Sed: template through a multiple line

Question

I have the following text:

[...]
<p class="title">ABC</p>
<p class="text">
<a href="https://site" target="_blank">
TEXT HERE   </a>
</p>
[...]

[...]
<p class="title">ABC</p>
<p class="text">
TEXT HERE  </p>
[...]

from the given text is necessary to get:

TEXT HERE<no space>
TEXT HERE<no space>

If the text was in one line, i.e.

<p class="title">ABC</p><p class="text"><a href="https://site" target="_blank">TEXT HERE   </a></p>
<p class="title">ABC</p><p class="text">TEXT HERE </p>

I would solve this problem in the following way: sed -n "s/.*title\">ABC<\/p>.*\">\([^<]*\).*/\1/p" ./file.txt

But I have a pattern that goes through a multiple line and I don't know how to solve the task in this case. Can somebody give the right direction for solving the problems?

Answer 1

This might work for you (GNU sed):

sed -nE '/"title">ABC<\/p>/{:a;s/<\/p>/&/2;tb;N;ba;:b;s/\n//g;s/ABC//;s/<[^>]*>//g;s/\s*$//;p}' file

Focus on line(s) with "title">ABC</p> and then keep appending lines (or not) until a second </p> is found.

Remove newlines if present.

Remove the text ABC.

Remove all tags.

Remove any trailing white space and print the result.