RE: svn commit: r1705391 - in /subversion/trunk: ./ subversion/include/ subversion/include/private/ subversion/libsvn_client/ subversion/libsvn_diff/ subversion/libsvn_fs_x/ subversion/tests/cmdline/ subversion/tests/libsvn_diff/

> -----Original Message-----
> From: Bert Huijben [mailto:bert_at_qqmail.nl]
> Sent: zaterdag 26 september 2015 10:06
> To: dev_at_subversion.apache.org; commits_at_subversion.apache.org
> Subject: RE: svn commit: r1705391 - in /subversion/trunk: ./
> subversion/include/ subversion/include/private/ subversion/libsvn_client/
> subversion/libsvn_diff/ subversion/libsvn_fs_x/ subversion/tests/cmdline/
> subversion/tests/libsvn_diff/
>
> A few comments inline.

> "===============================================================
> > ====\n"
> > + "diff --git a/iota b/iota\n"
> > + "old mode 100755\n"
> > + "new mode 100644\n"
> > + "Property changes on: iota\n"
> > + "-------------------------------------------------------------------\n"
> > + "Added: svn:executable\n"
> > + "## -0,0 +1 ##\n"
> > + "+*\n"
> > + )
>
> This specifies the addition of the svn:executable property with the non
> canonical "*\n" value. Is this really what you want to test?
> I think the actual property set code will change it to the proper value, but I
> would say you need the " \ No newline at end of property" marker here.
>
> The generated patchfile doesn't need it, but it does need the boolean to note
> the same thing.

The generated patch file might still need it to record a failed patch in the reject file.

In that case the hunk is copied there and it should have the 'has no eol' marker.
 
>
> Bert
Received on 2015-09-26 13:02:31 CEST