use awk to detect regex patterns and print by line?

Question

I have a text file in this format [ONE testing 1 2 3] [TWO lorem ipsum] [ONE 123]

I want to print `[ONE.+]` line by line.

An example output would be

[ONE testing 1 2 3]
[ONE 123]

I've tried awk '/\\[ONE.+\\]/ { print $1 }' but it didn't work. Can anyone teach me why? And what the correct way is?

Answer 1

awk works line by line, so the expression is only matched once per line. To do it in awk, you can use the match function in a loop. You'd also have to modify your regex to be less greedy, since your expression doesn't magically stop at the first ].

It might be easier to just use grep:

echo  "[ONE testing 1 2 3] [TWO lorem ipsum] [ONE 123]" | grep -o '\[ONE[^]]*\]'

Answer 2

You can try something like this

sed -re 's/(\[ONE[^\[]*\])/\n\1\n/g' temp.txt

Input

[ONE testing 1 2 3] [TWO lorem ipsum] [ONE 123]

Output

[ONE testing 1 2 3]
 [TWO lorem ipsum] 
[ONE 123]

---

If you want to remove the column with TWO then

sed -re 's/(\[ONE[^\[]*\])()/\n\1\n/g; s/(\[[^ONE][^\[]*\])//g' temp.txt

Output

[ONE testing 1 2 3]

[ONE 123]

Answer 3

If this is part of something bigger:

BEGIN { 
# Change the field-separator, from default blank, to the end-marker 
# for each "field"
    FS = "] "
}
# Get rid of lines which can't possibly match
!/\[ONE/ { next
    }
{
# Test and report each of three fields for starting with [ONE,
# "closing" the field with FS, except for the last which will 
# already be "closed"
if ( $1 ~ /^\[ONE/ ) {
    print $1 FS
    }
if ( $2 ~ /^\[ONE/ ) {
    print $2 FS
    }
if ( $3 ~ /^\[ONE/ ) {
    print $3
    }
}

The "if"s can be replaced by one in a loop if you are so inclined, but watch for the final one, as the FS (field separator) is not needed (unless you have a trailing blank in your data).

Answer 4

"awk" by default takes as a 'single space' as the separator and the 'print $1' command tries to retrieve the first value separated by the default separator.

Try this out :

Let there is a text file named as 'test.txt' containing the three lines.

cat test.txt

[ONE testing 1 2 3]

[TWO lorem ipsum]

[ONE 123]

grep -h '[ONE*' test.txt

[ONE testing 1 2 3]

[ONE 123]