Martin Kochanski <unicode at cardbox dot net> wrote: > I read, somewhere, that certain code point ranges had been allocated > properties (such as LTR/RTL) in the Unicode tables even though some > of them had not yet had characters defined for them. Possibly someone > can penetrate the vagueness of this memory and confirm or deny?
Broad ranges of Planes 0 and 1 have been tentatively blocked out on the Roadmap for RTL scripts. Software *MIGHT* be able to take advantage of this and make certain assumptions (to the extent it is allowed to make any assumptions at all about unassigned code points). But AFAIK there are no promises that an LTR character or script might not be assigned to these ranges, if space gets scarce enough. > If this is the case, have U+035D - U+035F already been assigned the > "COMBINING" property? This is probably even vaguer and less reliable than the LTR/RTL blocking. But again, I suppose you could make certain internal optimizations based on the assumption (as of today) that characters in that range are combining. -Doug Ewell Fullerton, California
