Radek Szymczyszyn

See also a screenshot of how the docs would be rendered in #456.

The idea is to make it handy in the shell, similar to `r3`, that's why the terseness. Another idea is that if we use that convention, then we could rewrite...

I implemented [`h()` for documentation access in docsh](https://github.com/erszcz/docsh/blob/b6d917e75270cd3220300fc00583e4d3cfa8b460/include/docsh_user_default.hrl) as a `user_default` extension before `shell_docs` existed and I think it's a viable option. However, I think `user_default` itself doesn't scale well...

Actually, I don't think `map_type_var` should pass - as the list element type is union of a map and an atom, I think the map pattern match might fail. I'll...

Looking forward to this happening 🤞

Hi, @gomoripeti! Thanks for the report. I checked the examples you provide with the current master branch (https://github.com/josefs/Gradualizer/commit/f647ae76b1510f6d3b367fec06dc2d68c7ede88c) and cannot reproduce the problem - no crash occurs. Could you confirm...

I think this can be closed now, WDYT @zuiderkwast?

Note to self: rechecking if that's still an issue with the current master branch would be insightful.

Yes, I think with simple guards like these it's possible to handle it. TBH I'm surprised it's not handled already, since we definitely have code for type refinement stepping through...

The original code is now tested against as https://github.com/josefs/Gradualizer/blob/e57106f4edb1a17288100f776e6c1c190920ec23/test/should_pass/minimised_gradualizer_fmt.erl and is checked correctly one the current master branch.