Parse Unikemet

[eggrobin]

No description provided.

[markusicu]

just FYI: I always stumble a little over "=" meaning "assign" on a "Let" line but "assertEquals" on a test line (without a statement prefix) ... It would be a disruptive change, but using "==" for "assertEquals" might make this file more obvious.

[eggrobin]

Or := (or ≔?) in Let (that would keep the actual invariant statements closer to normal mathematical operators on sets).

Not today, but good idea.

[markusicu]

Let with ASCII := would be ok

[markusicu]

lgtm but we need to remember that if UTC-180 disagrees with the PA/PVA additions, then we will need to move these back to the Extra files

[markusicu]

Let with ASCII := would be ok

[markusicu]

CI failure:

Error:  org.unicode.propstest.TestProperties.TestDefaults  Time elapsed: 2.912 s  <<< FAILURE!
org.opentest4j.AssertionFailedError: Null kEH_NoMirror for U+FFFF but property is Binary with default value type LITERAL. Add an @missing line to ExtraPropertyValueAliases.txt if needed. ==> expected: not <null>
	at org.unicode.propstest.TestProperties.TestDefaults(TestProperties.java:614)

[eggrobin]

CI failure:

Ah yes, I went too far in my removal in Extra, and there’s an illustration of the answer on needing those lines for the binaries 😩

[eggrobin]

lgtm but we need to remember that if UTC-180 disagrees with the PA/PVA additions, then we will need to move these back to the Extra files

Or if UTC does not take the decision recommended by https://github.com/unicode-org/sah/issues/496 (I kept Core in Extra in anticipation of that).