I'm sorry for my mistake.

The correct url is: http://cloudplus.me/EasyCoding/

#13572 is a duplicate of this ticket.

Confirm that this is a "regression", but jQuery 1.9.x was wrong.

The constructed object _IS_ a plain object, because the code is paving over the constructor's prototype:

function C() {}
C.prototype = {};
c = new C();
console.log(c.constructor);
// > function Object() { [native code] }

If the program resets the constructor, the expected results:

function C() {}
C.prototype = {
  constructor: C
};
c = new C();
console.log(c.constructor);
// > function C() {}

In jQuery.isPlainObject's case, the OPs example code only works in jQuery 1.9.x because it has properties on the object that was used as the prototype, which is what jQuery 1.9.x used to determine property ownership and therefore detect possible inheritance chains.

Everything works the way you want it to until you try to test this:

var c, extended;
function C() {}
C.prototype = {};
c = new C();

extended = jQuery.extend( true, {}, {
  target: c
});

strictEqual( extended.target, c, "Deep extended objects retain identity" );

... Without any properties on that prototype, the jQuery 1.9 way will fail as well: http://jsfiddle.net/rwaldron/7Jg8Q/

Unfortunately, the regression stands and I'll have to restore the behaviour, but it will still be broken (ie. it's correct, but the expectations of the calling code is wrong).

Fixes #13571. jQuery.isPlainObject 1.9.x compatibility

Signed-off-by: Rick Waldron <waldron.rick@gmail.com>

Changeset: 8d1c42296fefc4e79189b6c2d78a8a78fdd9576d

We've decided that this breaking change will remain in 2.x, but will be accompanied by documentation explaining the issue encountered here.

Reverted: https://github.com/jquery/jquery/commit/5c82d36f194a854a19f81bebd3ed1a6e7559c811

rwaldron

I can accept your decision easily.

wow, thanks for being so reasonable :) that's refreshing... you just made my night

I know that less code is fast code. Thanks for your good decision.

Take care for the same kind of claim for "isPlainObject" not to change again (and you try to revert again).

@timmywil, why is this reopened? (I'm just curious)

@rwaldron: Wait, I'm confused now. You reopened it. I just put it back to assigned.

Yep, you're right. Oops!

#13715 is a duplicate of this ticket.

Why the check:

!core_hasOwn.call(obj, "constructor")

is needed in jQuery 1.x anyway? All unit tests are passed by oldIE when this line is missing.

If it's really needed, a unit test for that should be written.

#14227 is a duplicate of this ticket.