Fix native HTTP server compatibility

[webknjaz]

• What kind of change does this PR introduce?

  • bug fix
  • feature
  • docs update
  • tests/coverage improvement
  • refactoring
  • other

• What is the related issue number (starting with #)
  #1377

• What is the current behavior? (You can also link to an open issue here)
  Various exceptions related to str/bytes py2/3 compat are being raised.

• What is the new behavior (if this is a feature change)?
  Will be working

• Other information:
  Still WIP, don't merge.
  I hope to get some help from @jaraco.

• Checklist:

  • I think the code is well written
  • I wrote good commit messages
  • I have squashed related commits together after the changes have been approved
  • Unit tests for the changes exist
  • Integration tests for the changes exist (if applicable)
  • I used the same coding conventions as the rest of the project
  • The new code doesn't generate linter offenses
  • Documentation reflects the changes
  • The PR relates to only one subject with a clear title
    and description in grammatically correct, complete sentences

[jaraco]

When I rebase your changes onto the bugfix/1377-add-tests branch, I notice that b936d6e fixes the failing test, suggesting that the second commit, 4c6b8ff, isn't necessary to fix the bug. Is there an additional bug that isn't caught by those tests, or can this second commit be omitted?

[jaraco]

I consider tonative to be deprecated. I'd like to avoid using it if possible and instead use either text or bytes explicitly on all Pythons.

[codecov]

Codecov Report

Merging #1712 into master will increase coverage by 0.45%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master    #1712      +/-   ##
==========================================
+ Coverage   80.35%   80.81%   +0.45%     
==========================================
  Files         104      105       +1     
  Lines       13546    13563      +17     
==========================================
+ Hits        10885    10961      +76     
+ Misses       2661     2602      -59

[jaraco]

I've stripped out the commit that seems to be unnecessary, including the one that fixes the reported issue. I believe this PR is ready to go.

[webknjaz]

@jaraco did it cause AppVeyor failures?

[webknjaz]

@jaraco there's or '' over the place defaulting to bytes/unicode all over the place. Shall we try from __future__ import unicode_strings?

[jaraco]

I'd very much like to do that. It was a feature that wasn't available in the early implementations (Python 2.3 compatibility and all), and even some people still advise against its use. It's my opinion, however, that using from __future__ import unicode_literals is the best way to achieve compatibility. Let's do it as time permits.

[webknjaz]

My concern is that from code it's not always obvious what kind of data the variable is expected to hold. Should we introduce some convention of prefixing b_ or suffixing _b to improve the situation?

[webknjaz]

@jaraco I guess we could explain what's going on in this test suite here.

[jaraco]

I do see appveyor failures; the failures seem unrelated, though I can't tell for sure. Is the startup/teardown of the native server causing conflicts with other tests?

[webknjaz]

I have no idea about the conflicts, but there's clearly some race conditions involved.

[webknjaz]

@jaraco it looks like this test doesn't catch the bug for some reason.

Ref:

• #1377 (comment)
• #1784 (comment)