Re: difference between ~ & ^ negation

[Bart Schaefer <schaefer@brasslantern.com>]

On Jan 3,  2:55am, Mikael Magnusson wrote:
}
} I think of ^foo as a * that won't match "foo", so you can have
} foo/bar^baz and it would match bar* but not barbaz, and if you think
} of it like that it's not really that confusing. ^foo/^bar is then just
} */* that won't match */bar or foo/*.

Unfortunately that's only helpful for understanding globbing, because in
pattern matching, ^foo/^bar is like * that won't match any foo/* except
specifically foo/bar.  As Peter said, it's implicitly ^(foo/^(bar)) in a
pattern; whereas it's (^(foo))/(^(bar)) in a glob.

There's no other "wildcard" with that weird difference in precedence in
the two cases.  The others differ only on whether "/" is an ordinary
character.  This is what fooled me into thinking it didn't work at all
in a pattern match; I assumed ^foo/^bar was two consecutive expressions,
(^foo/)(^bar), so I constructed the wrong test strings.