On Sat, 29 Dec 2001, Hugh S. Myers wrote: > Lingua is certainly obvious, but ZH doesn't tell me much. I suggest the > straight forward Lingua::Chinese as the namespace with the remaining names > the same. Can't say it will make much difference in my own scheme of things, > but I wouldn't make it harder to find than it needs to be! Now if you were > talking about Tengwar or something useful<g>!
Well, in the Lingua::* namespace the convention seems to be to call modules something like Lingua::<iso 2 letter language code>::Foo. ZH is the code for Chinese. For the curious, this is because in Mandarin, the characters for saying "Chinese language" can be romanized, using the Pin Yin system, as "zhong1 wen2" (the numbers are tone indicators). Its kind of a dumb code though because Chinese character sets also get used for other dialects, such Cantonese, where its pronounced differently (though I don't know exactly how). Anyway, I just noticed that there is already a Lingua::RU::Charset for Russian charset transformations, so I'm thinking that I could release the modules as: Lingua::ZH::Charset (convert between big5, gb, and unicode, maybe) Lingua::ZH::Romanization - romanize mandarin and cantonese pronunciations Lingua::ZH::HBF - read HBF font files and generate GD bitmaps That seems like a fairly rational set of choices. The other question is whether to release them as a unit (Lingua::ZH::Utils or something) or as three separate tarballs. -dave
