Data gathering

[hugolpz]

We currently look for database with { "glyph": "西", "phonetic": "xī" } (or xi1, or alternatives).

Sources possible, info to complete :

Moedict

• link (to complete)
• json format
• range : most common caracters, trad only ?

Unicode :

• link (to complete)
• xml
• range : traditional/modern ; -more complete for a font
• which phonetic format its provided also. ("glyph": "西", "phonetic": "xī" or xi1 ?)

CJKlib

• link

[edouard-lopez]

What about Unihan?

With the hexadecimal codepoint we can get the glyph like this in Python:

>>> print(chr(int('0x897F', 16)))
西

A JS solution would be better, but this is out of the scope of the project, we can do it anyway we think fits.

[hugolpz]

Please check out :

• unihan results
  -- npm unihan
  -- npm convertPinyin
  -- npm unihan-cjk

• PinNum2PinTones
  -- npm pinyinize
  -- pinyin-string -- best one !

[edouard-lopez]

Thanks for the link cjk-unihan might be useful for other projects.

I think it's better to limit the project to generating font and outsource the data gathering/validation to another project. This way we stay focus and efficient.

I'm closing as different users might have different needs hence handcraft their dictionaries.

[edouard-lopez]

I reckon the JS solution is in tobei/unihan code

const character = String.fromCodePoint(parseInt(code.substring(2), 16));

[hugolpz]

Did you gathered the data ?

[edouard-lopez]

Not yet, could you work on a project to do so?

[hugolpz]

Yup. See also peterolson/hanzi-tools#1 (comment)

[edouard-lopez]

@hugolpz I think you have a typo in your comment, there is a ratio of 1:10 between node-pinyin and unihan characters/phonetic pairs. Can you confirm/correct this number?

[hugolpz]

https://github.com/superbiger/pinyin4js/blob/master/src/dict/pinyin.dict.js

[edouard-lopez]

We can get the codepoint using punycode