Bug 200277 – AX: com.apple.WebKit.WebContent at com.apple.WebKit: -[WKAccessibilityWebPageObject accessibilityParameterizedAttributeNames]

• Component: Accessibility
• Status: RESOLVED FIXED

[chris fleizach]

6 WebKit: -[WKAccessibilityWebPageObject accessibilityParameterizedAttributeNames] <==
        6 AppKit: __NSAccessibilityEntryPointParameterizedAttributeNames_block_invoke.611.llvm.16512287332241650554
          6 AppKit: NSAccessibilityPerformEntryPointObject.llvm.16512287332241650554
            6 AppKit: NSAccessibilityEntryPointParameterizedAttributeNames
              6 AppKit: CopyParameterizedAttributeNames
                6 HIServices: _AXXMIGCopyParameterizedAttributeNames

[chris fleizach]

<rdar://problem/49475009>

[chris fleizach]

Created attachment 375166 [details]
patch

[Per Arne Vollan]

Comment on attachment 375166 [details]
patch

View in context: https://bugs.webkit.org/attachment.cgi?id=375166&action=review

R=me.

> Source/WebKit/WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:102
> +        if (auto corePage = m_page->corePage()) {

I think there always is a core page unless we're in the destructor of WebPage, so this check is possibly not needed.

[chris fleizach]

Comment on attachment 375166 [details]
patch

View in context: https://bugs.webkit.org/attachment.cgi?id=375166&action=review

>> Source/WebKit/WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:102
>> +        if (auto corePage = m_page->corePage()) {
> 
> I think there always is a core page unless we're in the destructor of WebPage, so this check is possibly not needed.

do you want me to remove, or should I leave for safety?
or is there a reference style method I should use instead of getting the pointer

[chris fleizach]

Comment on attachment 375166 [details]
patch

View in context: https://bugs.webkit.org/attachment.cgi?id=375166&action=review

>>> Source/WebKit/WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:102
>>> +        if (auto corePage = m_page->corePage()) {
>> 
>> I think there always is a core page unless we're in the destructor of WebPage, so this check is possibly not needed.
> 
> do you want me to remove, or should I leave for safety?
> or is there a reference style method I should use instead of getting the pointer

leaving in for safety for now.

[Per Arne Vollan]

(In reply to chris fleizach from comment #5)
> Comment on attachment 375166 [details]
> patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=375166&action=review
> 
> >>> Source/WebKit/WebProcess/WebPage/mac/WKAccessibilityWebPageObjectMac.mm:102
> >>> +        if (auto corePage = m_page->corePage()) {
> >> 
> >> I think there always is a core page unless we're in the destructor of WebPage, so this check is possibly not needed.
> > 
> > do you want me to remove, or should I leave for safety?
> > or is there a reference style method I should use instead of getting the pointer
> 
> leaving in for safety for now.

After closing the WebPage, the core page will be null. If this code is not reachable after closing, the null check could be removed, but it does not seem trivial to verify that.

[WebKit Commit Bot]

Comment on attachment 375166 [details]
patch

Clearing flags on attachment: 375166

Committed r248038: <https://trac.webkit.org/changeset/248038>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.