On Sun, 30 Dec 2001, Ken Williams wrote: > Yeah, that seems like a pretty good orthogonalization. I might shorten > 'Romanization' to 'Roman', though.
Roman seems confusing. Its fairly common to talk of "romanization schemes" for languages that don't use the roman alphabet. People interested in Chinese will recognize Romanization a little quicker than just Roman, I think. > HBF doesn't mean much to me (I'm not very savvy in this field, though), > perhaps a more descriptive name like 'Shape' or 'Picture' or 'Font', > maybe? Yeah, this is a very weird issue. HBF is a an _old_ (1995 or earlier) file format used to describe a bitmap font. Its basically a set of key/value(s) pairs, one per line. A snippet looks like this: HBF_START_FONT 1.1 HBF_CODE_SCHEME Big5 KC 6.22 FONT kcm24 Part of this info is a pointer towards the files that contain the actual bitmaps. The code I have reads this info and can then use GD to generate a bitmap for a given set of characters. Calling it HBF isn't all that helpful because people won't necessarily recognize that. OTOH, calling Lingua::ZH::Bitmap or something is misleading, because without some HBF files and fonts it won't work very well. On the other other hand, the HBF files I have say they are in the public domain and I believe that the bitmap files therefore are as well, in which case I could distribute them with the module and it'd just work. I'll have to try to investigate but this stuff is all really old and finding the original sources and authors is painful. The real kicker is that most of this work (except the romanization) will be obsoleted by Unicode in a few years. Oh well, its still useful for a while. > Plain old Lingua::ZH would seem to make sense, unless that's playing too > greedy with the namespaces. Another option is Lingua::ZH::Tools, which > is similar to existing module collections like File::Tools and so on. There is already a Lingua::ZH::HanConvert module which converts from traditional to simplified Chinese Unicode characters. Taking Lingua::ZH would be a bit rude. Lingua::ZH::Tools works fine. Also, I mentioned in my first message I had files with unknown copyrights that contained tables for implementing some of this. Fortunately, I just found a file under the OPL that contains Unicode code points for a huge number of characters and their corresponding romanizations in a number of schemes, including about 8 different hakka pronunciations, cantonese, and mandarin plus info about stroke counts and such. This can be very useful. -dave
