CODEWITHSUNDEEP
bashshellawksed
David Lovell
David Lovell

Reputation: 17

Finding text within a string and replacing some text on the next line

I have a file, which has some text similar to below:

#my-var="Group Name",*ItemInGroup
ThisIsTheResult/aaa/bbb/1
#my-var="Group Name",ItemInGroup
ThisIsTheResult/aaa/bbb/2
#my-var="Group Name",ItemInGroup2
ThisIsTheResult/aaa/bbb/3
#my-var="Another Group Name",*AnotherItemInGroup
ThisIsTheResult/aaa/bbb/4
#my-var="Another Group Name",AnotherItemInGroup
ThisIsTheResult/aaa/bbb/5

However, I would like to search through the document, and if it finds my-var="+++++++++++",* (where +++++++++++ could be anything), then it moves on to the next line, and replaces aaa/bbb with yyy/zzz. The expected outcome, is as below: 

#my-var="Group Name",*ItemInGroup
ThisIsTheResult/yyy/zzz/1
#my-var="Group Name",ItemInGroup
ThisIsTheResult/aaa/bbb/2
#my-var="Group Name",ItemInGroup2
ThisIsTheResult/aaa/bbb/3
#my-var="Another Group Name",*AnotherItemInGroup
ThisIsTheResult/yyy/zzz/4
#my-var="Another Group Name",AnotherItemInGroup
ThisIsTheResult/aaa/bbb/5

Was looking to do this in a shell script, but my regex and sed/awk experience isn't good enough... So hopefully someone know the answer easily :)

Cheers

David

Upvotes: 0

Views: 27

Answers (1)

Cyrus
Cyrus

Reputation: 88543

With GNU sed:

sed '/my-var=".*",\*/{n;s|aaa/bbb|yyy/zzz|}' file

If a lines matches regex my-var=".*",\* then read next line (n) and search (s|||) for aaa/bbb and replace with yyy/zzz.

Output:

#my-var="Group Name",*ItemInGroup
ThisIsTheResult/yyy/zzz/1
#my-var="Group Name",ItemInGroup
ThisIsTheResult/aaa/bbb/2
#my-var="Group Name",ItemInGroup2
ThisIsTheResult/aaa/bbb/3
#my-var="Another Group Name",*AnotherItemInGroup
ThisIsTheResult/yyy/zzz/4
#my-var="Another Group Name",AnotherItemInGroup
ThisIsTheResult/aaa/bbb/5

Please take a look at man sed and The Stack Overflow Regular Expressions FAQ.

Upvotes: 1

Related Questions