Pylint fixes for .pyi files

[avylove]

Pylint recently started checking .pyi files but there are some messages that either don't make sense (ex: unused argument) or aren't strictly necessary in a type hints file (ex: no function docstring).

There is some discussion in pylint-dev/pylint#9096 of quieting these, but rather than wait, I figured we can fix the ones that made sense to fix and disable the others.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 95.42%. Comparing base (c28b53f) to head (8c36af6).

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #265   +/-   ##
=======================================
  Coverage   95.42%   95.42%           
=======================================
  Files           9        9           
  Lines        1027     1027           
  Branches      216      216           
=======================================
  Hits          980      980           
  Misses         43       43           
  Partials        4        4           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[avylove]

Might have to look at this a little more. Getting this (Although only in the pull_request test run not the push test run. The references are wrong, but still points to something.

blessed/colorspace.py:1: [R0401(cyclic-import), ] Cyclic import (blessed.sequences -> blessed.terminal)

[jquast]

looks like it comes from the blessed/sequences.pyi file, importing blessed/terminal.py to use the Terminal type

blessed/blessed/sequences.pyi

Line 6 in c28b53f

from .terminal import Terminal

I'm not sure what could be done, except to move all of sequences.* into terminal.* so that they are in the same file

[jquast]

I am ok with moving all of sequences into terminal, so long as there is some alias that allows from blessed.sequences import .. statements to continue to work, there are several test cases covering this already.

[avylove]

@jquast, sorry for the delay. I went the route of using a guard if TYPE_CHECKING: in the .pyi files and quoting Terminal in the type annotations. This seems to make Pylint happy without breaking anything else.

We could use from __future__ import annotations to avoid using quotes, but this only works in 3.7+ so could cause issues with someone using 3.5 or 3.6 with typing.

[jquast]

Thanks for finding a less radical fix :) no need to apologize, looks great, hope you are well!

[avylove]

Looks like there are some errors due to missing type annotations for the ESC_DELAY changes. Not sure why that should matter, but looks like it does. I'll do a PR for that soon. Of course it doesn't fail locally, so not sure what's different in CI or why it didn't fail when this PR ran it's tests.