Only use current fix examples for tests #633

[TobiasNx]

See: #633

Not sure how to adjust the file:

https://github.com/metafacture/metafacture-core/blob/9ed5d4e8844998f899031861646a8384fc191c15/metafix/src/test/java/org/metafacture/metafix/MetafixBindTest.java

Also here is already an example of boolean operators for conditionals that are not supported yet:
https://github.com/metafacture/metafacture-core/blame/39ce013e731250b20698b61455733fadd80ae651/misc/vim/syntax/syntax-sample.fix#L15-L24 See #610

[blackwinter]

• MetafixBindTest.java: I couldn't find anything that needed changing, except for the @MetafixToDos. What do you have in mind?
• syntax-sample.fix: This one is tricky. It's used specifically to test the Vim plugin, which already has support for some syntax that isn't implemented yet. I don't think we should bother with it too much. As soon as the Metafix grammar is rewritten, I'll see to it that we get boolean operators :)

[TobiasNx]

• MetafixBindTest.java: I couldn't find anything that needed changing, except for the @MetafixToDos. What do you have in mind?

Yes, the TODOs which are "won't fix"s since the morph collectors and other morph concepts are not reused in Fix anymore. Should we just delete these tests? @blackwinter

* `syntax-sample.fix`: This one is tricky. It's used specifically to test the Vim plugin, which already has support for some syntax that isn't implemented yet. I don't think we should bother with it too much. As soon as the Metafix grammar is rewritten, I'll see to it that we get boolean operators :)

Fine with me

[blackwinter]

the TODOs which are "won't fix"s since the morph collectors and other morph concepts are not reused in Fix anymore.

Is that so? When was it decided?

Should we just delete these tests?

You should probably ask @fsteeg, since he's the one who introduced them (in f328a13).

If, in fact, we no longer plan to implement those, we should definitely delete the tests IMO.

[fsteeg]

the TODOs which are "won't fix"s since the morph collectors and other morph concepts are not reused in Fix anymore.

Yes, makes sense, these are probably leftovers from the initial, morph-based approach to implementing fix.

[TobiasNx]

Another thing that I spotted is that the java tests still have set_ in them, instead of add_. Should we adjust this here or in #631 ? @blackwinter

[blackwinter]

I stand by my earlier statement:

all the tests should be left untouched in order to surface any actual behaviour changes once the set_* implementations are changed. Just add new ones where appropriate.

I don't see why it should be addressed here at all. It still is and will be valid syntax.