Extract lines with specific patterns in separate files

Question

The input file has the following lines and segregate these using the 2nd field '+' symbol lines in one file and '-' symbol lines in another file:

24 +  I am the Five man    
22 -  Who are you?  The new number two!    
51 +  . . . And four on the floor    
42 +    
16 -  Who is number one?    
33 -  I three you.

is it possible when the $2 is '+', a=$1+500 and b=$1-500 at the same time when the $2 is '-', a=$1-500 and b=$1+500? 'a' and 'b' are the new variables.

is it possible to segregate above lines in two files basing on 2nd field, which is having '+' and '-' symbol using awk... — doc, Jun 23 '13 at 09:04
is it possible when the $2 is '+', a=$1+500 and b=$1-500 at the same time when the $2 is '-', a=$1-500 and b=$1+500? 'a' and 'b' are the new variables... — doc, Jun 24 '13 at 12:47

iruvar · Answer 1 · 2013-06-23T11:22:31.530

5

Another option

awk 'BEGIN{m["+"]="plus.txt";m["-"]="minus.txt"} $2 ~ /^[+-]$/{print>>m[$2]}'

edited Jun 23 '13 at 11:22

answered Jun 23 '13 at 10:34

iruvar

22,736
7
53
82

perreal · Answer 2 · 2013-06-25T00:05:30.580

5

With Perl:

perl -lne '/^\d+ -/ && print(STDERR) || print' input 2> minus > plus

in a slightly different form:

perl -lpe 'select(/^\d+ -/?STDERR:STDOUT)' input 2> minus > plus

Also possible using tee:

tee >(sed -n '/^[0-9]* -/p' > minus) < input | \
   sed -n '/^[0-9]* +/p' > plus

edited Jun 25 '13 at 00:05

answered Jun 23 '13 at 11:08

perreal

94,503
21
155
181

score 4 · Answer 3 · answered Jun 23 '13 at 09:33

4

This solution filters the output to files f1 and f2.

awk '{ if ($2 == "+") print >>"f1"; else if ($2=="-") print >>"f2"; }' datafile

answered Jun 23 '13 at 09:33

suspectus

16,548
8
49
57

score 4 · Accepted Answer · answered Jun 23 '13 at 21:58

4

This will put the "+" lines in file1 and the others in file2:

awk '{print > ("file" ($2~/+/?1:2))}' file

answered Jun 23 '13 at 21:58

Ed Morton

188,023
17
78
185

1

Hello Ed Morton, now I need to take the 1st field which are having '+' sign and assign to new variables by adding 500 and subtracting 500 of each record. – doc Jun 24 '13 at 13:21
What do you mean by "assign to new variables"? Just update your question with your new requirements along with sample input and expected output or open a new question. – Ed Morton Jun 24 '13 at 15:00

score 2 · Answer 5 · answered Jun 23 '13 at 11:03

2

Code for GNU sed:

sed '/\S\+\s\++/!D' file > plus.txt
sed '/\S\+\s\++/D' file > minus.txt

answered Jun 23 '13 at 11:03

captcha

3,756
12
21

score 2 · Answer 6 · answered Jun 23 '13 at 13:48

2

With awk you would simply do:

awk '$2=="+"{print>"f1";next}{print>"f2"}' file

Demo:

$ cat file
24 +  I am the Five man
22 -  Who are you?  The new number two!
51 +  . . . And four on the floor
42 +
16 -  Who is number one?
33 -  I three you.

$ awk '$2=="+"{print>"f1";next}{print>"f2"}' file

$ cat f1
24 +  I am the Five man
51 +  . . . And four on the floor
42 +

$ cat f2
22 -  Who are you?  The new number two!
16 -  Who is number one?
33 -  I three you.

answered Jun 23 '13 at 13:48

Chris Seymour

83,387
30
160
202

Hello sudo_O, now I need to take the 1st field which are having '+' sign and assign to new variables by adding 500 and subtracting 500 of each record. – doc Jun 24 '13 at 13:19

score -1 · Answer 7 · answered Jun 23 '13 at 09:13

-1

The question's title is a bit unclear, I'll edit it in a moment. Meanwhile here's the answer:

awk '/^[0-9]+ \+/{print > "a"} /^[0-9]+ -/{print > "b"}'

answered Jun 23 '13 at 09:13

Michał Kosmulski

9,855
1
32
51

1

You mean `'$2=="+" {print > "a"} $2=="-" {print > "b"}'` – SheetJS Jun 23 '13 at 09:15
@Nirk OK, I thought the first number was just a line number inserted by the editor and not part of the input file. Fixing the expression to handle that. – Michał Kosmulski Jun 23 '13 at 09:17

Extract lines with specific patterns in separate files

7 Answers7