I have some file:
aaaaaaaaaa
# Lines Start #
bbbbbbbbb
cccccccc
dddddddddd
# Lines End #
eeeeeeeeee
And some variable $newLines="new line"
How can I replace the lines from # Lines Start # to # Lines End # (including those line themselves) with variable's value using sed.
so I'll get:
aaaaaaaaaa
new line
eeeeeeeeee
I've tried using: sed -E 'N; s/# Lines Start #\.*\# Lines End #/$newLines/'
But its not doing it's job.
You can use awk if you like to try it:
awk '/# Lines Start #/ {f=1} !f; /# Lines End #/ {print "new line";f=0}' file
aaaaaaaaaa
new line
eeeeeeeeee
You can also use range like in sed, but its less flexible, so I normal avoid the range function.
awk '/# Lines Start #/,/# Lines End #/ {if (!a++) print "new line";next}1' file
aaaaaaaaaa
new line
eeeeeeeeee
Here is one way with sed:
sed '/# Lines Start #/,/# Lines End #/{
/# Lines End #/!d
s/# Lines End #/new line/
}' file
aaaaaaaaaa
new line
eeeeeeeeee
new lineThis might work for you (GNU sed):
var='first line\
second line\
third'
# N.B. multilines must end in "\" except for the last line
sed '/# Lines Start #/,/# Lines End #/c\'"$var" file
c \ replaces the selected lines with the text. The only caveat that potong rightly stated in his answer is if you use a variable with embedded new lines then they must be preceded by a backslash.