[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: " Inconsistent line-endings"

From: Philip Martin <philip_at_codematters.co.uk>
Date: 2004-04-08 00:31:07 CEST

Julian Foad <julianfoad@btopenworld.com> writes:

> Ben Reser wrote:
>> In the process of looking for this I went looking for the error text.
>> I couldn't find it. So I went and found it on the 1.0.x branch and
>> then
>> found the code location on trunk. The error text was removed in
>> r8908. We were outputing this in 1.0.x:
>> $ svn ci -m 'test2'
>> Sending foo
>> Transmitting file data .svn: Commit failed (details follow):
>> svn: Inconsistent line-endings in source stream
>> Now we output:
>> $ ~/wc/svn-trunk/subversion/clients/cmdline/svn ci -m 'test2'
>> Sending foo
>> Transmitting file data .subversion/libsvn_client/commit.c:747:
>> (apr_err=135000)
>> svn: Commit failed (details follow):
>> subversion/libsvn_subr/subst.c:454: (apr_err=135000)
>> Note the complete and utter lack of details that follow.
>> I think the changes to subversion/libsvn_subr/time.c and
>> subversion/libsvn_subr/subst.c in r8908 should be reverted. Opinions?
> Only if we can't make it work the way r8908 was intended to work,
> which is that the generic error message corresponding to
> SVN_ERR_IO_INCONSISTENT_EOL should be printed without having to put
> a copy of the message text in the source code at each point of call.

The problem appears to be caused by using svn_error_quick_wrap in
libsvn_client/commit.c:reconcile_errors. Given the new error handling
that avoids printing the whole error chain, I think reconcile_errors
needs to add a generic "commit failed" error rather than calling
quick_wrap. I suspect every current use of quick_wrap needs to be

Philip Martin
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Apr 8 00:31:25 2004

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.