Revision spec issues

[Matthew D. Fuller]

There are a couple difficulties currently with revision specs; mostly,
inconsistencies between what's said to work, and what does work.


1) 0.7 fixed the issue of ranges with no upper bound (e.g, "5.."), but
   neither 0.7 not bzr.dev handle ranges with no lower bound ("..5"),
   which "should" work according to the wiki and comments in the code.
   I'm not sure it really matters if we don't support it;
   "1..<whatever>" is virtually as easy to type (as opposed to
   "<whatever>..-1", which is a little less intuitive), but the docs
   and the code really should line up either way.


2) Date specs by name (today, etc) don't act like they're documented
   to.  Both the docstrings in the code and the wiki [I just changed
   the latter] say that "date:today..date:tomorrow" should refer to
   "all revisions today", but trying to actually DO that fails:

   % bzr log -rdate:today..date:tomorrow
   bzr: ERROR: Branch BzrBranch(u'/tmp/bzr/dtest/') has no revision
        date:tomorrow

   "date:yesterday..date:today", however, DOES appear to work
   correctly.  I don't have the knowledge of Python to tell whether it
   works correctly because that's the right way to do it, or whether
   it just SEEMS to work right, but is actually wrong.  I've updated
   the wiki so that it gives a command that at least mostly works,
   instead of one that dies, but somebody who understands this should
   take a gander at it and see which is right and beat the variations
   into submission.



-- 
Matthew Fuller     (MF4839)   |  fullermd at over-yonder.net
Systems/Network Administrator |  http://www.over-yonder.net/~fullermd/
           On the Internet, nobody can hear you scream.