Enable single-line case statements in LLVM clang-format style?

[arichardson]

@jh7370

This example got me thinking. Could clang-format be persuaded to do the following style?

switch (Lex.getKind()) {
default:                        CC = CallingConv::C;             return false;
case lltok::kw_ccc:             CC = CallingConv::C;             break;
case lltok::kw_fastcc:          CC = CallingConv::Fast;          break;
case lltok::kw_coldcc:          CC = CallingConv::Cold;          break;
case lltok::kw_cfguard_checkcc: CC = CallingConv::CFGuard_Check; break;
}

This doesn’t look so bad, because it is much more regular. I’d also agree that if you end up with some cases that don’t fit on one line, then none of them should in the above style.

I realize this is an ancient thread, but I ran into this issue as well recently while changing some larger switch statements and had to just ignore the clang-format warnings on the pull request.

Since Clang17, it appears there is a AlignConsecutiveShortCaseStatements setting which would generate the style listed in the initial post.

I would be in favour of updating the coding style to the following:

AllowShortCaseLabelsOnASingleLine: true
AlignConsecutiveShortCaseStatements:
  Enabled: true
  AcrossEmptyLines: true
  AcrossComments: true
  AlignCaseColons: false

It seems the single-line case statements are still very common in LLVM:

❯ git grep -E '^\scase\s+.+((return.;)|(break;))$' lib | wc -l
6733

I tried reformatting all of llvm/lib with the clang format config above and we now get almost 4x as many single line cases (rought estimate is that this is about 20% of all case statements).

❯ git grep -E '^\s*case\s+.+((return.*;)|(break;))$' lib | wc -l
24676

Counter-point against changing: this data means right now we are only using single-line case statements 25% of the time, so maybe too much time has passed and we should just stick to the style clang-format has mandated for the past years?