Recently discovered a wonderful little application called OpenVanilla ( http://www.openvanilla.org/
), which basically allows you to design and use custom IMEs for both Windows and Mac. I'd been thinking about doing one to help with transcription of Japanese, and so I downloaded it and decided to muck around. Between checking wiki and using what I already know, I think I've hammered out a decent Japanese one for the most part. Curiosity and a persistent interest in comparative phonology got me poking around some and I've started working on ones for Korean and Vietnamese as well (the latter having separate ones for the Hanoi and Saigon dialects).
That said, as I know there are some here who are much more up on their phonology than I am, I thought I'd put it out there for a bit of collaboration. I've bundled up the .cin files I've done so far in a zip file here: http://www.sakura-sake.net/kaji/ipa.zip
. If anyone wants to check through and proofread them or offer advice on how to make them more responsive to sound changes in each language (e.g. variation in the pronunciation of Korean consonants based on where they are located in the character).
For those who are interested, a brief description of the status of each:Japanese
Think this one's pretty well set for the most part. It's romaji based, and can accept either one English character at a time (useful for isolating phonemes) or one kana at a time (e.g. it'll recognize 'k' or 'ka', but not 'kare' since the last one hasn't been programmed in yet). I've also got it set up to recognize ん after words to make things a bit easier. ん on its own is n'
so as to distinguish it from the initial n's phoneme in isolation. Next planned step is to make it factor in long vowels.Korean
Think I have the basics put together so far. It's currently set up to handle any combination of a single initial consonant and a vowel, and should compensate for any allophonic shifts resulting from such combinations (e.g. it knows that ㅅ + ㅣ is supposed to be /ɕi/ and not /si/). Final consonants (if present) have to be added individually after the combined initial and vowel. Adding an apostrophe after a consonant will change its form to its variant if it has a separate form for the end of a character (e.g. a final ㅅ would be rendered as /t/ if entered as s'
). Jamo locations are the same as Apple's HNC Romaja input method. Consonant clusters have not been factored in yet, because I'm not quite sure what to render them as and none of the resources I can find address them.Vietnamese
Letter and tone locations are based on Apple's Vietnamese IME. I've scoured Wiki's references on Vietnamese phonology and set up just about all of the consonants, consonant clusters, vowels, and polyphthongs, as well as factoring in phonological differences between northern and southern dialects into two separate versions. Consonants and vowels presently need to be entered separately, but everything should work fine otherwise. If someone knows how to properly mark for tones in IPA I'm all ears; in the meantime I'm using the notation used in the article as a placeholder.
Here's hoping someone else finds some of this useful!