================ ---------------- ChuanqiXu9 wrote:
> I think that needs to happen eventually as well Agreed. > similar to preambles supporting in-memory storage, and it's actually not that > hard, GenerateModuleInterfaceAction already has an overrideable > CreateOutputFile method. I am slightly confused. Do you say to move the IO to the memory instead of on-disk files? If true, we're close to that since we can write them to /tmp directories. I thought you're saying we need to get rid of the reading and writing process for module files. > Also this comment is not solely about virtiualizing CDB, but also about the > way we're setting up tests. I don't think we need any of the compile flags > you're explicitly setting in the tests. The whole purpose of current > implementation is to rely on new module paths we provide into the > header-search-options. So there isn't any value in spelling those out > explicitly, rendering test setup more complicated. Agreed. In the new added test, I meant to use `TestTU` with `AdditionalFiles` to mimic a multiple module units project, but it fails. It looks like it can only accept headers. I feel there are a lot of spaces to improve. But I think it may be fine to leave it as is and improve this later. My plan after this patch is to make the module files reusable. I think the functionality to support modules in clangd is at least usable after that. https://github.com/llvm/llvm-project/pull/66462 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
