SVN Blame Returns Corrupt Data

2013-10-10 Thread T.J. Perovich
I'm having trouble running svn blame on a particular file.  It's returning
garbage.

In TortoiseBlame:
3341  TJP  ÿþO
3341  TJP

In the command line:
3341TJP  ■O
3341TJP


The file is 10.1k lines, not 2.  If I run the blame from revision 0 to 3341
it returns the correct information.

In WinMerge and TortoiseMerge, diffing the files shows about 10 lines
changing between 3340 and 3341 (it was merge).  However, the command line
diff shows the entire contents being changed with spaces between every
character. So "End Class" reads "E n d   C l a s s", etc..  Diffing a merge
post-rev# 3341 show the same spaces between every letter.

svn diff -r 3341:3489 svn://...

@@ -20032,7 +20058,7 @@

  F i l l _ d d l L o c a t i o n ( )
  F i l l _ d d l C o u n t r y ( )


Another strange thing is it's marking these as lines 20,032 and 20,058.
 But in Notepad++ they are lines 10,026 and 10,031.  The line numbers in
pre-rev# 3341 diffs match up between the Notepad++ and command line fine.

Any ideas on what happened or how to fix it?  I'm most concerned about
getting the blame working again.  Please let me know if you need more
information or if I was unclear with anything.  Thanks in advance!


Re: SVN Blame Returns Corrupt Data

2013-10-10 Thread T.J. Perovich
On Thu, Oct 10, 2013 at 2:27 PM, Ryan Schmidt <
subversion-20...@ryandesign.com> wrote:
>Sounds like you've converted the file from UTF-8 to UTF-16.

Thanks, you're absolutely right.  It changed from UTF-8 to UTF-16LE.

Any idea how to go about fixing it elegantly?  We have about 3 months of
commits since this happened.  Diff's in the GUIs have been working fine and
we don't blame too often which is why it was never noticed.


On Thu, Oct 10, 2013 at 3:15 PM, Bob Archer  wrote:

> > On Oct 10, 2013, at 11:29, T.J. Perovich  wrote:
> >
> > > I'm having trouble running svn blame on a particular file.  It's
> returning
> > garbage.
> > >
> > > In TortoiseBlame:
> > > 3341  TJP  ÿþO
> > > 3341  TJP
> > >
> > > In the command line:
> > > 3341TJP  ■O
> > > 3341TJP
> > >
> > >
> > > The file is 10.1k lines, not 2.  If I run the blame from revision 0 to
> 3341 it
> > returns the correct information.
> > >
> > > In WinMerge and TortoiseMerge, diffing the files shows about 10 lines
> > changing between 3340 and 3341 (it was merge).  However, the command
> > line diff shows the entire contents being changed with spaces between
> > every character. So "End Class" reads "E n d   C l a s s", etc..
>  Diffing a merge
> > post-rev# 3341 show the same spaces between every letter.
> > >
> > > svn diff -r 3341:3489 svn://...
> > >
> > > @@ -20032,7 +20058,7 @@
> > >
> > >   F i l l _ d d l L o c a t i o n ( )
> > >   F i l l _ d d l C o u n t r y ( )
> >
> > Sounds like you've converted the file from UTF-8 to UTF-16.
> >
> >
> > > Another strange thing is it's marking these as lines 20,032 and
> 20,058.  But in
> > Notepad++ they are lines 10,026 and 10,031.  The line numbers in pre-rev#
> > 3341 diffs match up between the Notepad++ and command line fine.
> >
> > Sounds like the line endings changed as well.
> >
>
> Sigh... if only svn would support Unicode encodings.
>
> BOb
>
>


Re: SVN Blame Returns Corrupt Data

2013-10-11 Thread T.J. Perovich
On Thu, Oct 10, 2013 at 5:49 PM, Bob Archer  wrote:

> I assume he was asking how to "fix" the blame. Cause, sure, he could open
> the file, convert it back to UTF-8 with CRLF line endings... and commit
> it... of course, now blame is going to show him on every line, since he
> just changed every line.
>
>
That's exactly what I meant.  You're correct with how the blame is handled.
 I committed the UTF-8 copy to a test branch, diff'd, and it showed every
line as being changed.  Unfortunately it looks like this is our best option.


On Thu, Oct 10, 2013 at 7:07 PM, Ben Reser  wrote:

> At current blame is not UTF-16 aware.
>
> About a year ago there was a "patch" (actually they just reposted an entire
> copy of blame.c) posted that helped, but it really didn't go anywhere
> since the
> original poster didn't continue following up.
>
> Perhaps you'll find the above useful.  Patches are of course welcome.
>

I'll take a look at that this weekend.  Thanks for all the input everyone.



On Thu, Oct 10, 2013 at 7:07 PM, Ben Reser  wrote:

> On 10/10/13 12:17 PM, T.J. Perovich wrote:
> > On Thu, Oct 10, 2013 at 2:27 PM, Ryan Schmidt <
> subversion-20...@ryandesign.com
> > <mailto:subversion-20...@ryandesign.com>> wrote:
> >>Sounds like you've converted the file from UTF-8 to UTF-16.
> >
> > Thanks, you're absolutely right.  It changed from UTF-8 to UTF-16LE.
> >
> > Any idea how to go about fixing it elegantly?  We have about 3 months of
> > commits since this happened.  Diff's in the GUIs have been working fine
> and we
> > don't blame too often which is why it was never noticed.
>
> At current blame is not UTF-16 aware.
>
> About a year ago there was a "patch" (actually they just reposted an entire
> copy of blame.c) posted that helped, but it really didn't go anywhere
> since the
> original poster didn't continue following up.
>
>
> https://mail-archives.apache.org/mod_mbox/subversion-dev/201207.mbox/%3CCAAF0CB13B282344AF95AD2DE3D1962215627A3C%40DAG-B.nexon.corp%3E
>
> and the followup
>
>
> https://mail-archives.apache.org/mod_mbox/subversion-dev/201208.mbox/%3ccab84ubvvrhffqyea5pf5gstmpxz+rh2jkvdvcqscocjv+rq...@mail.gmail.com%3E
>
> Perhaps you'll find the above useful.  Patches are of course welcome.
>


Re: Continuous Integration

2013-11-21 Thread T.J. Perovich
> I wish to create a web app and share it on SVN. I wish to use this SVN
url for Jenkins Integration.
>

I use Jenkins with SVN and it works just fine.  You'll need the SVN plugin
which is easily obtained through Jenkins.

> Do I need to take any precautions.

For SVN, make sure it's configured for https:// and not http, unless it's
internal.  For Jenkins, make sure you enable security.  By default it
allows for anonymous read/write.  You should also configure it for https
rather than http, unless it's internal.

Let me know if you have any specific questions.


On Tue, Nov 19, 2013 at 9:14 PM, Ryan Schmidt <
subversion-20...@ryandesign.com> wrote:

>
> On Nov 18, 2013, at 06:37, Navindian wrote:
>
> > I wish to create a web app and share it on SVN. I wish to use this SVN
> url for Jenkins Integration.
> >
> > Do I need to take any precautions.
>
> I don’t understand the question.
>
> I use Subversion and I’ve develop web sites whose code I’ve stored in a
> Subversion repository, but I haven’t used Jenkins. Perhaps you could
> explain what role you want Jenkins to play, and why you think it will be a
> problem.
>
>
>