利用AI大模型,一键生成高清短视频 Generate short videos with one click using AI LLM.
refactor: optimize codebase for better performance and reliability
## Key Optimizations: ### 1. video.py (app/services/video.py) - Remove unused variable calculation (req_dur) - Fix resource access after close by saving clip.duration before closing - Improve error handling in delete_files() with proper logging - Better code clarity and maintainability ### 2. material.py (app/services/material.py) - Fix thread-safety issue: replace global counter with thread-safe implementation using threading.Lock - Fix resource leak: ensure VideoFileClip is properly closed using try-finally block - Improve exception handling and resource cleanup - Prevents potential race conditions in concurrent scenarios ### 3. video.py controller (app/controllers/v1/video.py) - Fix critical bug: use correct variable 'file' instead of 'v' in file_to_uri() - Prevents NameError and ensures proper URI generation for task files ### 4. utils.py (app/utils/utils.py) - Improve error logging in to_json() with proper error messages - Add exc_info=True to background thread error logging for better debugging - Make background threads non-daemon for more predictable behavior ## Impact: - Better thread-safety for concurrent video generation tasks - Reduced resource leaks and memory issues - Improved error visibility and debugging capability - More robust error handling throughout the codebase - Fixed potential runtime errors in URI generation ## Testing: - All changes are backward compatible - No breaking changes to public APIs - Improved reliability for production use
C
Claude committed
44096d608651e66063d3abd75063626fc6f8844a
Parent: 6cb5f23