NoQ added inline comments.
================
Comment at: clang/include/clang/Analysis/AnyCall.h:170
+ return {ME->getArgs(), ME->getNumArgs()};
+ } else if (const auto *CE = dyn_cast<CallExpr>(E)) {
+ return {CE->getArgs(), CE->getNumArgs()};
----------------
xazax.hun wrote:
> As far as I remember `CXXMemberCallExpr` and `CXXOperatorCallExpr` having
> slightly different behavior with the former not including the implicit object
> argument in the argument list contrary to the latter. If that is the case,
> maybe this API should provide some uniformity, e.g. skipping the implicit
> object argument for `CXXOperatorCallExpr` (and potentially providing a
> separate function for getting the this/self). What do you think?
Excellent point yeah. And if we're going to do something so non-trivial it
probably makes sense for me to start covering `AnyCall` with unittests.
Repository:
rC Clang
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D101788/new/
https://reviews.llvm.org/D101788
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
