On 10/10/07 1:28 PM, "Ryan Schmidt" <subversion-2007b@ryandesign.com> wrote:
>
> On Oct 10, 2007, at 13:07, Reuben Avery wrote:
>
>> i have a website in a subversion repository. there are literally
>> hundreds and hundreds of orphan files and all kinds of other junk
>> that are being identified to be deleted.
>
> What do you mean, "orphan files"? What do you mean, "being identified
> to be deleted"? How are they being identified, and by whom or by what?
meaning I am working with a site tree that has many files and directories
which have been identified by human beings as no longer needed and have been
deleted from a shared directory.
>> 1) is there a way to create a "final" copy of this site sans all
>> the .svn directories but then be able to have subversion crawl
>> through the directory structure and create the .svn directories for
>> each directory that exists in the same structure within the
>> repository?
>
> I don't understand. I'm pretty sure the answer is "no"; the only way
> you get the .svn directories at all is by running "svn checkout" to
> make a new working copy from the repository. Or by running "svn
> update", if the repository is delivering a new directory to your
> existing working copy.
>
> What exactly are you trying to do?
I am trying to synchronize the repository directory tree with the physical
tree we now have post all the cleanup efforts. Before this process started,
I pulled down the entire site tree and checked it into the repository.
What I need to do is recursively walk the entire tree and for files that are
no longer there but within the repository, perform an svn delete, and for
directories which are no longer there but within the repository, do the
same.
I can use a software tool to compare the two directories (the SVN working
copy and this physical non-SVN site tree) and remove those files which are
in one but not the other (ignoring .svn directories, naturally).. So then,
if I could run some sort of script which could compare this slimmed down WC
to the repository and perform the necessary actions so that the repository
will reflect this new structure, I will have success.
>> and then
>> 2) perform an svn delete against the repository for each file and
>> directory which is in the repository but not in this directory
>> structure?
>
> I really don't understand, but you can write a script to run "svn
> delete" or any other operation you want to....
but this is so complicated and tricky and my scripting skills are rather
limited, I'm ashamed to say. Surely someone has written something like
this? For "webmasters", particularly those dealing with client websites
which are complete messes that we've been called in to rectify, this seems
like a fairly common sort of task.
Thanks again,
Reuben
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Wed Oct 10 22:57:04 2007