Copy one of fft-orig.cc or fft-improved.cc into fft.cc.
Then you can use the supplied Makefile (i.e. simply type make to build).
Notes for building:
Use c++23, otherwise std::println will not compile.
With Clang, use -Wno-mathematical-notation-identifier-extension to suppress the warning about subscript names (prevents the literal interpretation of a mistake by the Unicode consortium).*
With GCC, refrain from using -pedantic for the same reason.
*) “Apparently that is Unicode's fault. Subscript numbers 0-9 do not belong to (X)ID_Continue, because their category is Number, Other (No), not Number, Decimal (Nd).”
(Parabolicus)