Huge thanks for opening this PR and for the work you put into it.

The maintainer shop is currently full, so this may sit for a bit before it gets a proper review. We will come back to this as soon as possible with real feedback; I wanted to make sure it did not sit unacknowledged in the meantime.

Thanks for improving Swift extraction coverage. Before review can continue, please add a DCO Signed-off-by trailer to the non-merge commit and link the language coverage issue if there is one.

Thank you for the Swift extraction work, and apologies that this has been sitting.

The CI failure here is a small one and it is genuinely in your diff — clang-format violations on your own lines in internal/cbm/lang_specs.c, at lines 558 and 560. That is all that is red; nothing else about the change is failing.

One thing that will save you a frustrating round trip: our CI lint uses the Homebrew LLVM clang-format specifically. A standalone clang-format-20 produces spurious whole-file drift against our config, so if you format with that you will get a diff that looks nothing like the two lines actually at fault.

The reliable local check is:

make -f Makefile.cbm lint-ci

which runs the same cppcheck + clang-format combination CI does, so a green result there means a green result on the runner.

Once that is clean I will give the extraction change itself a proper review — Swift type coverage is squarely the kind of graph-quality improvement we want, so it is worth finishing.