Always open README with utf-8 encoding

[andrewmwilson]

This is another regression in 0.14.0

Fixes: #600

[codecov-io]

Codecov Report

Merging #601 into master will not change coverage.
The diff coverage is n/a.

@@          Coverage Diff          @@
##           master   #601   +/-   ##
=====================================
  Coverage     100%   100%           
=====================================
  Files           7      7           
  Lines        1501   1501           
  Branches      225    225           
=====================================
  Hits         1501   1501

---

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 f77546d...421dfe8. Read the comment docs.

[jadchaar]

@andrewmwilson Looks good. Any difference between with codecs.open and with open("README.rst", "r", encoding="utf-8") as f:?

[andrewmwilson]

@jadchaar I don't think the encoding parameter exists in Python 2.

There is also io.open in Python 2.6+.

[jadchaar]

You are very right--codecs is fine! Also, mind adding encoding='utf-8' to the reading of init.py?

Then we can get this merged in.

[andrewmwilson]

Done!

[jadchaar]

Thanks for the contribution @andrewmwilson. Apologies for these regressions; hopefully this will be the last one :).

[jadchaar]

LGTM