Full_Name: Zebediah Figura Version: trunk OS: (Kubuntu 16.04) URL: ftp://ftp.openldap.org/incoming/zebediah-figura-160930.patch Submission from: (NULL) (132.161.178.173)
The file ucgendat.c, which is used to generate a list of (among other things) Unicode case correspondences, only generates correspondences for the title-case characters which have both a corresponding lower-case and upper-case forms (Dž, Lj, Nj, Dz). This is because read_cdata() checks whether there exists both a lower-case and an upper-case form, but Unicode does not provide an upper-case form for title-case characters (Greek characters with iota subscript) which do not have a distinct upper-case form. Thus these are mistakenly treated as upper-case characters.
I have attached a patch which causes read_cdata() to treat a character as title-case if it has the property Lt (i.e. Letter, title) and has either an upper-case or lower-case form defined.