- UCS-2 and UCS-4 builds of Python
- Literals: u"\uabcd", u"\U00020000", or: # -*- coding: -*-
- unicode constructor: unicode (bytes, encoding name, errors=['strict'|'replace'
|'ignore']), or unichr (#)
- unicode_val.encode (encoding name, errors=) returns bytes
- str_val.decode (encoding_name, errors=) returns unicode
- codecs module to handle characters split across buffers
- unicodedata module: names, etc., and normalization functions
- sys.setdefaultencoding sets default coercion, defaults to 'ascii'
(partly because of disagreements over how else it could be set).
It can be set to 'undefined' to require explicit coercion.