Upgrade to flow 0.72 and tweak scripts

[TrySound]

Ref #712

$FlowIssue is not used
$FlowFixMe exists out of the box
flow check instead of flow won't start a server
Tweaked eslint to make it work with types and class properties

[TrySound]

@kof I added explicit check here. You can try to test it. I don't know code base and not sure what is the problem here.

[kof]

I am wondering if its a good thing to have data optional, why would someone call an update without data?

[TrySound]

Yep. These tests are failed if we do not allow to pass undefined.

FAILED TESTS:
  jss-default-unit
    support function values
      ‚úñ should add default unit
        Chrome 66.0.3359 (Mac OS X 10.12.6)
      Error: expected '' to equal '.a-id {\n  width: 1px;\n}'
          at Assertion.assert (webpack:///~/expect.js/index.js:96:0 <- tests/index.js:8804:13)
          at Assertion.be.Assertion.equal (webpack:///~/expect.js/index.js:216:0 <- tests/index.js:8924:10)
          at Assertion.(anonymous function) [as be] (webpack:///~/expect.js/index.js:69:0 <- tests/index.js:8777:24)
          at Context.<anonymous> (webpack:///~/jss-default-unit/tests/index.test.js:352:0 <- tests/index.js:20470:53)

  jss-nested
    function values
      ‚úñ should generate correct CSS
        Chrome 66.0.3359 (Mac OS X 10.12.6)
      Error: expected '' to equal '.a-id:hover {\n  color: red;\n}'
          at Assertion.assert (webpack:///~/expect.js/index.js:96:0 <- tests/index.js:8804:13)
          at Assertion.be.Assertion.equal (webpack:///~/expect.js/index.js:216:0 <- tests/index.js:8924:10)
          at Assertion.(anonymous function) [as be] (webpack:///~/expect.js/index.js:69:0 <- tests/index.js:8777:24)
          at Context.<anonymous> (webpack:///~/jss-nested/tests/index.test.js:470:0 <- tests/index.js:22087:45)

  jss-camel-case
    function values
      ‚úñ should generate correct CSS
        Chrome 66.0.3359 (Mac OS X 10.12.6)
      Error: expected '' to equal '.a-id {\n  font-size: 12;\n}'
          at Assertion.assert (webpack:///~/expect.js/index.js:96:0 <- tests/index.js:8804:13)
          at Assertion.be.Assertion.equal (webpack:///~/expect.js/index.js:216:0 <- tests/index.js:8924:10)
          at Assertion.(anonymous function) [as be] (webpack:///~/expect.js/index.js:69:0 <- tests/index.js:8777:24)
          at Context.<anonymous> (webpack:///~/jss-camel-case/tests/index.test.js:144:0 <- tests/index.js:19716:53)

  jss-vendor-prefixer
    prefix function values
      ‚úñ should generate correct CSS
        Chrome 66.0.3359 (Mac OS X 10.12.6)
      Error: expected '' to equal '.a-id {\n  display: flex;\n}'
          at Assertion.assert (webpack:///~/expect.js/index.js:96:0 <- tests/index.js:8804:13)
          at Assertion.be.Assertion.equal (webpack:///~/expect.js/index.js:216:0 <- tests/index.js:8924:10)
          at Assertion.(anonymous function) [as be] (webpack:///~/expect.js/index.js:69:0 <- tests/index.js:8777:24)
          at Context.<anonymous> (webpack:///~/jss-vendor-prefixer/tests/index.test.js:153:0 <- tests/index.js:22568:53)

  Integration: rules
    rule.toJSON()
      ‚úñ should handle function values
        Chrome 66.0.3359 (Mac OS X 10.12.6)
      Error: expected {} to sort of equal { color: 'red' }
          at Assertion.assert (webpack:///~/expect.js/index.js:96:0 <- tests/index.js:8804:13)
          at Assertion.eql (webpack:///~/expect.js/index.js:230:0 <- tests/index.js:8938:10)
          at Context.<anonymous> (webpack:///tests/integration/rules.js:438:45 <- tests/index.js:16268:56)

[kof]

It is optional here, because we have arguments an optional first argument, which results in potentially optional second argument (now I think it was a bad idea to make optional first argument, given the flow typing)

[kof]

Is there a way to express this?

if name was passed, data is not optional, if only name is passed and its an object - it is the data object.

[TrySound]

It will look really unmaintainable. Can't we just use named arguments?

[TrySound]

I just played with it. Seems like it's not possible to express with class method.

[kof]

Reversing the order would fix it too, but we would need to release a major version. Lets combine this with minorepo, new build and mb some other big changes and do it!

…

On Sat, May 19, 2018, 19:11 Bogdan Chadkin ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In src/RuleList.js <#716 (comment)>: > @@ -136,13 +129,14 @@ export default class RuleList { */ update(name?: string | Object, data?: Object): void { It will look really unmaintainable. Can't we just use named arguments? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#716 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADOWBDvKLjNnkE9QOAf4EvSP9zUTxxnks5t0FJUgaJpZM4T95bK> .

[TrySound]

Monorepo?)

[kof]

Phone typo))) Not touching computer today.

…

On Sat, May 19, 2018, 21:20 Bogdan Chadkin ***@***.***> wrote: Monorepo?) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#716 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADOWNmtCtGR4n-oYoDsZeWVHyGLHEfmks5t0HCJgaJpZM4T95bK> .

[TrySound]

@kof I achieved update types with istarkov suggestion. So can we merge then?

[kof]

Perfect, done!