# If you'd like to omit non-matching lines from the result; add ';d' to the end of the expression.
sed -E 's/\(([^)(]*)\)(?=(?:[^)(]*\([^)(]*\))*[^)(]*\))/$1/gm;t' <<< "Record ID
What is the best food? (choice=Nachos)
What is the best food? (choice=Tacos (big bent nachos))
What is the best food? (choice=Chips with stuff)
Complete?
So ((this) one (has) (many) (inside)) (none here) (but (here) and (here))
a(b(c)(c)b(c))(b((d)c(d)c(d)(d)))"
Please keep in mind that these code samples are automatically generated and are not guaranteed to work. If you find any syntax errors, feel free to submit a bug report. For a full regex reference for SED, please visit: https://www.gnu.org/software/sed/manual/html_node/The-_0022s_0022-Command.html