John Skopis wrote on Mon, 11 May 2009 at 10:54 -0500:
>
>
> >-----Original Message-----
> >From: Paul Burba [mailto:ptburba_at_gmail.com]
> >Sent: Monday, May 11, 2009 10:49 AM
> >To: Daniel Shahaf
> >Cc: B. Smith-Mannschott; David Glasser; John Skopis;
> >dev_at_subversion.tigris.org
> >Subject: Re: [PATCH] svnadmin load will not import dump with windows
> >newline character in svn:mergeinfo
> >
> >On Mon, May 11, 2009 at 10:51 AM, Daniel Shahaf <d.s_at_daniel.shahaf.name>
> >wrote:
> >> What Ben said.
> >>
> >> Also, I think the patch here is orthogonal to issue #3404 et al, since
> >> John said something about a segfault. John, could you say where the
> >> segfault is? Can you reproduce it without using svnadmin (by direct
> >> API usage)?
> >>
> >> B. Smith-Mannschott wrote on Mon, 11 May 2009 at 07:31 +0200:
> >>> > On May 8, 2009 3:23 PM, "John Skopis"
> ><jskopis_at_backstopsolutions.com> wrote:
> >>> >
> >>> > Hello,
> >>> >
> >>> > svnadmin load fails while importing a dump that contains a windows
> >>> > newline in svn:mergeinfo prop. I have not done extensive testing on
> >>> > this patch, but it should work (in that it doesn't segfault when I
> >>> > attempt to import a revision with \r\n in mergeinfo). Be advised I
> >am not actually a developer.
> >
> >John,
> >
> >
> >To follow-up on what Daniel asked: How exactly did svnadmin load fail?
> > Did you actually get a segfault or did you get an error something like
> >this?
> >
> > 1.6.2>svnadmin load ..\repositories\t1 < 131.dump
> >
> > <SNIP successful txn>
> >
> > ------- Committed revision 7 >>>
> >
> > <<< Started new transaction, based on original revision 8
> > * editing path : A_COPY ...svnadmin: Dumpstream data appears to
> >be malformed
> >
> >Or something else entirely?
> >
> >Paul
>
> Paul, here's what it looks like:
> ------- Committed revision 40997 >>>
>
> <<< Started new transaction, based on original revision 40998
> ' found in revision listter '
> # echo $?
> 1
>
> This made it extra hard to find where it actually failed as the \r in the printf made it nearly impossible to grep for the error printf string. =]
>
>
>
From the po files:
#: ../libsvn_subr/mergeinfo.c:408
#, c-format
msgid "Invalid character '%c' found in revision list"
which means that svn_mergeinfo_parse was called.
(But *why* was it called?)
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=2193292
Received on 2009-05-11 18:38:45 CEST