Currently, the GPT Reviewer plugin performs a code review on every file in a commit, regardless of whether those files were reviewed in a previous commit. This approach incurs unnecessary costs when using the maximum tokens for code review. We should optimize the plugin to avoid redundant reviews and save money by utilizing the previous review results.
Expected Behavior:
The GPT Reviewer plugin should intelligently track the reviewed files and their associated reviews to avoid redundant code reviews. It should utilize the existing review results when the reviewed files have not changed since the previous commit, thereby reducing costs when employing the maximum tokens for code review.
Issue Description:Currently, the GPT Reviewer plugin performs a code review on every file in a commit, regardless of whether those files were reviewed in a previous commit. This approach incurs unnecessary costs when using the maximum tokens for code review. We should optimize the plugin to avoid redundant reviews and save money by utilizing the previous review results.
Expected Behavior:The GPT Reviewer plugin should intelligently track the reviewed files and their associated reviews to avoid redundant code reviews. It should utilize the existing review results when the reviewed files have not changed since the previous commit, thereby reducing costs when employing the maximum tokens for code review.