Aug. 2, 2018
1:41 a.m.
The test cases will exercise that some EAI examples go through, but perhaps given the nature of Unicode there should be a few more test examples: control characters (esp. bidi controls) invalid characters fuzzy match equivalents to evaluate if they are treated the same or not, and across different MTAs and MUAs. Accented characters and Unicode normalization (composed and decomposed equivalences) Four-byte utf-8 characters