Speakeasy is a ruby library that maps a language to the minimum set of unicode codepoints required to represent it. The goal is to be minimal when including codepoints and exclude common punctuation characters when possible.

All languages will be referenced by their IANA language subtags.

Each language's data is a standalone YAML file contained in the data directory of this project.

data/
  ca
  cs
  de
  ...

Each YAML file will have the following keys and must be saved in UTF8:

• version (currently 2)
• anglicized_name (the language name in English)
• native_name (the name of the language in the language)
• codepoints (currently a list of ranges or a single codepoint see data/en for an example).

An example YAML file:

---
version: 2
anglicized_name: English
native_name: English
codepoints:
- !ruby/range 65..90 # Uppercase
- !ruby/range 97..122 # Lowercase

Note: codepoints should only include normalized unicode codepoints.

• Fork speakeasy
• Create a topic branch - git checkout -b my_lang
• Create or edit data/my_lang
• Verify it's correctly formatted - rake test[my_lang]
• Verify it has all the characters you want - rake visualize[my_lang]
• Push to your branch - git push origin my_lang
• Create a github pull request.
• That's it!

• Language Tags RFC5646
• http:https://www.langtag.net/
• http:https://github.com/runpaint/unicode-data

Matt Colyer / [email protected]

Copyright (c) 2010 Typekit. Licensed under the Apache 2.0 license.