CODEWITHSUNDEEP
bashawksed
ezzymn
ezzymn

Reputation: 33

Delete from pattern match until next pattern

From bash, I have a input file with various keywords and sometimes data on following lines. I would like to delete keywords starting from PATTERN1, until and not including lines starting with PATTERN2. Note :

I've tried sed such as 

sed -i.bak -e '/*FOO/,/*/d' -e '/*BAR/,/*/d $FILE

but this deletes the *KEEP line .

**START
*FOO
This wants to be deleted
*KEEP
*BAR
this also wants to be deleted
*KEEP
**END

should become

**START
*KEEP
*KEEP
**END

Thanks in advance for looking at this question

Upvotes: 3

Views: 93

Answers (2)

potong
potong

Reputation: 58558

This might work for you (GNU sed):

sed '/^*/h;G;/^*\(FOO\|BAR\)/M!P;d' file

Store each key as it presents itself, in the hold space and append it to the current line. Print the first of the two lines in the pattern space if either does not contain a line beginning *FOO or *BAR.

Upvotes: 0

anubhava
anubhava

Reputation: 786041

It is easier to do this using awk:

awk '/^*/{d = 0} /^*(FOO|BAR)/{d = 1} !d' file


*START
*KEEP
*KEEP
**END

If you're using gnu awk then use -i inplace to save changes inline:

awk -i inplace '/^*/{d = 0} /\*(FOO|BAR)/{d = 1} !d' file

If not using gnu awk then use:

awk '/^*/{d = 0} /^*(FOO|BAR)/{d = 1} !d' file > $$.tmp && mv file $$.tmp

Upvotes: 1

Related Questions