An identifier reduced to the form a database collation may compare it in: Unicode compatibility
decomposition (a KELVIN SIGN reads as K, a full-width letter as its ASCII form), accents and
other combining marks removed, control, format and default-ignorable characters (zero-width
joiners, direction marks, a soft hyphen) removed, case folded through upper case (ß and ẞ both read ss, a
final ς reads σ), and surrounding spaces trimmed.
An identifier reduced to the form a database collation may compare it in: Unicode compatibility decomposition (a KELVIN SIGN reads as
K, a full-width letter as its ASCII form), accents and other combining marks removed, control, format and default-ignorable characters (zero-width joiners, direction marks, a soft hyphen) removed, case folded through upper case (ßandẞboth readss, a finalςreadsσ), and surrounding spaces trimmed.