update docs to note future deprecation of .simulate

[browne0]

This PR updates the doc to explain that simulate doesn't work as intended. I included a link to the original issue that I opened on Friday. #2173

[browne0]

@ljharb Could I get a review on this? Thanks.

[azangru]

Don't know if it's too much to ask, but I was wondering whether you might include an example of the officially blessed way of testing components' behavior in response to DOM events.

I am particularly curious about DOM-mounted components. It makes total sense to test shallow-rendered components by directly calling functions that they pass to their children as props, but for a fully mounted component, you might want to test it via its interactions with the DOM.

[ughitsaaron]

Would it be considered out of scope/desirable to add a warning output to the console when .simulate is called? Happy to make a PR to follow up.

[ljharb]

@browne0 sorry for the long delay; i'll review this soon.

@azangru that would require a full simulation interface, which I want too, but is way too massive a project to be developed as part of enzyme. It's not as simple as clicks; you'd want to simulate things like "type foobar into the text field, hit tab, type baz, use the mouse to click a checkbox and then select from a dropdown, and then click into another text field and type quux, and then hit enter" - which involves a very large number of events, precisely timed and ordered, with a bunch of faked event objects.

@ughitsaaron no, I'm not interested in spamming warnings on every usage of the method, i'd rather just remove it in v4; if we used react's caching approach to only show the message once, it'll be ignored by humans anyways.