[SCM] Samba Shared Repository - branch master updated - 1d9c88b3885728aba3d7fef85d80501125011f1c

Andrew Bartlett abartlet at samba.org
Wed Nov 19 07:57:19 GMT 2008


On Wed, 2008-11-19 at 08:50 +0100, Stefan (metze) Metzmacher wrote:
> Andrew Bartlett schrieb:
> > On Sun, 2008-11-16 at 09:38 -0600, Stefan Metzmacher wrote:
> >> commit 69af236e7772e85b7b8dc7045c6737d85f4b189a
> >> Author: Stefan Metzmacher <metze at samba.org>
> >> Date:   Thu Nov 13 21:27:10 2008 +0100
> >>
> >>     librpc/ndr: let GUID_from_data_blob() accept a ndr blob or a hex
> >> string too
> > 
> > I had some (not committed) code which relied on this not working (strict
> > input validation on incoming LDIF).  Can you point me out what do you
> > use this for?  I'm just trying to decide if we should split this into
> > two functions or if I should just add an extra check on the ldif parsing
> > side?
> 
> For me GUID_from_data_blob() means convert the binary representation
> (ndr) into a GUID, that's why I added the ndr blob code. And as we
> already autodetect the string representations, I just added the one
> missing representation, the hex string one.
> 
> Why is it invalid to specify an guid as an base64 encoded ndr blob
> in LDIF?

I agree it isn't.  I decided later (but didn't get the gumption up to
writing) that being liberal in what you accept really should be the
philosophy here. 

Andrew Bartlett

-- 
Andrew Bartlett
http://samba.org/~abartlet/
Authentication Developer, Samba Team           http://samba.org
Samba Developer, Red Hat Inc.



More information about the samba-technical mailing list