Bash, match string and then replace everything after character in next line

Question

What I am looking to do is find a match and then replace everything after character in the next line.

Example file contents:

  super_service:
    app_version: 1.02
  duper_service:
    app_version: 4.0.1
  happy_service:
    app_version: 2.03
  turbo_service:
    app_version: 1.28.0

Expected outcome: What I am looking to do is match duper_service for example, and then replace everything after its' "app_version: " with a new version I define, 4.0.2 for example. I am able to do this with awk but need to write the change to the file.

cp originalFile original.back && awk '{ .... }' originalFile > output.tmp && mv output.tmp originalFile — αғsнιη, Apr 26 '21 at 23:08

xhienne · Answer 1 · 2021-04-26T23:39:48.937

0

sed seems a good candidate for this:

sed -i '/^[[:space:]]*duper_service:/ { N; s/[^:]*$/ 4.0.2/ }' file.txt

Explanation: when the "duper_service:" line is encountered, next line is appended and whatever ends that line, after the colon, ([^:]*$) is replaced with "4.0.2".

The -i option changes the file in-place. It's a GNU sed extension.

edited Apr 26 '21 at 23:39

answered Apr 26 '21 at 23:07

xhienne

17,793
2
53
69

This is great but a version may not always start with a number. – David Turner Apr 26 '21 at 23:13
Read the title again, "replace everything after character in next line". – David Turner Apr 26 '21 at 23:19
That looks great, it is not working for some reason though. – David Turner Apr 26 '21 at 23:26
Removing the ^ seems to have done the trick. – David Turner Apr 26 '21 at 23:28
No, that one is good, "^duper_service" was where the problem occurred for me. Thank you so much for your wonderful command and time spent. This is wonderful! – David Turner Apr 26 '21 at 23:38
My pleasure! Ok, I understand now why the first version didn't work: there are space characters before "duper_service". Answer edited accordingly. It's better not to /duper_service:/ as this regex may also match other lines like "extra_duper_service:" – xhienne Apr 26 '21 at 23:43

David Turner · Answer 2 · 2021-04-26T23:21:49.987

0

Looks like this works:

sed -i '/duper_service/!b;n;c\ \ \ \ app_version: 4.0.2' file.txt

edited Apr 26 '21 at 23:21

answered Apr 26 '21 at 23:14

David Turner

1

Bash, match string and then replace everything after character in next line

2 Answers2