wrap generator #3
Conversation
![mentatbot[bot]](https://avatars.githubusercontent.com/in/837875?s=60&v=4){kind=small}
| @@ -18,7 +18,7 @@ class SpiceResponse: | |||
| # if stream=False, all attributes will be available immediately | |||
| # if stream=True, will error if trying to access attributes before stream is exhausted | |||
|
|
|||
| def __init__(self, stream_generator, full_text=None, cost=None, usage=None): | |||
| def __init__(self, stream_generator=None, full_text=None, cost=None, usage=None): | |||
There was a problem hiding this comment.
Consider implementing a more robust way to check if a generator is exhausted. Returning True by default may lead to incorrect behavior if the generator still has elements to yield. You could potentially wrap the generator to track its state or explore other methods to check its exhaustion.
|
MENTAT CODE REVIEW IN ACTIVE DEVELOPMENT. Only in use on mentat and internal repos. The changes introduced in this pull request seem to aim at improving the handling of streaming responses from different API providers. The approach of wrapping the stream generator is a good one, but there are a few areas where the implementation could be improved for clarity and robustness, particularly around the handling of generator state and the definition of variables within the scope of functions. |
No description provided.