I suppose we'll do pgindent soon.
I have semi-manually gone through the various .l and .y files and fixed
up the formatting of the C code to be more in line with pgindent style.
Most of that code was old, so I don't expect this to be such a large
effort going forward. I also think a lot of that code started as
copy-and-paste, so having "correct" style will also help future code use
better style by default.
The patch is rather large, so I won't post it here, but you can see it here:
https://github.com/petere/postgresql/compare/master...petere:indent-gram-scan.patch
I propose to apply this after the main pgindent run is done.