We are using svn for development of a large web application, and we do periodic updates to production. The production server does not have access to svn (for security reasons).
What is the best way to push the changes since the last production release for a new release? We would like to avoid re-creating the whole site each time, since it is very large.
Well let me give a shot. You can parse the SVN update command output and generate copy instructions that only copy the files that were changed?
http://svnbook.red-bean.com/en/1.1/ch03s05.html
Or if you prefer a semi-manual solution you can use a diff tool like WinMerge or Araxis Merge to sync both directories.
EDIT:
I don’t think “svn update” will work exactly, but what I settled on was:
I wrote a script to capture the output of the “switch”. It translates U and A operations to “adds” to a zip archive. It translates D operations to delete commands to be executed on the production server. On the production server, we just need to unzip the archive and run the delete commands.