Charusso added a comment. In D70805#1776527 <https://reviews.llvm.org/D70805#1776527>, @NoQ wrote:
> Usually this kind of code is hard to re-use because all use cases require > different definitions of "has descendant". We already have at least 3 such > definitions floating around and we can't re-use them. > > Even in the world of ASTMatchers the actual `hasDescendant` matcher is > basically an anti-pattern as you always want to use a matcher for a > //specific// parent-child relation instead (cf. > http://lists.llvm.org/pipermail/cfe-dev/2019-September/063260.html). When I encounter the following: unsigned Foo = Foo.Bar[Baz]->Qux; malloc(strlen(src) + Foo + 1); At the moment I cannot lookup the `strlen(src)` and nor the `+ 1`. There is no way to say the `strlen(src)` which parent is a `malloc()`. I do not think we could make the full support of pattern-matching on symbolic level, so at least I wanted to see if it is possible. I am even thinking of using the ASTImporter to normalize the expression so that in a flatten/serialized form the AST-matcher could look for such constructs without any wonky `DeclRefExpr` or `ReturnStmt` matching. I believe the latter is the way we could benefit a lot, and the former, the current patch needs to be eliminated. For now I cannot relax the AST or invent a symbolic-matcher language so I would prefer this simple workaround and may we will have better ideas later. CHANGES SINCE LAST ACTION https://reviews.llvm.org/D70805/new/ https://reviews.llvm.org/D70805 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
