How to code sed to delete multiple lines after a keyword until next blank line reached

Question

Can anyone advise me on how to code a sed script that can be used to update a text file as follows:

Each time a line in the file containing the character string 'Header_one' is encountered: - delete the lines that immediately follow the 'Header_one' line until the next blank line is encountered. - but dont delete the 'Header_one' line itself

So as an example, a file containing this:

Header_one
dkrjng
djsje

Header_two
cklrgjsj
djrhg

Header_one
drgbyj
efgjjm
cddess

Header_three
sdfvgg
ddddfy

Would be changed to:

Header_one

Header_two
cklrgjsj
djrhg

Header_one

Header_three
sdfvgg
ddddfy

I would be grateful for any guidance or solution. Thanks, James.

welcome to SO, here you are expected to show what you've tried yourself to solve this problem.. if you are new to `sed`, you can go through https://stackoverflow.com/tags/sed/info for some faqs and learning resources.. after you've some code that you are stuck with, you could ask here for help.. https://stackoverflow.com/questions/38972736/how-to-select-lines-between-two-patterns/ would be a good starting point for this problem — Sundeep, Sep 28 '18 at 12:34

score 1 · Answer 1 · answered Sep 28 '18 at 15:06

Try the below snip

< InputFile sed -e '/Header_one/i Header_one' -e '/Header_one/,/s/d' > outputFile

The idea here is to replace the content between 2 rows and replace it with a header (i.e. Header_one). The second -e part of the codes does delete the data between Header_one and space; while the first -e part replaces it with a new header Header_one.

InputFile and OutputFiles are simple redirections.

You can also look into: https://askubuntu.com/questions/637003/delete-lines-between-2-strings

Hope this helps :)

score 0 · Answer 2 · answered Sep 28 '18 at 14:51

sed is a stream editor, and although one popular version of sed provides an option that appears to make it edit files, I strongly advise against its use. That said, you can produce the output you desire with:

sed -e '/Header_one/i\
Header_one\
\
' -e '/Header_one/,/^$/d' input

Note that it's a lot cleaner if you don't require keeping that blank line:

sed -e '/Header_one/p' -e '/Header_one/,/^$/d' input

Also: awk '/Header_one/{printf "Header_one\n\n"; next } 1' RS= ORS='\n\n' input

score 0 · Answer 3 · answered Sep 28 '18 at 14:56

0

Using Perl

perl -ne ' {  print "Header_one\n" if /Header_one/ ; print if not /Header_one/../^$/ } '

answered Sep 28 '18 at 14:56

stack0114106

8,534
3
13
38

How to code sed to delete multiple lines after a keyword until next blank line reached

3 Answers3