Print lines between two patterns to new file

Question

please help me with this. I have file of following pattern

ABC x  
bla bla bla  
bla bla  
bla  
XYZ  
ABC y   
bla bla bla  
bla bla  
bla bla bla  
XYZ  

ABC z  
bla bla bla
XYZ

I need output in file x.txt

ABC x  
bla bla bla  
bla bla  
bla    
XYZ

and

ABC y   
bla bla bla  
bla bla  
bla bla bla  
XYZ

in y.txt and so on for rest of patterns

You should try and ask specific questions, please check FAQ – Bruce Oct 21 '13 at 07:28 — Bruce, Oct 21 '13 at 07:28

Jotne · Answer 1 · 2013-10-21T08:13:06.250

2

Some like this:

awk '/ABC/ {close(f".txt");f=$2} {print >f".txt"}' file

This test if line starts with ABCand then set output file name to value in $2

Eks

cat y.txt
ABC y
bla bla bla
bla bla
bla bla bla
XYZ

EDIT: added close() to awk to close open file so it does not run out of space if it creates many files

edited Oct 21 '13 at 08:13

answered Oct 21 '13 at 07:33

Jotne

40,548
12
51
55

To avoid printing empty lines into these files, you can use `$0{print >f".txt}`, so that if `$0` is empty, the condition won't be matched. – fedorqui Oct 21 '13 at 08:25
That may be good and may be bad to do. If there are space between start end and end pattern, they are lost too. – Jotne Oct 21 '13 at 08:40

score 0 · Answer 2 · answered Oct 21 '13 at 07:29

0

Well, put all Input into one Array, walk through that with Regex until you find ABC something, open something.txt as Output, continue through your Array printing the lines to Output until you find XYZ and move on.

I assume you know how to write Perl.

answered Oct 21 '13 at 07:29

DeVadder

1,404
10
18

score 0 · Answer 3 · answered Oct 21 '13 at 07:48

How about:

my $fh;
open my $fh_in, '<', "input_file" or die "unable to open 'input_file: $!";
while (<$fh_in>) {
    chomp;
    if (/^ABC \w+/../XYZ/) {
        if (/^ABC (\w+)/) {
            open $fh, '>', "$1.txt" or die "unable to open '$1.txt': $!";
        }
        print $fh $_,"\n";
    }
}

score 0 · Answer 4 · answered Oct 21 '13 at 07:51

0

You can do it using sed as shown below:

sed -n '/ABC x/,/XYZ/p' file > x.txt

If you know all the patterns, you can use a loop:

for i in x y z
do
   sed -n "/ABC $i/,/XYZ/p" file > "$i".txt
done

answered Oct 21 '13 at 07:51

dogbane

266,786
75
396
414

score 0 · Answer 5 · answered Oct 21 '13 at 11:28

0

awk '/ABC/{name="file "$2".txt";}{print >name}' your_file

answered Oct 21 '13 at 11:28

Vijay

65,327
90
227
319

Print lines between two patterns to new file

5 Answers5

Linked