fix incorrect call to right_hand_split()

[JelleZijlstra]

The second argument to right_hand_split() is an int (line length), but we were passing supports_trailing_commas as a bool. Mypy didn't catch this because bool is a subclass of int. Found this while working on a fix for #759 because I changed the type of supports_trailing_commas.

This affects one test that is now formatted differently, and it also causes a fair number of changes in our codebase. Now that I look at it, in most cases I actually feel like the old formatting is better, so maybe we need to do something different than just fix the type mismatch.

[coveralls]

Pull Request Test Coverage Report for Build 971

• 1 of 1 (100.0%) changed or added relevant line in 1 file are covered.
• 1 unchanged line in 1 file lost coverage.
• Overall coverage increased (+0.03%) to 95.953%

---

Files with Coverage Reduction	New Missed Lines	%
black.py	1	94.79%

Totals
Change from base Build 966:	0.03%
Covered Lines:	3011
Relevant Lines:	3138

---

💛 - Coveralls

[coveralls]

Pull Request Test Coverage Report for Build 971

• 1 of 1 (100.0%) changed or added relevant line in 1 file are covered.
• 1 unchanged line in 1 file lost coverage.
• Overall coverage increased (+0.03%) to 95.953%

---

Files with Coverage Reduction	New Missed Lines	%
black.py	1	94.79%

Totals
Change from base Build 966:	0.03%
Covered Lines:	3011
Relevant Lines:	3138

---

💛 - Coveralls

[ambv]

Oh my, this one's embarrassing! It dates all the way back to 3eea3aa when I was fixing trailers (content with brackets) being unnecessarily exploded into their own lines after a dedented closing bracket.

[ambv]

@JelleZijlstra, please add a tiny description of the change to the change log for 19.4c0, please.

[rouge8]

This affects one test that is now formatted differently, and it also causes a fair number of changes in our codebase. Now that I look at it, in most cases I actually feel like the old formatting is better, so maybe we need to do something different than just fix the type mismatch.

I had the same experience in one of our codebases where the old formatting was generally better. Is there an issue or plans to address this?

[JelleZijlstra]

Filed #781 to follow up on this.