Mailing-List: contact dev-help@cocoon.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@cocoon.apache.org
Received-SPF: neutral (hermes.apache.org: local policy)
Message-ID: <4188AFDF.2050100@apache.org>
Date: Wed, 03 Nov 2004 11:15:59 +0100
From: Sylvain Wallez <sylvain@apache.org>
Organization: Anyware Technologies
User-Agent: Mozilla Thunderbird 0.8 (Macintosh/20040913)
MIME-Version: 1.0
To: dev@cocoon.apache.org
Subject: Re: [lazyweb] SVN checkout without keyword expansion?
References: <41889352.5000408@apache.org>
 <Pine.LNX.4.61.0411031016510.9358@lapgp.otego.com>
In-Reply-To: <Pine.LNX.4.61.0411031016510.9358@lapgp.otego.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit

Giacomo Pati wrote:

> On Wed, 3 Nov 2004, Sylvain Wallez wrote:
>
>> Hi all,
>>
>> A lazyweb question about SVN: is there a way to perform a checkout 
>> without expanding the $Id$ keyword? I couldn't find one by googling 
>> around, and that would be so useful for merging branches where many 
>> files only differ by their revision number.
>>
>> Currently I used a quickly hacked script that does the job, but that 
>> would be much cleaner to have it handled directly by svn.
>
>
> I usually compare different branches by use of the following gnu diff 
> command in a script:
>
> diff -rubBd -x CVS -x target -x .svn -x .cvsignore -x .settings \
>         -x .project \
>         --ignore-matching-lines=\\\$Author: \
>         --ignore-matching-lines=\\\$Date: \
>         --ignore-matching-lines=\\\$Header: \
>         --ignore-matching-lines=\\\$Id: \
>         --ignore-matching-lines=\\\$Locker: \
>         --ignore-matching-lines=\\\$Name: \
>         --ignore-matching-lines=\\\$RCSfile: \
>         --ignore-matching-lines=\\\$Revision: \
>         --ignore-matching-lines=\\\$Source: \
>         --ignore-matching-lines=\\\$State: \
>         --ignore-matching-lines=\\\$Version: $* $TARGET $SOURCE
>
> It is suitable for CVS as well as for SVN. Hope this helps.


Eeek! Thanks for the tip, but I prefer more graphical front-ends to 
reconcile changes :-)

I use Eclipse's diff utility by selecting the same folder in the two 
branches then "compare with/each other" in the contextual menu.

The nice thing about SVN is that it doesn't consider as having changed a 
file where "$Id: blah blah$" has been changed to "$Id$". Here's the 
script I use (named "rmkw" for "remove keywords")

#!/bin/sh
temp=rmkw.tmp
find . \( -name '*.java' -or -name '*.xml' -or -name '*.xmap' -or -name 
'*.js' -or -name '*.xsl' \) -print | while read file
do
  echo $file
  /usr/bin/sed -e 's/\$Id:.*\$/$Id$/' $file > $temp
  mv $temp $file
done

Sylvain

-- 
Sylvain Wallez                                  Anyware Technologies
http://www.apache.org/~sylvain           http://www.anyware-tech.com
{ XML, Java, Cocoon, OpenSource }*{ Training, Consulting, Projects }