feat: add fixer for no-dispatch-in-effects

[rafaelss95]

I went with a suggest instead of fix in this case because although it does not produce invalid code, it can cause some unexpected behavior if for example someone sets dispatch: false and still calls store$.dispatch manually (as in the test), there may be breaks in the expected flow.

[timdeschryver]

I don't think we can fix this automatically, the test only has a simple case but the store.dispatch could be nested, or it could even be called multiple times.

[rafaelss95]

Yep, that's why I implemented it as a suggestion. I find suggestions really useful, because you hover to see the problem and check if the suggestion solves the problem properly at the same time. Let me know your thoughts on this.

[timdeschryver]

I'm going to think on this a while longer.
While I see why we want to do it, I'm worried peoples will just use the suggestion and expect things to work.

[DaSchTour]

I suspect, that in many cases people use dispatch inside an effect in very complex cases. For example as part of a longer procedure in the middle of the pipe and not at the end. So in many cases this probably will lead to errors.
Just my two cents.

[timdeschryver]

I'm going to close this, since I'm still not convinced.
Thanks for the effort you've put int this @rafaelss95

[rafaelss95]

No problem, Tim! To be honest, I no longer remembered that this PR was still open 😄