CODEWITHSUNDEEP
linuxbashsedterminal
will
will

Reputation: 5

Sed search and append starting from the bottom of a file

How can I get sed to search a file from the bottom and find the first string it hits, then append after that? ie:

#file.test

Include sumeStuff1
Include sumeStuff2
Include sumeStuff3
Include sumeStuff4
Include sumeStuff5
Include sumeStuff6
Include sumeStuff7
Include sumeStuff8
Include SED_ENTRY

I know that I can do:

find -type f -name "file.test" | xargs sed -i "s/Include somestuff8/Include somestuff8\nInclude SED_ENTRY/g"

but, the problem is, I am building a script where I don't know what is the EXACT line is that is the last Include line, so I need to search from the bottom and then add a line after the last Include line.

Upvotes: 0

Views: 1242

Answers (3)

NeronLeVelu
NeronLeVelu

Reputation: 10039

sed -n '1 h
$ {/^Include/ s/$/\
/ }
1! H
$ {x;s/\(.*Include [^[:cntrl:]]\{1,\}\)\(\n\)\(.*\)/\1\2Include SED_ENTRY\2\3/p;}' YourFile

Upvotes: 0

Kent
Kent

Reputation: 195179

try this one-liner:

awk 'NR==FNR{if(/^YOUR_PAT$/)n=NR;next}FNR==n{$0="SED_ENTRY"}7' file file

for your example, it would be:

 awk 'NR==FNR{if(/^Include sumeStuff8$/)n=NR;next}FNR==n{$0="Include SED_ENTRY"}7' file file

Upvotes: 1

fedorqui
fedorqui

Reputation: 290025

With awk:

$ tac a | awk 'BEGIN{p=1} /^Include/ && p==1 {print "Include SED_ENTRY"; print; p=0}1' | tac
Include sumeStuff1
Include sumeStuff2
Include sumeStuff3
Include sumeStuff4
Include sumeStuff5
Include sumeStuff6
Include sumeStuff7
Include sumeStuff8
Include sumeStuff8
Include SED_ENTRY

Explanation

  • tac prints a file the other way round.
  • awk 'BEGIN{p=1} /^Include/ && p==1 {print "Include SED_ENTRY"; print; p=0}1' adds the "Include SED_ENTRY" before the first line starting with "Include" and flags it so that the line is not printed again.
  • tac finally reverses the output.

Upvotes: 0

Related Questions