James Hiew

It should be possible to create a new prompt on the runtime path and have it loaded dynamically, then it can be hooked into a flow (TBD) Relates to #22

Be robust to text edits that happen while fetching a completion

3

If while a completion is being fetched, the text that was used in the prompt is edited, then the resulting completion can be inserted into the wrong place or result...

* document how to create a new prompt and make it available to nvim-magic * a flow for using a simple custom prompt (only {{{language}}} and {{{snippet}}} tags)

Could do the following healthchecks for the OpenAI backend * `curl` is present * `OPENAI_API_KEY` envvar is set

Currently they default to `300` for code completion flow and `1000` for suggestions/docstring flows but they should be easily overrideable as OpenAI charges per token generated.

e.g. change `nvim-magic: replaced code with suggested code (1877 characters)` to something like `nvim-magic: fetched code suggestion (1877 characters) in ~4.3s`

**rust-analyzer version**: (eg. output of "rust-analyzer: Show RA Version" command, accessible in VSCode via Ctrl/⌘+Shift+P) rust-analyzer version: 0.3.1162-standalone (634cfe3d7 2022-08-07) **rustc version**: (eg. output of `rustc -V`) rustc 1.63.0 (4b91a6ea7...

Support for headers/footers for each page

22

Is there a way to output arbitrary header/footers (e.g. page numbers) similar to the Save to PDF functionality of the Chrome browser? EDIT: I read https://github.com/arachnys/athenapdf/issues/16 right after posting this...

astroid.node_classes.Keyword lineno is None

1

### Steps to reproduce `astroid.node_classes.Keyword` does not seem to have a populated `lineno`. Its `repr` does show a line number though. Is this a bug? ```python #!/usr/bin/env python3 import astroid...