CLN: use unicode literals instead of compat.u

[jbrockmendel]

This gets most of the non-test usages

[jbrockmendel]

Travis fail is pickle-plot that appears to be hitting everything right now

[jbrockmendel]

circle failure is Hypothesis

[codecov]

Codecov Report

Merging #24026 into master will decrease coverage by <.01%.
The diff coverage is 39.13%.

@@ Coverage Diff @@ ## master #24026 +/- ## ========================================== - Coverage 42.46% 42.46% -0.01%  ========================================== Files 161 161 Lines 51557 51555 -2 ========================================== - Hits 21892 21891 -1  + Misses 29665 29664 -1

Flag	Coverage Δ
#single	42.46% <39.13%> (-0.01%)	⬇️

Impacted Files	Coverage Δ
pandas/io/html.py	19.94% <0%> (ø)	⬆️
pandas/io/formats/latex.py	0% <0%> (ø)	⬆️
pandas/io/formats/html.py	0% <0%> (ø)	⬆️
pandas/core/arrays/integer.py	37.97% <100%> (ø)	⬆️
pandas/core/series.py	50.74% <100%> (ø)	⬆️
pandas/core/computation/pytables.py	92.07% <100%> (ø)	⬆️
pandas/io/parsers.py	48.54% <100%> (ø)	⬆️
pandas/core/indexes/base.py	55.93% <100%> (ø)	⬆️
pandas/core/config.py	61.12% <14.28%> (ø)	⬆️
pandas/core/arrays/categorical.py	41.89% <25%> (ø)	⬆️
... and 7 more

---

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 5b0610b...705c3e4. Read the comment docs.

[codecov]

Codecov Report

Merging #24026 into master will decrease coverage by <.01%.
The diff coverage is 95.58%.

@@ Coverage Diff @@ ## master #24026 +/- ## ========================================== - Coverage 92.21% 92.21% -0.01%  ========================================== Files 161 161 Lines 51684 51683 -1 ========================================== - Hits 47658 47657 -1  Misses 4026 4026

Flag	Coverage Δ
#multiple	90.61% <88.23%> (-0.01%)	⬇️
#single	43% <38.23%> (-0.01%)	⬇️

Impacted Files	Coverage Δ
pandas/core/config.py	87.04% <100%> (ø)	⬆️
pandas/io/html.py	92.66% <100%> (ø)	⬆️
pandas/core/arrays/categorical.py	95.4% <100%> (ø)	⬆️
pandas/io/json/json.py	93.09% <100%> (ø)	⬆️
pandas/core/series.py	93.68% <100%> (ø)	⬆️
pandas/core/panel.py	97.91% <100%> (ø)	⬆️
pandas/core/frame.py	97.02% <100%> (ø)	⬆️
pandas/io/formats/latex.py	100% <100%> (ø)	⬆️
pandas/util/testing.py	87.39% <100%> (ø)	⬆️
pandas/core/computation/pytables.py	92.37% <100%> (ø)	⬆️
... and 5 more

---

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 4b5f4d1...d637ef4. Read the comment docs.

[gfyoung]

Huh...that's a good catch! Maybe we can rid of it entirely perhaps later?

[jbrockmendel]

Maybe we can rid of it entirely perhaps later?

There's plenty of cases left with u("foo") that are easy to modernize. I didn't look too closely at the handful of u(bar) cases. Presumably it will all become simpler next month after py2 is dropped.

[jreback]

so the reason we have this is to provide a uniform way to do this across py2 and py3. you don't then have to worry about whether you are quoting unicode or not and just use u. Is there a reason you want to change this now, rather than just wholesale change after PY2 is dropped?

[jbrockmendel]

Is there a reason you want to change this now, rather than just wholesale change after PY2 is dropped?

Not especially. Just see cleanup opportunities when browsing through modules, figure being aggressive about them is better than the alternative.

so the reason we have this is to provide a uniform way to do this across py2 and py3. you don't then have to worry about whether you are quoting unicode or not and just use u

I'm guessing this got put in place before unicode literals got added back into py3. since there is a built-in way of doing this, that should be preferred over a home-built solution.

[jreback]

I'm guessing this got put in place before unicode literals got added back into py3. since there is a built-in way of doing this, that should be preferred over a home-built solution.

was this in 3.4? (e.g. this function was certainly done quite a long time ago).

[jbrockmendel]

Rebased. This is pretty low-priority, should either merge or close.

[jreback]

let's close for now.