On Tue, Apr 03, 2007 at 11:33:54PM +0100, James Youngman wrote: > When adding an item to struct predicate, please use a comment to > indicate which predicates it supports.
Done. > <attr/attributes.h> maynot be available on all platforms, so configure > should check for it. Yeah, that's a TODO for me. > fnmatch_sanitycheck(); -- this only needs to be called once, ever. Alright, but currently each predicate supporting globbing seems to be responsible of calling it. Maybe think of a way of centralizing that decision? > + error (0, 0, "missing attribute name (format: ATTR=VALUE)") ; > + state.exit_status = 1 return (false) ; > > Works, but this is easier & better: error (1, 0, "missing attribute > name (format: ATTR=VALUE)"); Thanks a bunch! > I see that you are expecting -xattr NAME=FOO and have plans for glob > support. This sounds good so far. When designing this though, please > think about how to support this sort of functionality on a variety of > systems (that is, not just on one filesystem or OS). Subject treated in other message. > Oh, and one piece of bad news: you have a potential name clash. Some > systems may appear to support xattr_get() just because we accidentally > linked in the curses library. Not sure how to protect ourselves > against that one. Well, let's avoid that name then. > Thanks for your work so far! Why certainly, hacking find's well-structured code is a breeze compared to other projects! :) Are you responsible for this, or did you find it when you got to findutils? Leslie -- NEW homepage: https://viridian.dnsalias.net/~sky/homepage/ gpg --keyserver pgp.mit.edu --recv-keys DD4EBF83
pgpYBaUwkdLjk.pgp
Description: PGP signature
_______________________________________________ Bug-findutils mailing list Bug-findutils@gnu.org http://lists.gnu.org/mailman/listinfo/bug-findutils
