class
Nikhil Pravin Pise- Added COOKIES_PATH attribute for optional Medium authentication - Fixes HuggingFace deployment error: 'Config' has no attribute 'COOKIES_PATH' - Path set to medium_cookies.json in project root for HF deployment05c4508 -
feat: comprehensive migration - merge Scraper + MCP Server
Nikhil Pravin Pise- Added 17 missing source modules from Medium-Scraper - Replaced app.py with comprehensive 6-tab UI - Replaced server.py with full MCP server (10 tools) - Resolved dependency conflicts (mcp 1.0.0, elevenlabs 1.0.0) - Fixed import paths for HuggingFace deployment - Added shared_config.py to src/ - Created merged config.py without Apify dependency - Updated requirements.txt with 55 consolidated dependencies - Removed ../Medium-Scraper sibling directory references Ready for local Docker testing and HuggingFace deploymentae588db
-
Fix DDG import to use 'ddgs' package
Nikhil Pravin Pise992973b -
Fix DuckDuckGo warnings: use 'ddgs' package and 'auto' backend
Nikhil Pravin Pisedf015e7 -
Stop tracking files listed in .gitignore
Nikhil Pravin Pise28e7693 -
Sync scraper features, fix audio paths, improve image extraction, and add gitignore
Nikhil Pravin Pisea01fb11 -
Implement RSS Feed Fallback for ultimate robustness
Nikhil Pravin Pise1f1aaa9 -
Implement Tag Fallback and update DDG version
Nikhil Pravin Pisea8f12af
-
Try multiple DDG backends for robustness
Nikhil Pravin Pise3bf941f -
Add debug logging for DDG fallback
Nikhil Pravin Pise375a2bb -
Fix fallback logic to run on challenge detection
Nikhil Pravin Pise1f67e61 -
Add DuckDuckGo fallback for search to bypass blocking
Nikhil Pravin Pise33d137c -
Improve scraper robustness and add audio logging
Nikhil Pravin Pise31bfe02 -
Remove unused apify dependency to fix Pydantic conflict
Nikhil Pravin Pise2ea0fbc -
Initial deploy
Nikhil Pravin Pisea80eeb8