bash: string variable contains asterisk. how to use this variable for searching etc with grep, sed?

Question

Have a nice day I have got text file (zz.txt):

Chemical name
3-Aminopropane-1-sulphonic acid
Synonym(s)
--
Homotaurine * Tramiprosate
--
Chemical name
Common name and synonyms
...

I have variable s="Homotaurine * Tramiprosate"; I try to (un)select text of this variable from text file:

s="Homotaurine * Tramiprosate"; cat zz.txt | grep -i "$s"

s="Homotaurine * Tramiprosate";cat zz.txt | sed -ne "/$s/p"

Does not work. But if I change $s to, for instance, "Chemical name", everything ok, grep and sed work correctly and similarly. Variable $s received from another text file. Next time it could contain asterisk or not. Yes I realize that asterisk is very special character. How can I solve my task with sed/grep? And additional question - what others similar characters could create headache?

Answer 1

For grep, you can use -F to use fixed strings instead of regexes. For sed, it's much more complex. I'd probably switch to Perl which can help you with its \Q/quotemeta function.

#!/bin/bash
s="Homotaurine * Tramiprosate"; grep -Fi -- "$s" zz.txt
s="Homotaurine * Tramiprosate"; s=$s perl -ne 'print if /\Q$ENV{s}/' -- zz.txt