- Updated the minimum confidence level from 0.95 to 0.97 to enhance the accuracy of search results.
- Adjusted confidence calculation in the findLevenshteinMatch function to ensure more reliable matching.
- Introduced a new Git fallback strategy for handling edits when confidence levels are low.
- Replaced the previous Git edit strategy with a more robust approach that utilizes temporary directories and commits for original, search, and replace states.
- Enhanced error handling and logging for better debugging during the edit process.
- Updated the main edit function to incorporate the Git fallback strategy, ensuring it is attempted when other strategies fail.
- Improved overall structure and readability of the applyGitFallback function, streamlining the process of applying changes using Git.
- Improved the applyContextMatching function to better track context lines and handle changes more accurately.
- Updated the applyEdit function to enforce a stricter confidence check, now requiring a confidence level of 1 for successful edits.
- Refined the processing logic for addition and removal changes, ensuring more reliable diff application.
- Enhanced debugging capabilities with clearer logging and validation checks for edit results.
- Enhanced the applyContextMatching and applyDMP functions to improve handling of context and edit changes.
- Updated confidence validation logic to ensure stricter checks, now requiring a minimum confidence of 1 for successful edits.
- Refined the way changes are processed, including better tracking of removal and addition changes.
- Improved the validation of edit results by incorporating strategy-specific checks and logging for better debugging.
- Adjusted the applyEdit function to ensure strategies are applied in a more robust manner, with clearer handling of debug mode.
Added useEffect hooks to GlamaModelPicker, OpenAiModelPicker, and OpenRouterModelPicker components to ensure the search term stays synchronized with the selected model ID from apiConfiguration. This prevents the search term from getting out of sync when the model is changed.
