reg.expr: how to detect the whole line?

Hi!


i have this sort of header on every page:

<p class="calibre1"> BAD WRITING AND HOWTO FIX IT</p>

So i need to detect a line where all the characters are capital letters and there's always more than one word. I also need to make sure that the script wouldn't delete similar strings inside of the text which is NOT header.

Is there someone who could help me with it please? So far i've figured out how to detect strings of capital characters: [A-Z]{2,} and tried to use () to repeat the pattern search but just got all confused...

:thanks: