BUG: Fixed tput output on windows

[TomAugspurger]

This came up in dask. For reasons I don't understand, you can end up with a bunch of warnings like tput: terminal attributes: No such device or address when pandas calls get_terminal_size

I was able to monkey patch those calls with shutil.get_terminal_size and verify that the warnings were fixed. However that's python3 only. This change uses shutil.get_terminal_size when possible, and falls back to our prior implementation otherwise.

Not sure about unit tests, but manually, the output matches on my mac.

[codecov]

Codecov Report

Merging #16496 into master will decrease coverage by 0.04%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #16496      +/-   ##
==========================================
- Coverage   90.43%   90.38%   -0.05%     
==========================================
  Files         161      161              
  Lines       51045    51049       +4     
==========================================
- Hits        46161    46140      -21     
- Misses       4884     4909      +25

Flag	Coverage Δ
#multiple	88.22% <100%> (-0.05%)	⬇️
#single	40.12% <100%> (-0.05%)	⬇️

Impacted Files	Coverage Δ
pandas/io/formats/terminal.py	16.43% <100%> (-31.39%)	⬇️

---

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 e41fe7f...d35dcad. Read the comment docs.

[codecov]

Codecov Report

Merging #16496 into master will decrease coverage by 0.04%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master   #16496      +/-   ##
==========================================
- Coverage   90.43%   90.38%   -0.05%     
==========================================
  Files         161      161              
  Lines       51045    51049       +4     
==========================================
- Hits        46161    46140      -21     
- Misses       4884     4909      +25

Flag	Coverage Δ
#multiple	88.22% <100%> (-0.05%)	⬇️
#single	40.12% <100%> (-0.05%)	⬇️

Impacted Files	Coverage Δ
pandas/io/formats/terminal.py	16.43% <100%> (-31.39%)	⬇️

---

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 e41fe7f...d35dcad. Read the comment docs.

[TomAugspurger]

Any thoughts on this @jreback? I don't think it should break anything.

[jreback]

lgtm.