Jurgen J. Vinju

Excellent issue! Are you looking for support in the abstract syntax suport and M3 relations too? or are we just looking for the extensions of the grammars? Maybe both would...

ok that's clear. we have several analysis features for Java which are based on first running the Eclipse compiler front-end to extract all kinds of facts and relations. That model...

I think that makes sense. It's not completely backward compatible because new reserved keywords were added here and there. It's smart to go ahead and implement Java17 first, and then...

this alternative with brackets is not ambiguous: ``` [ info("overrides Object.hashCode()", l) | l

Looks like the `

In what way was that branch still a work in progress? The current situation is very simple: * many tests are failing that accept random `datetime` parameters because large parts...

My current ideas go to a temporary change in the vallang library like so: ``` boolean partialDateTime = "true".equals(System.getProperty("vallang.random.partialDateTime")); try { if (partialDateTime && random.nextDouble() > 0.8) { LocalTime result...

Ok. ran a test and that would solve half of the issues Next issue: the Rascal implementation does not deal well with zone offsets: ``` [ERROR] jsonWithDatetime1: (org.rascalmpl.test.infrastructure.RascalJUnitParallelRecursiveTestRunner$ModuleTester) Time elapsed:...

Ah. We loose offsets when writing/reading JSON like this: ``` @Override public IValue visitDateTime(Type type) throws IOException { try { switch (in.peek()) { case STRING: // toEpocMilli is the wrong...

Very much agreed. Now working on Zulu time issues.