BUG: allow use of both default+input resolvers in df.eval, GH34966

[bubblingoak]

• closes DOC: Confusing 'resolvers' kwarg documentation for DataFrame.query() #34966
• tests added / passed (all tests passed except those which were auto-cancelled for taking too long)
• Ensure all linting tests pass, see here for how to run them
• whatsnew entry

Allows default resolvers to be used with input resolvers in pandas.dataframe.eval. Previously cases such as this would break:

import pandas as pd

df = pd.DataFrame({'a':[0,1,2], 'b':[7,8,9]}) 
d = {'c': 5}
df.eval('a+b*c', resolvers=[d])

with pandas.core.computation.ops.UndefinedVariableError: name 'a' is not defined as explained in #34966. Following this update, the above works as expected.
Added a test similar to the above: tests.frame.test_query_eval.TestDataFrameEval.test_eval_resolvers_combined

[jreback]

umm can you add a test which fails w/o this change

[pep8speaks]

Hello @bubblingoak! 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 2022-01-04 00:39:55 UTC

[bubblingoak]

umm can you add a test which fails w/o this change

Added the test tests.frame.test_query_eval.TestDataFrameEval.test_eval_resolvers_combined. The updated code passes, but current master fails on line 174:

result = df.eval("a + b * c", resolvers=[dict1])

with pandas.core.computation.ops.UndefinedVariableError: name 'a' is not defined

[jreback]

small commment, ping on green

[jreback]

thanks @bubblingoak