Delete consecutive lines if patterns match in each line

Question

Good day,

I was wondering how to delete to consecutive lines when this pattern is found: /^pt [a-z]\npy [0-9][0-9][0-9][0-9]\n//

Expected Input

pt a
py 01234
pt b
cd abc
py 5678

Expected Output

pt b
cd abc
py 5678

Thank so much in advance for any clue.

score 3 · Accepted Answer · answered Nov 01 '14 at 02:41

3

This should work:

sed '/^pt [a-z]/{N;/py [0-9][0-9][0-9][0-9]/d}' your_file

Expanation

If the current line matches /^pt [a-z]/ we execute what's between braces.
N appends the next line to the active buffer.
If the active buffer now matches /py [0-9][0-9][0-9][0-9]/, we delete (i.e. we don't print) the contents of the active buffer. This is accomplished by d.

answered Nov 01 '14 at 02:41

Joseph R.

Thank so much for helping me out, it works smoothly. Have a nice day : ) – Another.Chemist Nov 01 '14 at 02:48
You can substitute [0-9][0-9][0-9][0-9] by [0-9]\{4\} – Costas Nov 01 '14 at 08:26

score 0 · Answer 2 · answered Nov 01 '14 at 05:23

0

Through Perl,

$ perl -00pe 's/^pt [a-z]\npy [0-9]{5}\n//' file
pt b
cd abc
py 5678

answered Nov 01 '14 at 05:23

Avinash Raj

2 Answers2