David Holland <dholland-sourcechanges%netbsd.org@localhost> writes: > Well, it treats "this Sunday" as 6 days away. It apparently thinks > "this X" is in the next seven days and "next X" is the week past that. > > That is not entirely crazy, though it's not really consistent with the > conventional usage I know... but that usage isn't well defined. On > Monday, "this Sunday" and "next Sunday" are pretty clearly the same > day; but on Saturday, they probably aren't, and it's not clear what > the crossover is. That's a good point. I agree that in the US, on Saturday "next Sunday" is probably in 8 days. But Friday is iffy. Which means people say "Sunday the 11th", or fail to communicate. Which leads me to think that "next Sunday" should probably not be accepted by parsedate at all. > Meanwhile, these usages are regional. I've never understood the "next > Sunday week" or "this Sunday week" business no matter how many times I > run across it in lit from across the pond. "Sunday week" (no this or next) means the 2nd Sunday in the future, vs "Sunday" which is 1st Sunday in the future. I think it's less ambiguous than the US usage of "next". But it is probably worse than that. > Anyhow, if you care, I suggest writing a spec for parsedate }:-) Fair enough. My own strategy will be to avoid relying on it...
Attachment:
signature.asc
Description: PGP signature