Recursive function analysis

[AntoineBlaud]

Hi JusticeRage,

I was trying your tool, and first I found it super easy to use. Then, I got a little idea to improve a bit the analyze. When the current function calls others tiny functions without other calls inside, I think it will be great to first request analyze for the tiny one and rename it. Especially since it doesn't seem to be too wrong when it suggests a name for small functions. Then with some functions name filled, GPT-3 do a better job.

If you think the idea is good, I can make a pull request this week.

[JusticeRage]

Hi! While this is a good idea on paper, I think that going into functions recursively could be a little dangerous, especially when you get into the STL and the like where the rabbit hole can turn out to be pretty deep. It would be easy to have OpenAI costs spiral out of control for users.
Maybe in the future I'll think of doing this and implement a max-depth setting or something like this.

[AntoineBlaud]

Re JusticeRage,
Yes you're right. The cost can also quickly rise. But I was talking about calls to tiny functions without calls inside.
So MaxDepth=1
By the way I think that you-re uploading twice the function if we want both comments and variables name.
Given that the price depends of the size of the input, maybe we can find a way to upload it only once ?

[JusticeRage]

Correct, because the comments added during the first request help davinci-003 provide better suggestions for variable names.

[sudoaza]

I've experimented manually with this and have gotten great results. Knowing what the called functions do help the model see the big picture and understand better the function. You could add the analysis for the called functions as a comment before it is called and then pass it whole to the model, without displaying it to the user.

[JusticeRage]

This is definitely something I intend to look into in the near future.