1

I'm trying to combine two options for creating a tar archive, but they don't work well together. I don't know what's the problem.

I use:

tar -C /var/www/website html -cvzf html.tgz -T /tmp/htmllist.txt

If I use only the -C option it creates a tar archive with only a relative path "./html/", but ALL the files inside that folder. I don't want this.

Then if I use only the -T option it creates a tar archive with only the required files, but the path is absolute "/var/www/website/html/whatever.html". I don't want this either.

I want a combination of the two, but somehow it doesn't work.

The folder structure is:

/var/www/website/html/
  • Lots of .html files

The /tmp/htmllist.txt folder contains lines like:

  • /var/www/website/html/file1.html
  • /var/www/website/html/file5.html
  • /var/www/website/html/file32.html

And so on.

How do I solve this problem?

Improve this question
1
  • 1
    If you cd /var/www/website and use relative file paths in /tmp/htmllist.txt then tar will use relative paths in the archive. You can also use find and pipe the output to tar ... -T -.
    – AFH
    Jun 4, 2015 at 12:36

1 Answer 1

Reset to default
2

Tar works the other way around when using -T. It archives the files including full path. If you don't want the path, strip it on extracting, for example,

Create tarball

tar -C /var/www/website html -cvzf html.tgz -T /tmp/htmllist.txt

Extract tarball

tar -xvf html.tgz --strip-path=3

This strips three components from the output path, i.e. /var/www/website.

Improve this answer
4
  • Thanks for your answer! Well, this tarball is generated for the web users to download so I don't expect them to use strip path or any complicated thing. Plus I don't want them to read the full path. Is there anyother solution?
    – DMCGIVEN
    Jun 4, 2015 at 12:14
  • Or maybe I can compress, extract, and recompress again ? This is unefficient though....
    – DMCGIVEN
    Jun 4, 2015 at 12:18
  • The non-obvious solution would be to change into the source directory and do something like tar -cvzf foo.tgz *.html
    – bjanssen
    Jun 4, 2015 at 12:28
  • It's not *.html It's a specific selection of certain .html files
    – DMCGIVEN
    Jun 4, 2015 at 12:55

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .