Fix UN flag test by returning the correct value.

[peterwilsoncc]

In #40616 the method for testing flags was changed to allow for changes to the canvas engines in browsers, the test was reversed.

If the two sets of characters render the same, the flag test has failed and the polyfill should be used. WP is doing the reverse.

I'm not really following. Could you provide some steps to replicate the issue and why it doesn't apply to the England flag?

Replying to ocean90:

I'm not really following. Could you provide some steps to replicate the issue and why it doesn't apply to the England flag?

The tests are considered to fail (return false) if both flags look the same.

The UN test is effectively returning flag === flag2, ie true if identical.

The English flag returns flag !== flag2, ie false if identical.

The way the tests are written means the affect is minimal as users are more likely to get a false negative than a false positive.

I've put together a reduced test case at ​https://jsbin.com/palagay/2/edit?js,output

1. Visit the ​test case to see if your browser supports the UN flag (Chrome 61/MacOS Sierra will not)
2. The outcome will display the images.
3. To test in core, change the UN flag test to end:

if ( flag !== flag2 ) {
        console.log( 'fail' );
        return false;
}
console.log( 'pass' );

Thanks, makes sense now.

In 41386:

Emoji: Invert value returned by UN test for accuracy.

The JavaScript testing for UN flag support was passing when it failed, and failing when it passes.

This inverts the result to take more traditional approach of returning the correct result.

Unprops Pento.
Fixes #41852.

Reopening for backporting to 4.8 branch.

In 41396:

Emoji: Invert value returned by UN test for accuracy.

The JavaScript testing for UN flag support was passing when it failed, and failing when it passes.

This inverts the result to take more traditional approach of returning the correct result.

Merge of [41386] to the 4.8 branch.

Unprops Pento.
Fixes #41852.