On Sat, 2005-07-30 at 12:49 +0200, Matthieu Moy wrote: > # By default, no regexp. > $ baz rbrowse Matthieu.Moy/ > Cannot connect to archive Matthieu.Moy.
Nice. What about glob by default?
> # Exact match on archive
> $ baz rbrowse [EMAIL PROTECTED]/ | head
> [EMAIL PROTECTED]
> authinfo--main--0
> base-0 ... patch-2
[...]
Cool.
> # Regexp matching. The regexp must match the full string.
> $ baz rbrowse --id-regex [EMAIL PROTECTED]/
Why not just call it "--regex"?
> # You can explicitely ask for substring with ".*"
> $ baz rbrowse --id-regex '[EMAIL PROTECTED]/' | head
> [EMAIL PROTECTED]
> authinfo--main--0
> base-0 ... patch-2
Glob?
> # LIMIT must match a prefix. Not necessarily the full string.
> $ baz rbrowse --id-regex '[EMAIL PROTECTED]/bazaar' | head
>
> [EMAIL PROTECTED]
> bazaar--a-test-for-file-history--1.5
> base-0 ... patch-2
[...]
> # You can ask for full string with "$"
> $ baz rbrowse --id-regex '[EMAIL PROTECTED]/bazaar$' | head
>
I would prefer a full match, as that would be more consistent with the
syntax that could be used in arguments of comands that can take several
branch names or globs/regexps.
> # exact match also matches prefix.
> $ baz rbrowse '[EMAIL PROTECTED]/bazaar' | head
>
> [EMAIL PROTECTED]
That would not be consistent with the syntax used in commands that can
accept several commands. I regard "browse" essentially as a "ls" in the
branch directory.
> # ... but not arbitrary substring
> $ baz rbrowse '[EMAIL PROTECTED]/zaar' | head
Cool.
--
-- ddaa
signature.asc
Description: This is a digitally signed message part
_______________________________________________ Gnu-arch-users mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnu-arch-users GNU arch home page: http://savannah.gnu.org/projects/gnu-arch/
