topic Re: Print from matched string on.... in Operating System - Linux

Print from matched string on....

dev44 — Thu, 24 Jan 2008 13:56:28 GMT

Hi,

How can I print or get the output from a certain string on...like everything after that string printed out to the screen?

Thanks,
Sally

Re: Print from matched string on....

James R. Ferguson — Thu, 24 Jan 2008 14:07:27 GMT

Hi Sally:

# cat stuff
this is a file
of stuff
which I want to
dump from here
to the end.
Hence, here it is...
from there to the end.

# sed -ne '/here/,$p' stuff
dump from here
to the end.
Hence, here it is...
from there to the end.

Regards!

...JRF...

Re: Print from matched string on....

Pete Randall — Thu, 24 Jan 2008 14:11:49 GMT

Sally,

For questions like this I always turn to "Handy One-Liners for Sed" (attached) which I found here on the Forums years ago.

Pete

Re: Print from matched string on....

Hein van den Heuvel — Thu, 24 Jan 2008 14:13:48 GMT

Try this:

awk '/certain string/,0' file.txt

Hein.

Re: Print from matched string on....

dev44 — Thu, 24 Jan 2008 14:16:51 GMT

Thanks guys!

Re: Print from matched string on....

James R. Ferguson — Thu, 24 Jan 2008 14:23:02 GMT

Hi Sally:

If you mean literally "everything AFTER that string" but not anything that preceeds it on the same line, I'd do it thusly:

# perl -nle 'if ($n) {print} elsif (m/(here.*)/) {print $1;$n++}'

It this example, the string to match is "here". As written it could be a part of the string "where". Thus, if you wanted only the isolated word "here" (at a boundary) , change this to:

# perl -nle 'if ($n) {print} elsif (m/(\bhere\b.*)/) {print $1;$n++}'

Now use this variation and compare the difference of the two:

# cat stuff
this is a file
of stuff
where I want to
dump from here to
the end.
Hence, here it is...
from there to the end.

Regards!

...JRF...

Re: Print from matched string on....

Hein van den Heuvel — Thu, 24 Jan 2008 14:53:20 GMT

This topic is closed, problem solved. Excellent!
But I had an extra thought...

'Normally' (is there such thing?) the awk range is used as: /start/,/end/

In my example I used a hardcoded 0 as end test, so the end match is never true (near). It ends when the data ends.

Sometimes I want to see a few lines after a match. I just realized that using a similar technique one could use:

awk '/error/,!++x%5' file.log

This is only a quick and dirty solution.
One would probably want a seperator between matching zones.
And what to do on re-match while in the print zone?
If you solve all that, then it quickly becomes a full blown script.
Cute?

Hein.