I need to extract a URL that is wrapped with <strong> tags. It's a simple regular expression, but I don't know how to do that in shell script. Here is example:
line="<strong>http://www.example.com/index.php</strong>"
url=$(echo $line | sed -n '/strong>(http:\/\/.+)<\/strong/p')
I need "http://www.example.com/index.php" in the $url variable.
Using busybox.
This might work:
url=$(echo $line | sed -r 's/<strong>([^<]+)<\/strong>/\1/')
url=$(echo $line | sed -n 's!<strong>\(http://[^<]*\)</strong>!\1!p')
You don't have to escape forward slashes with backslashes. Only backslashes need to be escaped in regular expressions. You should also use non-greedy matching with the ?-operator to avoid getting more than you want when there are multiple strong tags in the HTML sourcecode.
strong>(http://.+?)</strong
Update: as busybox uses ash, the solution assuming bash features likely won't work. Something only a little longer but still POSIX-compliant will work:
url=${line#<strong>} # $line minus the initial "<strong>"
url=${url%</strong>} # Remove the trailing "</strong>"
If you are using bash (or another shell with similar features), you can combine extended pattern matching with parameter substitution. (I don't know what features busybox supports.)
# Turn on extended pattern support
shopt -s extglob
# ?(\/) matches an optional forward slash; like /? in a regex
# Expand $line, but remove all occurrances of <strong> or </strong>
# from the expansion
url=${line//<?(\/)strong>}
