[PATCH] D139834: [clang-format] AllowShortCompoundRequirementOnASingleLine

[MyDeveloperDay via Phabricator via cfe-commits]

MyDeveloperDay added inline comments.


================
Comment at: clang/lib/Format/Format.cpp:1265
   LLVMStyle.AllowShortCaseLabelsOnASingleLine = false;
+  LLVMStyle.AllowShortCompoundRequirementOnASingleLine = true;
   LLVMStyle.AllowShortEnumsOnASingleLine = true;
----------------
HazardyKnusperkeks wrote:
> MyDeveloperDay wrote:
> > Backl1ght wrote:
> > > MyDeveloperDay wrote:
> > > > why would the default be true, is that what happens today?
> > > yes
> > just to clarify so I'm sure (without me having to try it myself), if you hadn't introduce this option it would be the equivalent of true.
> > 
> > The only reason I say is we get complained at when we change the default from not doing something to doing something. Even if that means before we left it alone.
> > 
> > So mostly we normally find the options go through an evolution from   bool->enum->struct, sometimes it can be better to introduce an enum so we can have "Leave" as the default
> > 
> > in such circumstances you let the old behaviour be the default, that way we know. That previously unformatted compound statements won't be touch in any way. 
> > 
> > ```
> > else if (Style.AllowShortCompoundRequirementOnASingleLine != Leave  && .......)
> > {
> > ```
> > 
> > Users then have to "positively" buy into your style change one way or another. Rather than us imposing a default even if that default seems perfectly reasonable.
> The `true` as default is correct.
Then if @HazardyKnusperkeks  you think this is ok, I'm ok with it too.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D139834/new/

https://reviews.llvm.org/D139834