sed - comment a matching line and x lines after it

Question

I need help with using sed to comment a matching lines and 4 lines which follows it. in a text file.

my text file is like this:

[myprocess-a]
property1=1
property2=2
property3=3
property4=4

[anotherprocess-b]
property1=gffgg
property3=gjdl
property2=red
property4=djfjf

[myprocess-b]
property1=1
property4=4
property2=2
property3=3

I want to prefix # to all the lines having text '[myprocess' and 4 lines that follows it expected output:

#[myprocess-a]
#property1=1
#property2=2
#property3=3
#property4=4

[anotherprocess-b]
property1=gffgg
property3=gjdl
property2=red
property4=djfjf

#[myprocess-b]
#property1=1
#property4=4
#property2=2
#property3=3

Greatly appreciate your help on this.

score 24 · Accepted Answer · answered Jul 28 '12 at 19:32

24

You can do this by applying a regular expression to a set of lines:

sed -e '/myprocess/,+4 s/^/#/'

This matches lines with 'myprocess' and the 4 lines after them. For those 4 lines it then inserts a '#' at the beginning of the line.

(I think this might be a GNU extension - it's not in any of the "sed one liner" cheatsheets I know)

answered Jul 28 '12 at 19:32

Flexo

87,323
22
191
272

`sed: -e expression #1, char 12: unexpected ','` :-( – Mausy5043 May 13 '17 at 14:07
What version of see? Odds are it's either old or not got the gnu extensions. – Flexo May 13 '17 at 19:28
sed 4.2.2. It's the Debian default so, yeah, probably missing GNU extensions – Mausy5043 May 14 '17 at 08:03
gnu documentation for this: https://www.gnu.org/software/sed/manual/html_node/Range-Addresses.html – wisbucky Jul 04 '19 at 08:22
How does one do it for all the lines after a matching pattern? – Coder Apr 13 '22 at 21:24
Off the top of my head I think changing +4 to $ would do that. – Flexo Apr 14 '22 at 06:07

perreal · Answer 2 · 2012-07-29T04:35:27.530

3

sed '/\[myprocess/ { N;N;N;N; s/^/#/gm }' input_file

edited Jul 29 '12 at 04:35

answered Jul 28 '12 at 19:37

perreal

94,503
21
155
181

slitvinov · Answer 3 · 2023-05-30T21:11:22.220

3

Using string concatenation and the default action in awk.

awk '/myprocess/{f=1} f>5{f=0} f{f++; $0="#" $0} 1'  foo.txt

Alternatively, if the block always ends with an empty line:

awk '/myprocess/{f=1} !NF{f=0} f{$0="#" $0} 1'  foo.txt

edited May 30 '23 at 21:11

answered Jul 28 '12 at 22:17

slitvinov

5,693
20
31

sed - comment a matching line and x lines after it

3 Answers3

Linked