Add slotscheck

[ariebovenberg]

I have made things!

@sobolevn kindly invited me to add slotscheck to the CI.

Checklist

• I have double checked that there are no unrelated changes in this pull request (old patches, accidental config files, etc)
• I have created at least one test case for the changes I have made n/a
• I have updated the documentation for the changes I have made
• I have added my changes to the CHANGELOG.md

Related issues

related: #1147, #1144

Open questions

• I've kept the same slots rules as present in test_slots.py. This means returns.contrib.mypy is not checked. We can either:
  a) add slots to mypy plugin classes
  b) keep the current ignore
  c) run slotscheck on this part of the code with more relaxed settings (there is an issue open multi-setting support if you have an opinion)
• The UnwrapFailedError class reports an error. This is because BaseException actually has a __dict__. We can either:
  a) remove slots from UnwrapFailedError
  b) keep the current ignore

[sobolevn]

Thanks!

[sobolevn]

Suggested change

	slotscheck = "^0.5.1"
	slotscheck = "^0.5"

[sobolevn]

Why do we exclude it? 🤔
I think it has __slots__: https://github.com/dry-python/returns/blob/master/returns/primitives/exceptions.py#L12

[ariebovenberg]

BaseException (Python stdlib) doesn't have slots (see questions in 'open questions' in top comment)

[sobolevn]

I've kept the same slots rules as present in test_slots.py. This means returns.contrib.mypy is not checked. We can either:

Let's go with b. No need to check mypy plugin for slots.

The UnwrapFailedError class reports an error. This is because BaseException actually has a dict. We can either:

I think having a __slot__ definition is still makes attr access a bit faster, doesn't it? If so, I am happy with special ignoring it.

[ariebovenberg]

• Indeed there is still an advantage to slots, even if the base class doesn't support it. It may be common enough to add a --no-require-exceptions flag to slotscheck to disable slots check for exception subclasses.
• any idea what the docs error is about? (Before I start debugging locally)

[codecov]

Codecov Report

Merging #1233 (99e954d) into master (af08071) will not change coverage.
The diff coverage is n/a.

@@            Coverage Diff            @@
##            master     #1233   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files           79        79           
  Lines         2434      2434           
  Branches       211       211           
=========================================
  Hits          2434      2434           

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update af08071...99e954d. Read the comment docs.

[sobolevn]

@ariebovenberg it is unrelated, I broke it recently somehow (sphinx 😒 ).

Fix in a separate PR is highly appreciated.

[sobolevn]

Thanks again!

[ariebovenberg]

@ariebovenberg it is unrelated, I broke it recently somehow (sphinx 😒 ).

Fix in a separate PR is highly appreciated.

An issue is a start: #1234