jimingham wrote: > A high-level concern that I have with this is the grammar is full of > C++-specific nodes, which is not something I would like to see in living in > `Core`. Have we previously reached any consensus on how we want to handle > different languages in DIL? I would be more comfortable with moving most of > this into the `Plugins/Language/CPlusPlus` directory and having only a > generic, language-agnostic driver live in `Core`.
My .02: The problem here is that the commands that use the DIL often have no idea what language the path expression they are being asked to parse is going to end up being. `(lldb) target var some_DIL_expression ` could find C/ObjC/C++/Rust/Swift variables that match that expression. The idea for the DIL is to have enough syntax to express sensible data lookup and presentation that would cover the variety of languages we support. That's going to end up looking like the languages we are using it to model, but it shouldn't be driven by them. I think that's better than having the DIL have language specific variants and the user somehow has to know to pick the right one for the context. https://github.com/llvm/llvm-project/pull/95738 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
