use pybridge to capture stderr for transcript tests

[crotwell]

use pybridge to capture stderr when running transcripts as tests. Think this addresses at least part of #1185, as stderr is part of the comparison. But this does not address outputing stderr when creating a transcript using history -t.

Possible that simple concatenation is not the right way to create the result for comparison, but at least any self.stderr output will be noticed by the test.

[codecov[bot]]

Codecov Report

Base: 98.58% // Head: 98.58% // Increases project coverage by +0.00% :tada:

Coverage data is based on head (65cae71) compared to base (031832a). Patch coverage: 100.00% of modified lines in pull request are covered.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1210   +/-   ##
=======================================
  Coverage   98.58%   98.58%           
=======================================
  Files          22       22           
  Lines        5705     5707    +2     
=======================================
+ Hits         5624     5626    +2     
  Misses         81       81           

Impacted Files	Coverage Δ
cmd2/transcript.py	93.10% <100.00%> (+0.12%)	:arrow_up:

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.

[tleonhardt]

@crotwell Phillip, thanks for the PR! Can you please run black to auto-format your changes? That way all of our tests will pass.

@anselor @kmvanbrunt You guys are more familiar with the pybridge stuff than I am, could one of you please review this PR and make sure it won't break any existing known workflows?

[tleonhardt]

@crotwell Formatting is still broken due to isort.

I added a new invoke task you can run with inv format which will auto-format code using both black and isort.

[tleonhardt]

@kmvanbrunt @anselor Now that all tests are passing, could one or both of you please review this PR?

IMO I don't like that it is only a partial solution. On the other hand, perhaps that is better than no solution?

[tleonhardt]

@crotwell Thank you very much for this PR. We appreciate your contribution.

My understanding is that this solution is incomplete and likely not the best long-term approach. This PR has existed for almost a year and I am going to close it now.

If someone would like to work on a more complete and appropriate solution in the future, that would be great.