jquery-migrate handling stray text different than jQuery

[dana-ross]

Researching a WordPress issue: http://core.trac.wordpress.org/ticket/23055

Server-side code for an AJAX call was adding a stray "0" character outside the root element of an HTML fragment. For example:

<li>example</li>0

(we've patched the code that was doing this)

When this HTML is passed to jQuery.fn.init, jQuery 1.8.3 and 1.9.1 both ignore the trailing 0 as it isn't matched by rquickExpr.

In jquery-migrate's version of jQuery.fn.init, the HTML is passed through jQuery.parseHTML before calling the original jQuery.fn.init(). The parseHTML method is returning a two-element jQuery collection with the LI element and the 0.

Since jQuery.fn.init is passed a collection, there's no need to compare against rquickExpr and the 0 never ends up getting stripped out.

Sure, it's bad HTML that never should have been generated in the first place, but I'm thinking jquery-migrate shouldn't be changing the behavior here.

[nacin]

Just a note, this doesn't affect a stable version of WordPress. But, it could affect plugins. I don't care all that much, but it would be nice if this could be fixed.