You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
However, in practice, VSCode determines the type of language based on the code you input or by selecting the language for the current text being edited. Therefore, it should be possible to determine the language used in the current unnamed file through this information. Other extensions can achieve this, such as Copilot, Lingma (developed by Alibaba), or Twinny (https://github.com/twinnydotdev/twinny).
We should be able to do at least a decent job without needing to parse the AST. I actually think there might be a glaring bug where we don't pass any context to the model in unsaved files. I'll look into this and see if there is something so obvious
I have found that the 'continue' statement can autocomplete code in an unsaved Jupyter file, but it can't in other unsaved files. I hope this is helpful for you to find bugs.
Before submitting your bug report
Relevant environment info
Description
No response
To reproduce
No response
Log output
No response
The text was updated successfully, but these errors were encountered: