[PATCH 0 of 8] Persistent tag caching

Greg Ward greg-hg at gerg.ca
Sun Jul 12 20:02:03 CDT 2009

OK, this is it: a working tag cache that actually does not smell bad
and seems to pass the whole test suite.  (Running it now on Ubuntu
9.04, and I'll go run it on OS X and NetBSD soon.)

This is basically the code I posted a few days ago, with some minor
cleanups and comment clarifications.  So please review the hell out of
it.  Tear it to shreds.  Be merciless.  I'd rather find problems now
than after it goes out with 1.4.  ;-)

Oh yeah: it turns out that some of the refactoring I did is not
strictly necessary.  Specifically, splitting readtags() in two and
changing one of the halves from a closure to a method wasn't actually
needed, since I don't try to cache .hgtags content.  However, it's a
nice refactoring and I'm rather fond of it.  So I'll leave the patches
in; if either one really bugs you, say so and I can drop one or both
of them. (I think splitting it in two is a readability win, so I'd
like to keep that.  Changing the closure to a method is not important,
and I could drop that change easily.)


More information about the Mercurial-devel mailing list