Using BASH sed command to strip a line - bash

Question

I have one line in a html file which i located using
grep -m 1 'argument'
That line looks a lot like this
<tag option="something" option="something"><span option="something"> Text1 </span> - <span option="something"> Text2 </span></tag>
I need to extract Text 1 and Text 2 using seperate lines, what do I do? I get that I need to use sed, I have removed tag and span at the begginng leaving me with
Text1 </span> - <span...........</tag>
but I need only Text1 and i realy dont know how to remove that non-static Text2

Answer 1

If the lines are exactly always looking like the example you provide, you can do it with a regexp.
But in all other cases, you should really use a XML parser instead (for example, use perl : twig, or others)
So here is a regexp, but you've been warned ^^
#replace each <...> with "|", so you can easily separate each fields
sed -e 's/<[^>]*>/|/g'
You can then fetch each section by using that new "simple" separator, |:
grep 'argument' | sed -e 's/<[^>]*>/|/g' | awk -F'|' '{print $3}' #shows Text1. Change $3 to $5 to fetch the Text2

Answer 2

Here is much shorter way to do that using grep and perl regular expressions.
$ cat testfile # I've placed your line in this file
<tag option="something" option="something"><span option="something"> Text1 </span> - <span option="something"> Text2 </span></tag>
$ grep -Po '<span[^>]*>\K[^<]*' testfile
Text1
Text2
But if you want to get exactly Text1 and Text2 you need a bit more complicated regexp. And here it is:
$ grep -Po '<span[^>]*>( )?\s*\K.*?(?=\s*( )?</span>)' testfile
Text1
Text2
Some explanation:
This regex is using lookaround syntax or zero-width assertions. You can read about it here
\K might be unfamiliar too. It is very similar to zero-width assertions and is greatly explained here. Here is a quote from that link:
There is a special form of this construct, called \K , which causes the regex engine to "keep" everything it had matched prior to the \K and not include it in $& . This effectively provides variable-length look-behind. The use of \K inside of another look-around assertion is allowed, but the behaviour is currently not well defined.
Ok, but why?
One of the greatest things in this approach is that you're already using grep, you can possibly combine this regex with your search pattern so you will only need one grep command, unlike Oliver`s answer which uses grep, sed and awk.
But still, there are special tools to parse xml, please use them instead of this regex porn.