Preserve metaclass when slots=True

[YevIgn]

Fixes #154

[hynek]

fyi the tests fail on py2/pypy2

[YevIgn]

Sorry, doubt I can fix the new test, without bringing analog of six.add_metaclass to the code, or using an ugly hack with eval.

[Tinche]

Hm. Can we bring in six as a test dependency?

[hynek]

I don’t mind. I don't want runtime deps to check for PY3 but reimplementing metaclass magic just for tests seems like a bridge too far.

[codecov]

Codecov Report

Merging #155 into master will not change coverage.
The diff coverage is 100%.

@@          Coverage Diff          @@
##           master   #155   +/-   ##
=====================================
  Coverage     100%   100%           
=====================================
  Files           8      8           
  Lines         551    551           
  Branches      122    122           
=====================================
  Hits          551    551

Impacted Files	Coverage Δ
src/attr/_make.py	100% <100%> (ø)	✅

---

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 2c8bd46...39a4299. Read the comment docs.

[hynek]

Tin please take this; I have willfully forgotten everything I ever knew about meta classes.

[Tinche]

Sure thing. I'm sick atm though so it'll have to wait a little

[hynek]

Get well soon!

[Tinche]

LGTM. It's a simple change, instead of assuming the metaclass of the class we're changing is type, first fetch the metaclass with the type() builtin, then use that.

The docs are failing because of the logo (Warning, treated as error:

CHANGELOG.rst:None: WARNING: nonlocal image URI found: https://attrs.readthedocs.io/en/latest/_static/attrs_logo.png), so it's unrelated.

[hynek]

I’ve updated the branch and it passes now.

Thanks for the input and I hope you feel better!