This will work with Unicode strings, using a technique called "case-folding" to handle the vast majority
of case-sensitive human languages regardless of system locale. It can deal with expanding values: a
German Eszett character can compare against two ASCII 's' chars and be considered a match, for example. A
notable exception: it does not handle the Turkish 'i' character; human language is complicated!
Since this handles Unicode, it expects the string to be well-formed UTF-8 and not a null-terminated
string of arbitrary bytes. Bytes that are not valid UTF-8 are treated as Unicode character U+FFFD
(REPLACEMENT CHARACTER), which is to say two strings of random bits may turn out to match if they convert
to the same amount of replacement characters.
Note that while this function is intended to be used with UTF-8, maxlen specifies a _byte_ limit! If the
limit lands in the middle of a multi-byte UTF-8 sequence, it may convert a portion of the final character
to one or more Unicode character U+FFFD (REPLACEMENT CHARACTER) so as not to overflow a buffer.
maxlen specifies a maximum number of bytes to compare; if the strings match to this number of bytes (or
both have matched to a null-terminator character before this number of bytes), they will be considered
equal.