================
Comment at: docs/ClangFormatStyleOptions.rst:422-424
@@ -421,1 +421,5 @@
+**ObjCSpaceAfterSynchronized** (``bool``)
+ Add a space after ``@synchronized`` in Objective-C, i.e. use
+ ``\@synchronized (lock)`` instead of ``@synchronized(lock)``.
+
----------------
strager wrote:
> djasper wrote:
> > Don't write these by hand. They should be auto-generated with
> > cfe/docs/tools/dump_format_style.py.
> Good to know. I assume that applies for my other diffs as well.
Yes.
================
Comment at: lib/Format/FormatToken.h:287
@@ -286,1 +286,3 @@
}
+ bool is(clang::tok::ObjCKeywordKind Kind) const {
+ return Tok.getObjCKeywordID() == Kind;
----------------
strager wrote:
> djasper wrote:
> > remove clang::
> >
> > Also, do you mind cleaning up the other usage of getObjCKeywordID() ==.
> > There shouldn't be that many, so this can be done in the same patch.
> Sure. I'll do it in a separate patch, along with similar types like
> preprocessor keywords.
AFAICT, there are only three other places in clang-format that could make use
of this. Please do that in this patch. Otherwise, we'll just forget.
================
Comment at: lib/Format/TokenAnnotator.cpp:1850-1851
@@ -1848,2 +1849,4 @@
(!Left.Previous || Left.Previous->isNot(tok::period))))) ||
+ (Style.ObjCSpaceAfterSynchronized &&
+ Left.is(tok::objc_synchronized)) ||
(Style.SpaceBeforeParens == FormatStyle::SBPO_Always &&
----------------
strager wrote:
> djasper wrote:
> > Can you swap the operand order? I would find that easier to read.
> You mean:
>
> ````
> (Left.is(tok::objc_synchronized) &&
> Style.ObjCSpaceAfterSynchronized)
> ````
>
> ?
>
> I was keeping consistent with lines 1852 and 1844, where the `Style` check
> comes first.
It is what I meant, but you are right, keeping consistent with the rest of this
(awful ;-) ) expression is probably better.
My thinking is that I like to read it better the other way round, but having
the style flag first can actually bring a performance improvement for the other
parts of this expression. So just leave as is.
http://reviews.llvm.org/D10371
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
_______________________________________________
cfe-commits mailing list
cfe-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
