You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Carlo Kok <ck...@carlo-kok.com> on 2006/01/26 19:37:37 UTC
svn assertion
I have an older checkout done by rapidsvn that has entries with urls
like:
url="svn://buurtnet/projects/trunk/" ;
svn 1.3 seems to give runtime assertions when I execute:
svn status -Nu
in such a directory.
Assertion failed: is_canonical (base, blen), file
C:\Home\brane\src\svn\releases\subversion-1.3.0\subversion\libsvn_subr\path.c,
line 114
Having to fix this in each and every directory is quite a hassle, would
it be possible for SVN to just accept this and strip of the ending slash?
--
Carlo Kok
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn assertion
Posted by "Peter N. Lundblad" <pe...@famlundblad.se>.
On Sun, 29 Jan 2006 kfogel@collab.net wrote:
> Carlo Kok <ck...@carlo-kok.com> writes:
> > I have an older checkout done by rapidsvn that has entries with urls
> > like:
> >
> > url="svn://buurtnet/projects/trunk/" ;
> >
> > svn 1.3 seems to give runtime assertions when I execute:
> >
> > svn status -Nu
> >
> > in such a directory.
> >
> > Assertion failed: is_canonical (base, blen), file
> > C:\Home\brane\src\svn\releases\subversion-1.3.0\subversion\libsvn_subr\path.c,
> > line 114
> >
> > Having to fix this in each and every directory is quite a hassle, would
> > it be possible for SVN to just accept this and strip of the ending slash?
>
> I'm a little leery of having SVN silently clean up bad data in the
> working copy administrative area. Not that this data is fatally bad,
> but it's not something SVN would have written, and presumably modern
> versions of RapidSVN wouldn't write it either. The working copy
> metadata is pretty sacred stuff; better to be strict than loose about
> it, that way we can detect it when tools start to stray.
>
OTOH, assert isn't the best input validator either:-) We need to validate
the entry when we read it from disk. (And RapidSVN needs to learn to not
throw invalid paths at our APIs.)
Thanks,
//Peter
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn assertion
Posted by Carlo Kok <ck...@carlo-kok.com>.
kfogel@collab.net wrote:
> Carlo Kok <ck...@carlo-kok.com> writes:
>> I have an older checkout done by rapidsvn that has entries with urls
>> like:
>>
>> url="svn://buurtnet/projects/trunk/" ;
>>
>> svn 1.3 seems to give runtime assertions when I execute:
>>
>> svn status -Nu
>>
>> in such a directory.
>>
>> Assertion failed: is_canonical (base, blen), file
>> C:\Home\brane\src\svn\releases\subversion-1.3.0\subversion\libsvn_subr\path.c,
>> line 114
>>
>> Having to fix this in each and every directory is quite a hassle, would
>> it be possible for SVN to just accept this and strip of the ending slash?
>
> I'm a little leery of having SVN silently clean up bad data in the
> working copy administrative area. Not that this data is fatally bad,
> but it's not something SVN would have written, and presumably modern
> versions of RapidSVN wouldn't write it either. The working copy
> metadata is pretty sacred stuff; better to be strict than loose about
> it, that way we can detect it when tools start to stray.
Actually it's RapidSVN 0.9 which uses SVN 1.2.3 that I managed to check
this out with (I used svn://buurtnet/projects/trunk/ as an url, not
knowing that that could cause harm). That's the newest RapidSVN
available and isn't that old. Although I agree it's easy to fix myself,
I just wanted to report this.
>
> Fixing it by hand shouldn't be such a big hassle. (By the way, that
> semicolon isn't really there, right?)
Nope it's not really there.
--
Carlo Kok
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: svn assertion
Posted by kf...@collab.net.
Carlo Kok <ck...@carlo-kok.com> writes:
> I have an older checkout done by rapidsvn that has entries with urls
> like:
>
> url="svn://buurtnet/projects/trunk/" ;
>
> svn 1.3 seems to give runtime assertions when I execute:
>
> svn status -Nu
>
> in such a directory.
>
> Assertion failed: is_canonical (base, blen), file
> C:\Home\brane\src\svn\releases\subversion-1.3.0\subversion\libsvn_subr\path.c,
> line 114
>
> Having to fix this in each and every directory is quite a hassle, would
> it be possible for SVN to just accept this and strip of the ending slash?
I'm a little leery of having SVN silently clean up bad data in the
working copy administrative area. Not that this data is fatally bad,
but it's not something SVN would have written, and presumably modern
versions of RapidSVN wouldn't write it either. The working copy
metadata is pretty sacred stuff; better to be strict than loose about
it, that way we can detect it when tools start to stray.
Fixing it by hand shouldn't be such a big hassle. (By the way, that
semicolon isn't really there, right?)
-Karl
--
www.collab.net <> CollabNet | Distributed Development On Demand
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org