update test results for no tags by default #573
Conversation
![mentatai[bot]](https://avatars.githubusercontent.com/in/850097?s=60&v=4){kind=small}
| @@ -213,20 +213,20 @@ def func_4(string): | |||
| mock_session_context.config.auto_context_tokens = 8000 | |||
|
|
|||
| code_message = await code_context.get_code_message(0, prompt="prompt") | |||
| assert mock_session_context.llm_api_handler.spice.count_tokens(code_message, "gpt-4", is_message=True) == 95 # Code | |||
| assert mock_session_context.llm_api_handler.spice.count_tokens(code_message, "gpt-4", is_message=True) == 85 # Code | |||
There was a problem hiding this comment.
Although the change from including tags like (search-result, user-included) in the code message output to a simplified version without these tags could make the output cleaner, it's important to consider if any downstream processes or users rely on these tags for context or filtering. If so, it would be beneficial to include a rationale or address potential impacts in the pull request description.
| @@ -51,7 +51,7 @@ async def test_partial_files(mocker, mock_session_context): | |||
| """\ | |||
| Code Files: | |||
|
|
|||
| dir/file.txt (user-included) | |||
| dir/file.txt | |||
There was a problem hiding this comment.
Similar to my previous comment, ensure that removing (user-included) from the code message doesn't negatively affect how users interact with or understand the output, especially in environments where clarity of source or inclusion reason might be significant.
|
MENTAT CODE REVIEW IN ACTIVE DEVELOPMENT. Only in use on mentat and internal repos. The pull request aims to update test conditions for |
Pull Request Checklist