.attr() returns -1 on missing attribute instead of undefined

[anonymous]

When element(s) are missing the attribute, attr works as follows.

$('elem').attr('noattr') == -1

This is a change from 1.4 where .attr() returned "undefined" or some other "falsy" response. Since -1 isn't considered falsy by javascript the following code is now broken.

if ($('elem').attr('noattr')) { doStuff();}
else { doOtherStuff();}

If the intended attribute check is now -1 please update documentation so we can update code, or fix the regression that makes this fail a boolean check.

[timmywil]

Please provide a test case on ​http://jsfiddle.net. Non-existing attributes should return undefined. ​http://jsfiddle.net/timmywil/2Ymea/

[anonymous]

i think i got this one right.

​http://jsfiddle.net/uyLQL/6/

I'm loading jquery in the head, but the code that runs is in a document ready function.

[anonymous]

Replying to anonymous:

i think i got this one right.

Meaning i've never used JSfiddle before, so i'm not 100% sure i replicated the test here correctly.

[timmywil]

Nice catch! You are correct and I never noticed that before.

Nevertheless, this has been fixed in the upcoming attrHooks updates for jQuery 1.6. ​http://jsfiddle.net/timmywil/uyLQL/9/

[anonymous]

Replying to timmywil:

Nice catch! You are correct and I never noticed that before.

Nevertheless, this has been fixed in the upcoming attrHooks updates for jQuery 1.6. ​http://jsfiddle.net/timmywil/uyLQL/9/

As long as I know it's fixed later on, I suppose I can wait. :P Back to vanilla dom for now i guess.

[timmywil]

If it helps, the way we retrieve maxLength in 1.6 is by using node.getAttribute('maxLength') rather than the property node.maxLength. Perhaps we could put a quick update for maxlength in 1.5.2, but since attr is getting a rewrite, my guess is that it's not the highest priority.

An example with raw js: ​http://jsfiddle.net/timmywil/uyLQL/10/

[rwaldron]

​http://bugs.jquery.com/ticket/8457