<html>
<head>
<base href="https://bugs.webkit.org/" />
</head>
<body>
<p>
<div>
<b><a class="bz_bug_link
bz_status_NEW "
title="NEW - [GTK] Get rid of DataObjectGtk::forClipboard and cleanup pasteboard code"
href="https://bugs.webkit.org/show_bug.cgi?id=161907#c3">Comment # 3</a>
on <a class="bz_bug_link
bz_status_NEW "
title="NEW - [GTK] Get rid of DataObjectGtk::forClipboard and cleanup pasteboard code"
href="https://bugs.webkit.org/show_bug.cgi?id=161907">bug 161907</a>
from <span class="vcard"><a class="email" href="mailto:cgarcia@igalia.com" title="Carlos Garcia Campos <cgarcia@igalia.com>"> <span class="fn">Carlos Garcia Campos</span></a>
</span></b>
<pre>(In reply to <a href="show_bug.cgi?id=161907#c2">comment #2</a>)
<span class="quote">> Comment on <span class=""><a href="attachment.cgi?id=288687&action=diff" name="attach_288687" title="Patch">attachment 288687</a> <a href="attachment.cgi?id=288687&action=edit" title="Patch">[details]</a></span>
> Patch
>
> View in context:
> <a href="https://bugs.webkit.org/attachment.cgi?id=288687&action=review">https://bugs.webkit.org/attachment.cgi?id=288687&action=review</a>
>
> > Source/WebCore/platform/gtk/PasteboardGtk.cpp:99
> > -DataObjectGtk* Pasteboard::dataObject() const
> > +const DataObjectGtk& Pasteboard::dataObject() const
> > {
> > - return m_dataObject.get();
> > + return *m_dataObject;
> > }
>
> You should also change m_dataObject to be a (non-const) reference, since
> it's always initialized in the constructor initializer lists and there are
> also ASSERTs to ensure it's not null. Then you can get rid of the ASSERTS,
> and don't have to dereference it throughout the file anymore.</span >
Yes, I tried that, but it's not that easy because of the way DragData works, I plan to do another cleanup after the clipboard is moved to the UI process, to not delay more that patch.
<span class="quote">> > Source/WebCore/platform/gtk/PasteboardHelper.cpp:311
> > + // When gtk_clipboard_set_with_data fails the callbacks are ignored, so we need to leak the data we were passing to clearClipboardContentsCallback.
> > + data.release();
>
> This confused me because I assumed this was on the failure path, because
> your comment starts with "when it fails," but in fact this is the success
> case. I see that you need to leak it only on success, so the code is right,
> just the comment is confusing.</span >
Agree, I modified the existing comment that was in the failure path, and it's not clear enough now, I'll rephrase it.
<span class="quote">> > Source/WebKit2/WebProcess/WebCoreSupport/gtk/WebDragClientGtk.cpp:69
> > + DragData dragData(const_cast<DataObjectGtk*>(&dataObject), clientPosition, globalPosition, dataTransfer.sourceOperation());
>
> Wouldn't it be better to provide a non-const accessor as well, to avoid the
> const_cast here?</span >
This is temporal, I'll get rid of all const_casts in the second cleanup.</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>