MAINT: Condense TIMEZONE_IDS construction

[gfyoung]

Follow-up to #26596

[codecov]

Codecov Report

Merging #26600 into master will decrease coverage by <.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master   #26600      +/-   ##
==========================================
- Coverage   91.87%   91.86%   -0.01%     
==========================================
  Files         174      174              
  Lines       50661    50661              
==========================================
- Hits        46547    46542       -5     
- Misses       4114     4119       +5

Flag	Coverage Δ
#multiple	90.4% <ø> (ø)	⬆️
#single	41.76% <ø> (-0.11%)	⬇️

Impacted Files	Coverage Δ
pandas/io/gbq.py	78.94% <0%> (-10.53%)	⬇️
pandas/core/frame.py	97% <0%> (-0.12%)	⬇️
pandas/util/testing.py	90.6% <0%> (-0.11%)	⬇️

---

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 891a419...d51f7e2. Read the comment docs.

[h-vetinari]

@gfyoung @jreback this is IMO a step back - several of these objects don't have nice str/reprs, and it really makes sense to invest the extra 4(!) lines to have a clean output of pytest -v.

Comparing master vs this PR with the following dummy test:

def test_demo_tz_id(tz_aware_fixture):
    pass

this is the state of master:

test_demo_tz_id[UTC] PASSED                                                                   [  7%]
test_demo_tz_id[US/Eastern] PASSED                                                            [ 15%]
test_demo_tz_id[Asia/Tokyp] PASSED                                                            [ 23%]
test_demo_tz_id[dateutil/US/Pacific] PASSED                                                   [ 30%]
test_demo_tz_id[dateutil/Asia/Singapore] PASSED                                               [ 38%]
test_demo_tz_id[dateutil.tz.tzutz()] PASSED                                                   [ 46%]
test_demo_tz_id[dateutil.tz.tzlocal()] PASSED                                                 [ 53%]
test_demo_tz_id[pytz.FixedOffset(300)] PASSED                                                 [ 61%]
test_demo_tz_id[pytz.FixedOffset(0)] PASSED                                                   [ 69%]
test_demo_tz_id[pytz.FixedOffset(-300)] PASSED                                                [ 76%]
test_demo_tz_id[datetime.timezone.utc] PASSED                                                 [ 84%]
test_demo_tz_id[datetime.timezone.+1] PASSED                                                  [ 92%]
test_demo_tz_id[datetime.timezone.-1.named] PASSED                                            [100%]

resp. of this PR:

test_demo_tz_id[UTC] PASSED                                                                   [  7%]
test_demo_tz_id[US/Eastern] PASSED                                                            [ 15%]
test_demo_tz_id[Asia/Tokyo] PASSED                                                            [ 23%]
test_demo_tz_id[dateutil/US/Pacific] PASSED                                                   [ 30%]
test_demo_tz_id[dateutil/Asia/Singapore] PASSED                                               [ 38%]
test_demo_tz_id[tzutc()] PASSED                                                               [ 46%]
test_demo_tz_id[tzlocal()] PASSED                                                             [ 53%]
test_demo_tz_id[pytz.FixedOffset(300)] PASSED                                                 [ 61%]
test_demo_tz_id[<UTC>] PASSED                                                                 [ 69%]
test_demo_tz_id[pytz.FixedOffset(-300)] PASSED                                                [ 76%]
test_demo_tz_id[datetime.timezone.utc] PASSED                                                 [ 84%]
test_demo_tz_id[datetime.timezone(datetime.timedelta(seconds=3600))] PASSED                   [ 92%]
test_demo_tz_id[datetime.timezone(datetime.timedelta(days=-1, seconds=82800), 'foo')] PASSED  [100%]

In tests with more fixtures (like I'm working on in #25637), this would make the output absurdly wide.

[jreback]

comments

[gfyoung]

@jreback : If you could have another look. The failures seem unrelated...

(e.g. see this build here for a branch that I made off of master just now)

[gfyoung]

https://dev.azure.com/pandas-dev/pandas/_build/results?buildId=12180

Really not sure why numpy and scipy aren't playing nice together. They're the same versions as in this passing build on latest master:

https://dev.azure.com/pandas-dev/pandas/_build/results?buildId=12180

[jreback]

@gfyoung weird, can you merge master and repush and see

[gfyoung]

https://dev.azure.com/pandas-dev/pandas/_build/results?buildId=12207

@jreback : Alas, no luck after merging and repushing

[gfyoung]

@jreback : It seems that dropping the fixture decorator is causing issues with that one Azure case. Why that's the case? No idea - it's quite bizarre.

I'm inclined to just use repr to generate TIMEZONE_IDS for the time being, and we can then see later what's happening with that decorator.

Thoughts?

[jreback]

@gfyoung sounds fine

[gfyoung]

@jreback : Everything is cleaned up and green. PTAL.

[jreback]

thanks @gfyoung