Closed Bug 170413 Opened 22 years ago Closed 22 years ago

User-defined elements are all created as children of the <head>

Categories

(Core :: DOM: HTML Parser, defect)

Product:
Core
Component:
DOM: HTML Parser
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED DUPLICATE of bug 167493

People

(Reporter: rbs, Assigned: jesup)

References

Details

(Keywords: regression) 

This is a regression from bug 167493.

This patch used for that bug had some issues. I traced what goes on when a
user-defined tag is outside the <HEAD> and saw something strange happening.
Suppose <MyElement> is inside the <BODY> instead of the <HEAD>. The code will
roll into HandleStartToken() and do what I illustrate below (traced in the
debugger):

nsresult CNavDTD::HandleStartToken(CToken* aToken) {
[...]
//XXXrbs
const nsAString& tag = ((CStartToken*)aToken)->GetStringValue();
if (tag.Equals(NS_LITERAL_STRING("myElement")))
  NS_BREAK();

      PRBool theExclusive=PR_FALSE;
      theHeadIsParent=nsHTMLElement::IsChildOfHead(theChildTag,theExclusive);
>>>with the patch, this *always* returns |theHeadIsParent=true| for a
>>>user-defined tag
     
[...]

#if 0
        //we'll move to this approach after beta...
      if(!isTokenHandled) {

        if(theHeadIsParent && (theExclusive || mOpenHeadCount>0)) {
          result=AddHeadLeaf(theNode);
        }
        else result=HandleDefaultStartToken(aToken,theChildTag,theNode); 
      }
#else
        //the old way...
      if(!isTokenHandled) {
        if(theHeadIsParent || 
          (mOpenHeadCount>0  && (eHTMLTag_newline==theChildTag ||
eHTMLTag_whitespace==theChildTag))) {
            result=AddHeadLeaf(theNode);

>>>since |theHeadIsParent| is true, the smippet above
>>>will *always* add userdefined as a _leaf_ of the head

        }
        else result=HandleDefaultStartToken(aToken,theChildTag,theNode); 
      }
#endif
      //now do any post processing necessary on the tag...
      if(NS_OK==result)
        DidHandleStartTag(*theNode,theChildTag);
    }//if
  } //if

[...]
}
-> re-assigning to rjesup who handled bug 167493.
Assignee: harishd → rjesup
Depends on: 167493
Keywords: regression
Duping to bug 167493; all the discussion is there.  I'm going to back the patch
out until we can figure out a solution.

*** This bug has been marked as a duplicate of 167493 ***
Status: NEW → RESOLVED
Closed: 22 years ago
Resolution: --- → DUPLICATE
*** Bug 171324 has been marked as a duplicate of this bug. ***
You need to log in before you can comment on or make changes to this bug.