Neil-N4 wrote:

> I know I said I was deferring to other reviewers, but now I see parsing here 
> in the base patch. I'm not sure where the communication breakdown is, but 
> this is kind of a major issue that can't really be ignored, since we had 
> about a 30 minute call discussing what was to be done, and one of the key 
> items we discussed was splitting the base data structure into its own easy to 
> review PR and following up w/ parsing and other work in later PRs.
> 
> I'm going to try and ignore this PR for a bit to give you some time to 
> address this, but to be clear: this patch should _only_ contain the AST data 
> structures. It should test that their methods work correctly, w/o introducing 
> any kind of parsing into the mix. Any reviewer in LLVM would expect the 
> classes to have methods to access (and if appropriate to update) their 
> private data members, unless there is a clear benefit to not doing so, which 
> normally requires some documentation, discussion, or explanation. I don't 
> expect to see any complicated usage of BumpPtrAllocator for now, as I don't 
> see how that's relevant to the core behavior of the AST's design I've seen 
> thus far.

Thats my bad. I was keeping the parser in a stacked branch on my fork and when 
i merged review feedback changes back to this branch the parser commits came 
along with it. stripped it back to just node definitions, print/dump, and node 
tests that exercise the children. parser and ASTContext are in the stacked PR 
on my fork.

https://github.com/llvm/llvm-project/pull/205609
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to