awk: get lines starting from String1 and finish String2

frenchn00b · 10-02-2007, 05:32 PM

Hi, I'd like an easy way for getting lines starting from String1 and finish String2 with awk

Code:

cat myfile | awk '  {

bla ... 

 split($0,vk,string1) ; print vk[2] 

... bla 
 } '

I can but it ll be quite complicate... I am sure there is easy way since awk is very powerful

jschiwal · 10-02-2007, 06:15 PM

If you want to include the entire lines in the range, simply precede the command with /<pattern1>/,/<pattern2>/.

example:

Code:

awk '/gdm/,/lp/{ print $0 }' /etc/passwd
gdm:x:105:108:Gnome Display Manager daemon:/var/lib/gdm:/bin/false
haldaemon:x:101:102:User for haldaemon:/var/run/hal:/bin/false
icecream:x:102:104:Icecream Daemon:/var/cache/icecream:/bin/false
lp:x:4:7:Printing daemon:/var/spool/lpd:/bin/bash

You can also use sed:

Code:

sed -n '/gdm/,/lp/p' /etc/passwd

ghostdog74 · 10-02-2007, 10:31 PM

shorter version of awk

Code:

awk '/pattern1/,/pattern2/' file

frenchn00b · 10-04-2007, 01:05 AM

Thanks
actually , I should have explained more
I do:

Code:

cat index.htm | grep http:// | awk '  blabla'

and Id like that teh awk do return

Code:

http://linklinklink

:
from

Code:

blablaaref="http://linklinklink"blablabal

so that I can have an output:

Code:

http://linklinklink
http://linklinklink
http://linklinklink
http://linklinklink
..

radoulov · 10-04-2007, 03:22 AM

Post a sample from your file.

angrybanana · 10-04-2007, 09:12 AM

You need to give us an actual sample. But from what I'm understanding you want something like this.

Code:

$ cat file
blablaaref="http://linklinklink"blablabal

$ sed -n 's/.*blablaaref="\(.*\)"blablabal.*/\1/p' file
http://linklinklink

sed -n 's/.*pattern1\(.*\)pattern2.*/\1/p' file

this won't work if you have multiple urls on one line..
so yea, post a sample...

frenchn00b · 10-05-2007, 07:35 AM

Quote:

Originally Posted by angrybanana

You need to give us an actual sample. But from what I'm understanding you want something like this.

Code:

$ cat file
blablaaref="http://linklinklink"blablabal

$ sed -n 's/.*blablaaref="\(.*\)"blablabal.*/\1/p' file
http://linklinklink

sed -n 's/.*pattern1\(.*\)pattern2.*/\1/p' file

this won't work if you have multiple urls on one line..
so yea, post a sample...

Code:

$ sed -n 's/.*blablaaref="\(.*\)"blablabal.*/\1/p' file
http://linklinklink

Exactly !!

By chance would you have that with awk, I am awk fan now. It's so great tool, C way and vb way ...

thanks !!

ghostdog74 · 10-05-2007, 07:58 AM

Code:

# awk -F '"' '{print $2}' file
http://linklinklink

frenchn00b · 10-28-2007, 04:56 PM

If I do this, I do not get all the files to download, then I ll get to look for each nunber to download separately ...lot of work

and this : "[","]"

Code:

 elinks mysite.htm  | grep  ZIP | awk '  "[","]"         {     print $0      }  '

is not working

sorry, I am learning
but awk is super cool !: http://www.grymoire.com/Unix/Awk.html

angrybanana · 10-29-2007, 08:16 AM

Not sure what you want exactly. Can you give an example of the input data and what you want for an output.