ngbDropdownToggle double triggers on child element

[karptonite]

Bug description:

When you have the ngbDropdownToggle on a button element, and a span inside that button element, clicking on the span (inside the button) does not toggle the dropdown (or, more specifically, I think it double-toggles it back to its original state).

This bug is new as of alpha 6--it may be related to the change from HostListener to (click) for handling the event. The bug was observed using chrome version 53.

Link to minimally-working plunker that reproduces the issue:

http://plnkr.co/edit/TyaAkbqHeLq7HidFzpNf?p=preview

Version of Angular, ng-bootstrap, and Bootstrap:

Angular: 2.0.0? whatever the sample plunker defaults to--it does not specify

ng-bootstrap: 1.0.0-alpha.6

Bootstrap: 4.0.0-alpha.4

[pkozlowski-opensource]

@karptonite thnx for reporting, this is indeed a bug on our side (luckily with an easy fix!). What needs changing is this line to listen for child elements as well (this._toggleElement.contains).

A new test should be added, similar to the existing one:

ng-bootstrap/src/dropdown/dropdown.spec.ts

Lines 167 to 191 in a173e40

	it('should toggle dropdown on click', () => {
	const html = `
	<div ngbDropdown>
	<button ngbDropdownToggle>Toggle dropdown</button>
	</div>`;
	const fixture = createTestComponent(html);
	const compiled = fixture.nativeElement;
	let dropdownEl = getDropdownEl(compiled);
	let buttonEl = compiled.querySelector('button');
	expect(dropdownEl).not.toHaveCssClass('open');
	expect(buttonEl.getAttribute('aria-haspopup')).toBe('true');
	expect(buttonEl.getAttribute('aria-expanded')).toBe('false');
	buttonEl.click();
	fixture.detectChanges();
	expect(dropdownEl).toHaveCssClass('open');
	expect(buttonEl.getAttribute('aria-expanded')).toBe('true');
	buttonEl.click();
	fixture.detectChanges();
	expect(dropdownEl).not.toHaveCssClass('open');
	expect(buttonEl.getAttribute('aria-expanded')).toBe('false');
	});

PRs welcomed!

[JNKielmann]

Is anyone working on this? Otherwise I would try to make a PR.

[pkozlowski-opensource]

@JNKielmann looks like we've got a PR for this one already #810