Peter Kirk scripsit: > I have just reviewed this list and found it odd that Hebrew presentation > forms are included but Arabic ones are not.
The specification actually called only for Latin, Greek, and Cyrillic; I added Hebrew pour la lagniappe. If someone wants to add Arabic, I encourage them to do so. > the Hebrew presentation forms but also most of the precomposed > characters are redundant in this list. True; however, the current list indicates the scope of what actually happens, even if it is overlong. > It is therefore > necessary to list in the specification of the folding only all (?) > combining marks, which are to be deleted, I believe that all Mn-class characters, and only they, are deleted by this. > I note that 0429 is not folded to 0428 etc, and this is correct because > within the Cyrillic writing system these are entirely separate > characters. But the difference between these two is in fact exactly the > same descender which is removed in 0496 etc. I don't think that matters. Long historical practice has made SHCHA a separate letter, just as G, J, U, and W are now separate Latin letters from C, I, V, and VV-ligature. > I am also surprised to note > that no folding is given for 0419/0439; although in some ways this is > desirable because Russians do not consider this breve to be a diacritic > (and after all we would not want the dot on i to be removed as a > diacritic!), these characters have canonical decompositions to 0418/0438 > and breve and the principle of canonical equivalence and the folding > algorithm (which works on decomposed characters) more or less demand > that the breve be deleted. Also 048A/048B should then fold to 0418/0438 > rather than 0419/0439. I think I agree with this: i-breve does not have the same universal status as shch. -- John Cowan www.reutershealth.com www.ccil.org/~cowan [EMAIL PROTECTED] 'Tis the Linux rebellion / Let coders take their place, The Linux-nationale / Shall Microsoft outpace, We can write better programs / Our CPUs won't stall, So raise the penguin banner of / The Linux-nationale.