Document use of custom character sets (python)

[ixcat]

Use of non-latin1 character sets should be documented - Currently, in python, this means:

• configuring server my.cnf for character set/collation or manually creating/altering schemas:
  • for globally-applicable server my.cnf :

    character-set-server = utf8
    collation-server = utf8_unicode_ci
    skip-character-set-client-handshake

  • for per-schema / per table / per attribute: use ALTER TABLE, etc as needed note: since client setting is global (see below), this may still yield compatibility issues in some mixed-character set cases
• setting dj.config['connection.charset'] in all clients to match desired character set for client network encoding

MATLAB tbd.