ENH: return RangeIndex from difference, symmetric_difference

[jbrockmendel]

• closes ENH: Return RangeIndex for .difference and .symmetric_difference #12034
• tests added / passed
• passes black pandas
• passes git diff upstream/master -u -- "*.py" | flake8 --diff
• whatsnew entry

[jbrockmendel]

@iamlemec can you take a look and tell me what needs to be done here to be consistent with what you're doing in #36413?

[iamlemec]

@iamlemec can you take a look and tell me what needs to be done here to be consistent with what you're doing in #36413?

I think the most consistent way here would be to use _wrap_setop_result on the final return statement and _get_reconciled_name_object on the quick return statements up top. I actually didn't really touch difference and symmetric_difference in #36413, but it would make sense that they get a similar treatment.

My only concern regarding difference is that maybe since it's not a symmetric operation, it's naming rule should be left-biased. Like, if you're just dropping a few nameless elements, should that clobber the original name?

[jbrockmendel]

added name checks to tests

[jreback]

i guess this is technically a change, can you add a note; ping on green.

[jbrockmendel]

ping