Question

Using sed I’m having a problem trying to match and delete blog entries from a txt file before creating a sitemap.xml

# Contents of filename:
# http://www.example.com/2008/10/article3.html
# http://www.example.com/2009/11/article7.html

#!/bin/bash
hostName="www.example.com"
hostTLD="$(echo ${hostName}|cut -d . -f 3)" # results in "com"

sed -i '/\.'"${hostTLD}"'\/\([0-9]{4}\)\/\([0-9]{2}\)/d' filename

I can’t figure out how to match the year/month bits. I want to remove all lines that contain “.TLD/year/month/”

I know the $hostTLD part works because I’m using it with a different match:

sed -i '/\.'"${hostTLD}"'\/category\//d' filename # works! ".TLD/category/"

Answer 1

You were close, but you needed to use double-quotes around your sed command and escape the braces. Try this instead:

sed -i "/\.$hostTLD\/[0-9]\{4\}\/[0-9]\{2\}/d" filename

For your second command, use this:

sed -i "/\.$hostTLD\/category\//d" filename