[PLUGIN] bzr changeset, time to pull again

[Robey Pointer]

On 2 Jul 2005, at 1:19, John A Meinel wrote:

> The current format uses tabs quite liberally for tokenizing. This is
> because 'path\twith\ttab'.encode('unicode_escape') will escape tabs,
> which means you can str.split('\t') on them, and not be ambiguous.
> But some people don't like them, because they don't look any different
> from spaces.
> If someone wants to offer something else, they are welcome to. But to
> the best of my knowledge, the only character that is not allowed in a
> unix *path* is a null (\0). And the only extra one not allowed in a
> filename is a '/'.
> So you're going to have to escape something. And it gets really  
> ugly if
> you have to escape a space character. (Though again, python handles
> u'\u0020' just fine :) But I don't think we'd like to see:
>
> path\u0020to\u0020my\u0020file.txt
> file-id:path\u0020to\u0020my\u0020file.txt-1092309182390-109820381234

If you normalize the path, then "//" would be an unambiguous  
terminator (and you could follow with spaces for readability):

moon landing.txt//  file-id:moon landing.txt-12389127-123871223784

I think this looks ugly, but decided to bring it up for  
completeness.  I agree that the tabs are kinda gross but have no  
better solution.

robey