ENH: add abiility for json.dumps to parse and add timezones

[jeliashi]

• closes to_json converts to UTC when encoding ISO formatted datetimes #12997
• tests modofied
• Ensure all linting tests pass, see here for how to run them
• whatsnew entry

Ok, so this is a doozy AND it's not done, but it can be merged as is.
What do we do:
We add timezones to json output!
Before:

ts = pd.Timestamp('2020-10-5 4:00:05', tz='US/Pacific')
print(pd.io.json.dumps(ts, iso_dates=True))
print(ts.isoformat())

would provide
"2020-10-05T11:00:05.000Z"
and
"2020-10-05T04:00:05.000-07:00"
respectively.

Now they both produce
"2020-10-05T04:00:05.000-07:00"

So what do I mean by not complete?
This still doesn't work for arrays that have datetime values. This only works for single datetime values.
So:
pd.DataFrame({'foo': range(5)}, index=pd.date_range('2020-10-12', freq='H', periods=5, tz='US/Eastern')).to_json(date_format='iso')

will still produce:
'{"foo":{"2020-10-12T04:00:00.000Z":0,"2020-10-12T05:00:00.000Z":1,"2020-10-12T06:00:00.000Z":2,"2020-10-12T07:00:00.000Z":3,"2020-10-12T08:00:00.000Z":4}}'

and not
'{"foo":{"2020-10-12T00:00:00.000-04:00":0,"2020-10-12T01:00:00.000-04:00":1,"2020-10-12T02:00:00.000-04:00":2,"2020-10-12T03:00:00.000-04:00":3,"2020-10-12T04:00:00.000-04:00":4}}'

Working on it currently. Open to recommendations!

[pep8speaks]

Hello @jeliashi! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! 🍻

Comment last updated at 2021-05-26 15:36:58 UTC

[simonjayhawkins]

@jeliashi can you close this one and carry on with #41573 if possible, much easier to track.

[jeliashi]

@jeliashi can you close this one and carry on with #41573 if possible, much easier to track.

sure thing