Need help using sed to stop interpreting \n as new line in linux bash scripts - regex

Question

I am new to linux and any bash scripts and have the following problem:
I have this kryptokey:
-----BEGIN CERTIFICATE-----\n
MIICSTCCAfCgAwIBAgIRAMsLZqD4PavC7NJz7+5ld+EwCgYIKoZIzj0EAwIwdjEL\n
MAkGA1UEBhMCVVMxEzARBgNVBAgTCkNhbGlmb3JuaWExFjAUBgNVBAcTDVNhbiBG\n
cmFuY2lzY28xGTAXBgNVBAoTEG9yZzEuZXhhbXBsZS5jb20xHzAdBgNVBAMTFnRs\n
c2NhLm9yZzEuZXhhbXBsZS5jb20wHhcNMTgxMjMxMTA1ODA5WhcNMjgxMjI4MTA1\n
ODA5WjB2MQswCQYDVQQGEwJVUzETMBEGA1UECBMKQ2FsaWZvcm5pYTEWMBQGA1UE\n
BxMNU2FuIEZyYW5jaXNjbzEZMBcGA1UEChMQb3JnMS5leGFtcGxlLmNvbTEfMB0G\n
A1UEAxMWdGxzY2Eub3JnMS5leGFtcGxlLmNvbTBZMBMGByqGSM49AgEGCCqGSM49\n
AwEHA0IABEbH7l3CiqLA4N4wgfilYgyEuxDrMAqDX6BrFOfWhymNosjh5FlJDHtN\n
GPDKhjtrI6e1q0NC0l6wh9h9TrBn7N2jXzBdMA4GA1UdDwEB/wQEAwIBpjAPBgNV\n
HSUECDAGBgRVHSUAMA8GA1UdEwEB/wQFMAMBAf8wKQYDVR0OBCIEIH7OaekSLJda\n
S0yuV9PCsuasGTt/+/35aVBXTVbII2rCMAoGCCqGSM49BAMCA0cAMEQCIEd+YP/6\n
tCzG/LueYTEio8ApQSyz94ju07pmc3LZJDKBAiALu66LKhOpKhogY9XEFg4TScOt\n
el4dC6OnMMTmRsEtoA==\n-----END CERTIFICATE-----\n
saved in a file $replacementOrg1 (is the path to that file).
Now I want to replace in a template $file "INSERT_ORG1_CA_CERT" with this certificate and safe it in $org1. But I need to keep the "\n" Character.
The result should keep the \n and write it into one line.
I already tried:
sed -e "s#INSERT_ORG1_CA_CERT#$(cat $replacementOrg1)#g" $file > $org1
but it interprets the "\n" as new line.
So the final Output should look like this, 1 String in 1 Line:
"-----BEGIN CERTIFICATE-----\nMIICSTCCAfCgAwIBAgIRAMsLZqD4PavC7NJz7+5ld+EwCgYIKoZIzj0EAwIw djEL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\n
el4dC6OnMMTmRsEtoA==\n-----END CERTIFICATE-----\n"
Anybody can help?
Thank you

Answer 1

That is not a valid key. What someone has done is "half-encoding" (I don't know a better term) the newlines - they have added the literal string "\n" before every newline. What you very likely want is either the original key with no "\n" strings or a single line string where every newline has been replaced with "\n".
With the original value you can use replace instead - it supports newlines in the replacement value:
$ replace foo $'foo\nbar' <<< $'x\nfoo\ny'
x
foo
bar
y
Your case should be simply replace 'INSERT_ORG1_CA_CERT' "$(< $replacementOrg1)" "$file" > "$org1".

Answer 2

The substitute command isn't very good with multi-line replacement strings. But we can use GNU sed's read command to work around that:
echo "${replacementOrg1}" |
sed -e '/INSERT_ORG1_CA_CERT/{r /dev/stdin' -e ';d}' ${file} > ${org1}
How it works:
echo the multi-line string, piping it to /dev/stdin.
When sed finds the target "INSERT_ORG1_CA_CERT" it reads /dev/stdin and outputs the contents
then deletes the search string line, (which is presumed to contain no other text).
The tricky part is the inadequately documented r command -- sed assumes everything after the r is part of the filename. If we tried '/INSERT_ORG1_CA_CERT/{r /dev/stdin;d}' it would bomb with the error:
unmatched '{'
Because sed would think the filename was literally "/dev/stdin;d}". But the error message doesn't complain about the missing file, because sed never complains about a missing r filename. Instead sed complains that there's no } closing brace, because sed thinks the } is part of the filename.
To avoid that error we stick an ' -e ' in there.