Re: compiler for Chinese development language

Oliver Wong <owong@castortech.com> wrote:
>...It's not too bad to memorize an alphabet. With English, that's
>only 52 characters (you have to learn both the uppercase and lowercase
>version of every character, as they differ significantly)...


The number is actually a bit higher than that, because there are a few
letters which vary in basic shape from font to font; English speakers
are so used to this that they seldom notice it, but newcomers have to
learn the variations as separate forms. In italics, "f" grows a tail,
"Q"'s tail grows to almost an underline, and "a" is a completely
different shape (loop with a slight tail, rather than low loop with a
roof over it); in Helvetica and a lot of other sans-serif fonts, "g"'s
tail is a line rather than a loop. Printer and terminal(-emulation)
fonts pick one or the other almost at random.


(It's easy to dismiss the changes in tails in particular as trivial, but
differences between letters often are no bigger -- "j" is just "i" with
a tail, for example.)


>Even the Japanese Katakana alphabet has around 100 characters...
> Incidentally, the Japanese Katakana alphabet has a completely
>unambiguous pronounciation: Each chararacter represents one syllable...


The two facts are, of course, connected: you need more characters to
give each syllable a unique one.


By the way, a linguistic nitpick: technically an alphabet is a writing
system with (approximately) one character per sound, like the English one,
and Katakana is a syllabary, not an alphabet. People who independently
invent writing systems generally invent either syllabaries or ideographic
systems; it appears that the alphabet concept was invented just once, by
some obscure neighbors of the Phoenicians, and all other alphabets derive
at least inspiration from theirs.
--
spsystems.net is temporarily off the air; | Henry Spencer
mail to henry at zoo.utoronto.ca instead. | henry@spsystems.net
[Fascinating though this discussion is, it's veered away from compilers, so
unless someone can tell us about compiling kana into kanji or something,
this thread is at an end. -John]

Return to the comp.compilers page.
Search the comp.compilers archives again.