dan-blank / javatomips_testcases Goto Github PK

@Jonast the file jt_ValidOverride.java should not be valid because of BaseType != SubType.

Thanks to you I found a bug in my typechecker! :)

More at: http://softwareengineering.stackexchange.com/questions/267310/overriding-methods-by-passing-as-argument-the-subclass-object-where-the-supertyp

And less helpfull, but more official:

https://docs.oracle.com/javase/tutorial/java/IandI/override.html

At the line 21 of ok/db_expressions we have:
true = !false;

However, we have only production id = Exp, which means that true should be id, but I guess it should be not possible, because true is a keyword.

Dann füge ich noch meine Tests hinzu. Danke!

At the line 23 of ok/db_expressions we have:
new Object().apply(2,3,4,new OtherObject());

That is incorrect because it is Exp, and we expect Statement, we do not have production from one to another.

@Jonast in jt_InvalidMemberOverride:

It is okay that the subclass has a member with the same name, even it is of a different type.

https://docs.oracle.com/javase/tutorial/java/IandI/hidevariables.html

@SgnJp some of your ok-testcases contain overloading, which is not allowed in MiniJava, source: https://proglang.informatik.uni-freiburg.de/teaching/compilerbau/2016ws/ex2.pdf

Sorry that this comes up so late :/

Let's change the structure from
exercise<number>/<initials>/[ok|error]/<initials>_<test_case_name>.java
to
exercise<number>/[ok|error]/<initials>_<test_case_name>.java

I guess it would be easier to use it directly, and the test case owner is defined by the initials at the begging of test name

@Jonast : in Class B, there is the method

public int test(int a, boolean c) {
int[] a;
a = new int[5];
return a[3];
}

This should be illegal according to the java specification

I researched a bit to make sure and found this thread with the relevant quote in the top-answer: http://stackoverflow.com/questions/4623334/question-about-variable-scope-and-shadowing-in-java