--include vs. --exclude

[Phil Howard]

On Sun, Mar 28, 2004 at 07:16:26PM -0600, John Van Essen wrote:

| On Fri, 26 Mar 2004, Phil Howard <phil-rsync-2 at ipal.net> wrote:
| >
| > So I have on my server a big file tree.  I want to use rsync to download
| > only the PDF files, which make up a small portion of that tree.  So I try
| > it this way:
| > 
| >   rsync -aHPvz --include '*.pdf' --exclude '**' user at host:<source> <destination>
| > 
| > which gives me nothing.
| 
| Hi Phil,
| 
| Your goal seems to be to create a tree that consists only of *.pdf
| files and their necessary directory nodes.  Right?

Right.


| The closest you can get to that with standard includes/excludes is
| with this combination:
| 
|   --include '*/'     (process/create all directory nodes)
|   --include '*.pdf'  (include all *.pdf files)
|   --exclude '*'      (exclude everything else)
| 
| The disadvantage is that you will get the entire tree, even the
| branches with no pdf files.

I was not including that first --include '*/' for the directories.
I was under the impression rsync simply created directories "as needed".


| rsync 2.6.0 has a new 'files-from' option which will do what you need,
| but you have to create the file list yourself (easy enough to do with
| the 'find' command).  But you are at a disadvantage in that you are
| doing a pull and may not have access to the source.
| 
| Here's a possibility.  Do a --dry-run first to determine the names
| of the pdf files.  Grep the -v output of rsync for '\.pdf$', store
| that list in a file, then do a real run with the files-from option.
| 
| I didn't try this - so I haven't verified that this will work.

Maybe a feature like "--make-needed-directories" would do well.  Of course
for directories that exist at the source, using the same metadata should
still be done.

-- 
-----------------------------------------------------------------------------
| Phil Howard KA9WGN       | http://linuxhomepage.com/      http://ham.org/ |
| (first name) at ipal.net | http://phil.ipal.org/   http://ka9wgn.ham.org/ |
-----------------------------------------------------------------------------