<html>
<head>
<base href="https://bugs.webkit.org/" />
</head>
<body>
<p>
<div>
<b><a class="bz_bug_link
bz_status_NEW "
title="NEW - Web Inspector: Show upcoming values of non-native iterators"
href="https://bugs.webkit.org/show_bug.cgi?id=152682#c5">Comment # 5</a>
on <a class="bz_bug_link
bz_status_NEW "
title="NEW - Web Inspector: Show upcoming values of non-native iterators"
href="https://bugs.webkit.org/show_bug.cgi?id=152682">bug 152682</a>
from <span class="vcard"><a class="email" href="mailto:joepeck@webkit.org" title="Joseph Pecoraro <joepeck@webkit.org>"> <span class="fn">Joseph Pecoraro</span></a>
</span></b>
<pre><span class="quote">> If they use iterators, why are they going to have any less trouble
> duplicating the iterator state than we are?</span >
Our native iterators have a wrapper and an internal object. When we peek, we create a new wrapper, with the same data as the original wrapper, advance the new wrapper, then drop it on the floor. See JSMapIterator::clone, JSSetIterator::clone, etc.
I don't think we can make the same assumptions about any user script's object. And I'm not sure we can call execute _anything_ on the user object in a way that will guarantee no side-effects, as anything can be a getter.
The cop-out, which is what I'm suggesting, is a public function with a contract of "don't do side-effects, if you do it is your fault". The inspector can then call these. Problem is it requires libraries to author those functions.</pre>
</div>
</p>
<hr>
<span>You are receiving this mail because:</span>
<ul>
<li>You are the assignee for the bug.</li>
</ul>
</body>
</html>