Standardize binary operations between int and str columns

[xinrong-meng]

Notes

• New test cases related to Pandas only are added for PR reviewers to visualize Pandas' behavior.
  They are commented out and will be removed before the PR merge (after PR review).
• This PR doesn't aim to cover all binary operations: only +, -, *, /, //, % are concerned.

Proposal

Make behaviors of binary operations (+, -, *, /, //, %) between int and str columns consistent with respective pandas behaviors.

• Standardize binary operations as follows:

  +: raise TypeError between int column and str column (or string literal)
  *: act as spark SQL repeat between int column(or int literal) and str columns; raise TypeError if a string literal is involved
  -, /, //, %(modulo): raise TypeError if a str column (or string literal) is involved

• Add def repeat(col, n): in databricks/koalas/spark/functions.py

  repeat defined in scala API only accepts integer literal as the method's second parameter.
  But internal StringRepeat accepts IntegerType as the method's second parameter.

  In order to pass int columns as the second parameter, we take advantage of callUDF.

Test

databricks/koalas/tests/test_dataframe.py

Resolves #1819

[itholic]

Just FYI: You can run formatter and linter using dev/reformat and dev/lint-python to perform static analysis. dev/reformat automatically rearrange your code if it's needed :)

[xinrong-meng]

@itholic Thank you for reminding me!

[itholic]

Otherwise, Looks fine to me.

[ueshin]

LGTM.

[ueshin]

Thanks! merging.

[xinrong-meng]

@ueshin Thank you for merging!

[itholic]

Nice 👍