Filter RPM URLs with awk

Luke Hebert

- 2 minutes read - 223 words

When you are digging through a long log file or a package manifest and only need the RPM download URLs, a single awk command saves you from scrolling, copying, and pasting. The trick is to match lines that contain both a keyword (like autoinstall) and the .rpm extension, then print just the last field, which is typically the URL or file path.

The Command

awk '/autoinstall/ && /\.rpm/ {print $NF}'

How It Works

  • /autoinstall/ matches any line containing the string autoinstall.
  • && requires both patterns to be true on the same line.
  • /\.rpm/ matches any line containing the literal string .rpm (the backslash escapes the dot so it is not treated as a regex wildcard).
  • {print $NF} prints the last whitespace-delimited field on the matching line, which is usually the URL or file path.

Example

Given a log line like:

2022-01-24 10:32:01 INFO autoinstall package http://mirror.example.com/repo/epel/x86_64/epel-release-8-11.el8.noarch.rpm

The command outputs:

http://mirror.example.com/repo/epel/x86_64/epel-release-8-11.el8.noarch.rpm

Tips

  • Swap autoinstall for whatever keyword identifies the lines you care about.
  • If your URLs are not always the last field, replace $NF with the specific field number (e.g., $5).
  • Pipe the output to xargs or curl to download the packages in one step:
awk '/autoinstall/ && /\.rpm/ {print $NF}' install.log | xargs -n1 curl -O

This keeps your terminal clean and your package list accurate without manual copy-paste.