Arabic character semantics are not established in the cmap subtable, as would be expected. By itself, the Windows Symbol encoding implies no interoperable character semantics, but rather implies font-specific semantics. These fonts use Windows Symbol encoding - that is, they have a cmap subtable for platform ID 3, encoding ID 0. Font encoding and character set declarations
However, they make use of undocumented details not defined in the TrueType or OpenType specs, and do not conform to current specifications for how to implement Unicode fonts for Arabic. ttf files that, in general, follow the TrueType spec. These are TrueType fonts, in the sense that they are.
This topic provides some details regarding this class of fonts. There is a class of legacy fonts for Arabic that use not-well-documented mechanisms and legacy shaping implementations in Windows.
