tags:

views:

57

answers:

2

Hi I need to run 'sed' command on file1.txt and have it extract all names that are in between StringA and StringB... ex: "Nickname":"bad_name", extract bad_name and then save all results too Output.txt. is this even possible or should I be looking at another command?

+2  A: 

without showing more sample of your file1.txt, i am assuming you have consistent data format. If it is, then use awk

echo '"Nickname":"bad_name"' | awk -F":" '{print $NF}'

For a file, just input the file name

awk -F":" '{print $NF}' file > output.txt

Otherwise, provide more sample data for us to work with.

ghostdog74
A: 

Using sed:

dmedvinsky@home:~$ cat file1 
"Nickname":"bad_name"
"First Name":"Dmitry"
bad line
"Dog Name":"Chuppy"
another bad line

dmedvinsky@home:~$ cat file1 | sed -e '/\(.\+\)":"\(.\+\)"/!d' -e 's/"\(.\+\)":"\(.\+\)"/\2/'
bad_name
Dmitry
Chuppy

The first expression deletes badly formatted lines, the second one replaces the pattern with 2nd matched group -- the contents of second quotes.

d.m
the file has other user information in it so i have to isolate "Nickname":" (with the quotations) and extract bad_name with the ending string of ",ex: "Nickname":"bad_name",output: bad_namethe : can't be used by itself to determine where the name starts only because its not unique in the other data sets us the : as well
acctman
Well, in that case you need to change the first part of the expression (first subpattern) to `"Nickname"`.So the command will look like:`cat file1 | sed -e '/"Nickname":"\\(.\+\\)"/!d' -e 's/"Nickname":"\\(.\+\\)"/\1/'`
d.m
I tried the command you list and it just displayed the file content. the sed command didn't run or output anything. cat file1.txt | sed -e '/"Nickname":"\(.\+\)"/!d' -e 's/"Nickname":"\(.\+\)"/\1/'
acctman
Yep, my bad. Try this:`cat file1 | sed -e '/"Nickname":"\\([^"]\+\\)"/!d' -e 's/.*"Nickname":"\\([^"]\+\\)".*/\1/'`To output to file just use ` > output`.
d.m