I agree, no need to risk breaking things by trying to be clever.

If I can be of more help when it comes time to implement the change, I'd
love to contribute.  I haven't worked on any open source projects before,
but now is as good a time as any to start.

Dan

On Mon, Nov 8, 2010 at 12:54 PM, Jeff Forcier <[email protected]> wrote:

> On Mon, Nov 8, 2010 at 10:34 AM, Dan Craig <[email protected]> wrote:
> > In my
> > use case, it's still awkward to do the replace(r"\/", "/"), but I may be
> > unusual in passing a fully escaped string.
>
> That would be my guess -- I'd bet most people using those functions
> don't have regexes in the actual strings being commented/uncommented.
> Though it's certainly understandable given how many config, etc files
> have regexes in them.
>
> > Perhaps the built-in escaping in
> > sed() should only escape non-escaped slashes?  I'm not sure that wouldn't
> > have some other side effect, but it might work.
>
> Escaping/quoting in general has been a big pain with Fabric, since we
> wrap so many shell related things. The deeper you go the more oddball
> corner cases you find.
>
> In this particular case I think I'd probably just opt for applying the
> escaping uniformly and expecting the user to do any additional
> escaping, instead of risking further potential craziness. At least,
> until an overall re-examination of escaping, which is planned at some
> point.
>
> At any rate, I've documented this in ticket #250:
>
>    http://code.fabfile.org/issues/show/250
>
> Best,
> Jeff
>
> --
> Jeff Forcier
> Unix sysadmin; Python/Ruby engineer
> http://bitprophet.org
>
_______________________________________________
Fab-user mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/fab-user

Reply via email to