Bug 17497 - eliminate DeprecatedValueList
Summary: eliminate DeprecatedValueList
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Platform (show other bugs)
Version: 528+ (Nightly build)
Hardware: All OS X 10.5
: P2 Normal
Assignee: Darin Adler
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-02-23 00:59 PST by Darin Adler
Modified: 2008-12-13 14:59 PST (History)
2 users (show)

See Also:

Attachments
patch to do this for SegmentedString (7.67 KB, patch)
2008-02-23 01:05 PST, Darin Adler 		no flags Details | Formatted Diff | Diff
patch (60.17 KB, patch)
2008-12-13 13:16 PST, Darin Adler 		andersca: review+
Details | Formatted Diff | Diff
Show Obsolete (1) View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Darin Adler 2008-02-23 00:59:23 PST 
Need to get rid of this class.
Comment 1 Darin Adler 2008-02-23 01:05:23 PST 
Created attachment 19290 [details]
patch to do this for SegmentedString
Comment 2 Sam Weinig 2008-02-23 14:26:44 PST 
Comment on attachment 19290 [details]
patch to do this for SegmentedString

r=me, though I assume this requires the changes the to Deque first.
Comment 3 Darin Adler 2008-02-23 23:39:54 PST 
Comment on attachment 19290 [details]
patch to do this for SegmentedString

Committed revision 30546.
Comment 4 Darin Adler 2008-02-24 23:02:21 PST 
These are the three remaining clients:

    1) CSSMutableStyleDeclaration::m_values -- can be a Vector, I think
    2) Document::m_windowEventListeners -- Vector? ListHashSet?
    3) EventTargetNode::m_regdListeners  -- Vector? ListHashSet?
Comment 5 Anders Carlsson 2008-02-25 17:05:27 PST 
I think that CSSMutableStyleDeclaration might be a great candidate for ListHashMap, something like

ListHashMap<int, CSSProperty>

(the item accessor will be O(n), but that is already the case)
Comment 6 Darin Adler 2008-12-13 13:16:25 PST 
Created attachment 26003 [details]
patch
Comment 7 Anders Carlsson 2008-12-13 13:53:50 PST 
Comment on attachment 26003 [details]
patch

>  #endif
> Index: WebCore/dom/EventTargetNode.cpp
> ===================================================================
> --- WebCore/dom/EventTargetNode.cpp	(revision 39276)
> +++ WebCore/dom/EventTargetNode.cpp	(working copy)
> +const RegisteredEventListeners& EventTargetNode::eventListeners() const
> +{
> +    if (hasRareData()) {
> +        if (RegisteredEventListeners* listeners = rareData()->listeners())
> +            return *listeners;
> +    }
> +    static const RegisteredEventListeners* emptyListenersVector = new RegisteredEventListeners;
> +    return *emptyListenersVector;

Clever!

>  void EventTargetNode::removeEventListener(const AtomicString& eventType, EventListener* listener, bool useCapture)
>  {
> -    if (!m_regdListeners)
> +    if (!hasRareData())
>          return;
>  
> -    RegisteredEventListenerList::Iterator end = m_regdListeners->end();
> -    for (RegisteredEventListenerList::Iterator it = m_regdListeners->begin(); it != end; ++it) {
> -        RegisteredEventListener& r = **it;
> +    RegisteredEventListeners* listeners = rareData()->listeners();
> +    if (!listeners)
> +        return;
> +
> +    size_t size = listeners->size();
> +    for (size_t i = 0; i < size; ++i) {
> +        RegisteredEventListener& r = *(*listeners)[i];

I prefer listeners->at(i).

>  void EventTargetNode::removeInlineEventListenerForType(const AtomicString& eventType)
>  {
> +
> +    size_t size = listeners->size();
> +    for (size_t i = 0; i < size; ++i) {
> +        RegisteredEventListener& r = *(*listeners)[i];
> +        if (r.eventType() != eventType || !r.listener()->isInline())
>              continue;
>

Same thing here.
  
> +
> +typedef Vector<RefPtr<RegisteredEventListener> > RegisteredEventListeners;
>

I think RegistereEventListenerVector is better here.
  
r=me!

I usually delete the files in a separate commit. Boy it feels great to see those red patches on webkit-changes ;)