Add back CopyConstructors for instantiated array templates

[aymanhab]

Fixes issue #0

Brief summary of changes

Updating the Array class resulted in missing Copy constructors for all Array<> classes in Java/Matlab. This PR adds them back so that GUI code and possible user code that depends on them continue to work.

Testing I've completed

Building, checking resulting Java files

Looking for feedback on...

1. Should we try to do this for Python as well? Not sure how this is handled now.
2. If 1 is affirmative, should we make this change in Array directly or keep it off as a SWIG limitation

CHANGELOG.md (choose one)

• no need to update because fixing unintended change not user facing yet

---

This change is 

[aymanhab]

@adamkewley for your feedback, thank you

[adamkewley]

If this is because SWIG doesn't emit them in the absence of a copy constructor in C++, it might be easier to write Array(const Array&) = default; in Array.h to tell the compiler that it has to emit them?

Maybe rule of 5 it:

• Array(Array const&) = default;;
• Array(Array&&) noexcept = default;
• Array& operator=(Array const&) = default;
• Array& operator=(Array&&) noexcept = default;
• ~Array() noexcept = default

To ensure that all possible ctor/assignments are covered

(note: the C++ compiler will never emit template methods in the binary, but i'm guessing this relates to SWIG's parser)

[aymanhab]

Thanks @adamkewley, that also works and was my other question above. Since you seem to agree with that I'll undo the swig change and implement in Array.h. I also added a Java test case to exercise so if the test pass then we're good. Thank you

[aymanhab]

Done, and tests pass so ready to go.

[adamkewley]

Great! Sorry for the hiccup, I'll try to keep the behavior in mind in the future!

[aymanhab]

Thanks much @adamkewley much appreciated, and no worries at all 💯