I use Awk to search a keyword <replaceme> and replace it by a string. This string contains the chars "\n", but Awk interprets it as a new line. I really want to keep both characters, \ and n.
I've a string containing a real new line and I need to replace it by the chars "\n". It works:
mytext=$'hello\nhow are you'
mytext2=${mytext//$'\n'/\\n}
echo "$mytext"
hello
how are you
echo "$mytext2"
hello\nhow are you
Then I have a keyword <replaceme> in a text and I would like to replace it by the content of $mytext2
echo 'this is the text: <replaceme>' |
awk -v srch="<replaceme>" -v repl="$mytext2" '{
gsub(srch,repl,$0); print $0 }'
I get this result.
this is the text: hello
how are you
But I would like:
this is the text: hello\nhow are you
How to proceed, please?
Note: in fact, the text "this is the text: " is in a file and I use cat myfile.txt | awk...
You may use this awk solution:
mytext=$'hello\nhow are you'
echo 'this is the text: <replaceme>' |
awk -v srch="<replaceme>" -v repl="${mytext//$'\n'/\\\\n}" '{gsub(srch,repl)} 1'
this is the text: hello\nhow are you
BASH replacement ${mytext//$'\n'/\\\\n} will replace each instance of \n with \\n before passing that string as awk argument repl.
Pure awk solution (without bash):
echo 'this is the text: <replaceme>' |
awk -v srch="<replaceme>" '
BEGIN {
repl = ARGV[1]
gsub(/\n/, "\\n", repl)
delete ARGV[1]
}
{
gsub(srch, repl)
} 1' "$mytext"
this is the text: hello\nhow are you