How can I grep text from file?

@ngelot · 06-12-2007, 09:22 AM

I would like to grep the following text from a file called file.txt:

Code:

<p class="maintxt">
- You should include as much detail as possible in your message,
including exact error messages (where applicable) and what
you have done so far. The more detail you include the more we 
can help.</p>

If I do:

Code:

cat file.txt | grep "<p class=\"maintxt\">" > file2.txt

I only get one line with <p class="maintxt"> in file2.txt

How can I grep so that I get all the text from <p class="maintxt"> to </p>?

Thanks!

_______________________
@ngelot

nx5000 · 06-12-2007, 09:43 AM

Maybe this:

Code:

sed -n -e '/<p class="maintxt">/,/<\/p>/p'

pwc101 · 06-12-2007, 09:51 AM

Or

Code:

grep -A4 "<p class=\"maintxt\">" file.txt > file2.txt

Assuming that text you want is always going to be exactly 4 lines after the match.

edit: @ngelot, UUOC (uselss use of cat) - grep will take a file name as an input argument, therefore no need to cat the file first.

@ngelot · 06-13-2007, 04:10 AM

Thanks!

I found out that awk does the job:

Code:

awk '/<p class=\"maintxt\">/, /<\/p>/' file.txt > file2.txt

Liked the UUOC - thats the essens of computing : do things easier!

_______________________
@ngelot

syg00 · 06-13-2007, 04:33 AM

Quote:

Originally Posted by UUOC

If it's only one file, concatenating it with nothing at all is a waste of time

Nope.
Not if you want to read it and "cat" it to STDOUT ...

pwc101 · 06-13-2007, 04:38 AM

Quote:

Originally Posted by syg00

Nope.
Not if you want to read it and "cat" it to STDOUT ...

Point taken, but there are also tools specifically for that (less and more to name but two). Such is the *nix way!

edit: On occasion, I've caught myself doing something like this:

Code:

cat file.txt | less

d'oh!

tfault · 06-13-2007, 04:44 AM

What about this way?

Code:

grep "<p class=\"maintxt\">.*</p>" file.txt > file2.txt

(Note: I have not tested this, only theory!)

Also, don't pipe cat to grep. Just grep directly from the file!