Section 4.1.3 IDNA2008: Convert ASCII domain name to Unicode
In one of the test:
Input:
xn----f38am99bqvcd5liy1cxsg.xn-- rhqv96g
Expected Output:
普遍接受-测试.世界 or 普遍接受-测试。世界
The second expected output had a Chinese punctuation period instead of a dot. While many modern browsers can handle the Chinese period mark, it is not part of the domain. This , IMO, should be treated as negative test case or taken out.
Section 4.2.1 Domain name: syntactic check
I noticed of the TODO tagged to add more tests, and recommend here to make sure we should have test cases that involved CONTEXTO and CONTEXTJ.
Section 4.3.2 Domain name: decompose into components
ua-test.link ---> ua-test, link
I recommend to add ccTLD with subzones to the test suite such as *.
co.uk, *.com.au, etc. This test will give users more insight of whether the tool considered "
ua-test.co.uk" has "uk" or "
co.uk" as TLD. This will make the scoring decision more complicate, but I would love to see this kind of tests conducted and documented.
Rather than writing all above in the doc, I want to bring this up for more discussion. I also have few edit related comments in the doc.
Best,
Joseph