Replace 0 and NULL with nullptr in WebCore/dom.

• Component: WebCore Misc.
• Status: RESOLVED FIXED

[Hunseop Jeong]

Use the 'nullptr' instead of '0' and 'NULL' in WebCore/dom.

[Hunseop Jeong]

Created attachment 264548 [details]
Patch

[Hunseop Jeong]

Created attachment 264549 [details]
Patch

[Darin Adler]

Comment on attachment 264549 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=264549&action=review

Looks good.

> Source/WebCore/dom/Attr.cpp:49
> +    , m_element(nullptr)

This initialization of m_element to nullptr should be done in the header, as you did with m_ignoreChildrenChanged. It’s OK if other constructors initialize it to something else.

> Source/WebCore/dom/ContainerNode.cpp:67
> +ChildNodesLazySnapshot* ChildNodesLazySnapshot::latestSnapshot = nullptr;

This isn’t needed. A static data member like this will automatically be initialized to null without saying "= nullptr".

> Source/WebCore/dom/DocumentOrderedMap.h:-77
> -            : element(0)

I would have initialized count to zero where it’s defined below too, just like element.

> Source/WebCore/dom/Event.cpp:-53
> -    , m_currentTarget(0)

I would have initialized all those other data members where they are defined too, just like we are doing for m_currentTarget.

> Source/WebCore/dom/EventListenerMap.cpp:-201
> -    : m_map(0)

I would have initialized the other data members where they are defined too.

> Source/WebCore/dom/Node.h:707
> +        DataUnion() { }

I don’t think we need to write this constructor out explicitly. I think it’s what will automatically be generated.

> Source/WebCore/dom/PositionIterator.h:40
> -        : m_anchorNode(0)
> -        , m_nodeAfterPositionInAnchor(0)
> -        , m_offsetInAnchor(0)
> +        : m_offsetInAnchor(0)

I would have initialized m_offsetInAnchor where it's defined too.

> Source/WebCore/dom/ProcessingInstruction.cpp:-45
> -    , m_cachedSheet(0)

I would have initialized those other data members where they are defined too.

> Source/WebCore/dom/RangeBoundaryPoint.h:-70
> -    , m_childBeforeBoundary(0)

I would have initialized m_offsetInContainer where it's defined too.

> Source/WebCore/dom/ScriptElement.cpp:60
> +    , m_cachedScript(nullptr)

This isn’t needed. CachedResourceHandle already initializes itself to null without explicit initialization.

> Source/WebCore/dom/StyledElement.cpp:49
> +    PresentationAttributeCacheKey() { }

I think you can omit this entirely now that it’s empty. You should at least try it.

[Hunseop Jeong]

Created attachment 264670 [details]
Patch

[Hunseop Jeong]

(In reply to comment #3)
> Comment on attachment 264549 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=264549&action=review
> 
> Looks good.
> 
> > Source/WebCore/dom/Attr.cpp:49
> > +    , m_element(nullptr)
> 
> This initialization of m_element to nullptr should be done in the header, as
> you did with m_ignoreChildrenChanged. It’s OK if other constructors
> initialize it to something else.
Okay. I moved the constructor initializer to the header.
> 
> > Source/WebCore/dom/ContainerNode.cpp:67
> > +ChildNodesLazySnapshot* ChildNodesLazySnapshot::latestSnapshot = nullptr;
> 
> This isn’t needed. A static data member like this will automatically be
> initialized to null without saying "= nullptr".
Oops, I will be kept in mind.
> 
> > Source/WebCore/dom/DocumentOrderedMap.h:-77
> > -            : element(0)
> 
> I would have initialized count to zero where it’s defined below too, just
> like element.
I moved the initialized count where it's defined.
> 
> > Source/WebCore/dom/Event.cpp:-53
> > -    , m_currentTarget(0)
> 
> I would have initialized all those other data members where they are defined
> too, just like we are doing for m_currentTarget.
I moved the constructor iniailizers to they are defined.
I will keep working to change the codes that not only using the 'nullptr' but also moving the constructor initailizers to they are defined.
> 
> > Source/WebCore/dom/EventListenerMap.cpp:-201
> > -    : m_map(0)
> 
> I would have initialized the other data members where they are defined too.
Done. And I removed the default constructor because it is empty.
> 
> > Source/WebCore/dom/Node.h:707
> > +        DataUnion() { }
> 
> I don’t think we need to write this constructor out explicitly. I think it’s
> what will automatically be generated.
I deleted it.
> 
> > Source/WebCore/dom/PositionIterator.h:40
> > -        : m_anchorNode(0)
> > -        , m_nodeAfterPositionInAnchor(0)
> > -        , m_offsetInAnchor(0)
> > +        : m_offsetInAnchor(0)
> 
> I would have initialized m_offsetInAnchor where it's defined too.
Done. And I removed the default constructor of PositionIterator because it is empty.
> 
> > Source/WebCore/dom/ProcessingInstruction.cpp:-45
> > -    , m_cachedSheet(0)
> 
> I would have initialized those other data members where they are defined too.
Done.
> 
> > Source/WebCore/dom/RangeBoundaryPoint.h:-70
> > -    , m_childBeforeBoundary(0)
> 
> I would have initialized m_offsetInContainer where it's defined too.
Done.
> 
> > Source/WebCore/dom/ScriptElement.cpp:60
> > +    , m_cachedScript(nullptr)
> 
> This isn’t needed. CachedResourceHandle already initializes itself to null
> without explicit initialization.
I removed the explicit initailaization. And initialized other data members where they are defined.
> 
> > Source/WebCore/dom/StyledElement.cpp:49
> > +    PresentationAttributeCacheKey() { }
> 
> I think you can omit this entirely now that it’s empty. You should at least
> try it.
Done.

Could you check again?

[Darin Adler]

Comment on attachment 264670 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=264670&action=review

> Source/WebCore/dom/IdTargetObserverRegistry.h:43
> +    IdTargetObserverRegistry() { }

I think this can be removed.

[WebKit Commit Bot]

Comment on attachment 264670 [details]
Patch

Clearing flags on attachment: 264670

Committed r191955: <http://trac.webkit.org/changeset/191955>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.