Re: PGDN and CVS

________________________________

From: Gevik babakhani [mailto:gevik(at)xs4all(dot)nl]
Sent: 21 September 2005 10:41
To: pgsql-www(at)postgresql(dot)org
Cc: Dave Page
Subject: PGDN and CVS

Hi Dave,

Hi Gevik,

Regarding the PGDN.

1. Let's say we produce a directory with all kinds of
static html,images sample sql scripts etc... How do we want to update
the whole PGDN within our CVS.

We need a script that will traverse the directory structure and perform
the appropriate CVS add/remove operations based on what it finds, and
then commits the whole tree. It sounds hard at first thought, but I
don't think it is when thought about further - for example, we
hierarchically traverse the tree, and for each directory or file we
find, we check CVS/Entries to see if it exists in CVS. If not, we do a
CVS ADD on it. When we've processed the files, we then go line by line
through CVS/Entries, and CVS REMOVE any files or directories that don't
exist any more. That should be a relatively trivial perl script for
someone familiar with perl.

1. Do we want to put the static PGDN files in CVS anyway?
Should we keep the PGDN files separate not mixing with the current
site's CVS?

Yes, we do want them in CVS because we build the entire from there & the
backend database. we don't want to start pulling in content from
additional places as that just makes it much harder to rebuild the
servers if there is ever a problem.

Regards, Dave.