347496 - Dynamic changes to MathML table lines don't work

Status: RESOLVED FIXED

(Core :: MathML, defect)

Description

[Jesse Ruderman]

Comment 1

[Jesse Ruderman]

Attachment: testcase

Comment 2

[rbs]

The fix will follow-up from bug 179619.

Comment 3

[rbs]

Attachment: extended testcase

Comment 4

[rbs]

Attachment: extended testcase

Comment 5

[rbs]

Attachment: patch

Follow-up to bug 179619, where AttributeChanged was hooked on mtr and mtd.

Here, 
- AttributeChanged is hooked on mtable itself. Plus some cosmetics:
- remove the utopic loop over the rowgroups to face reality.
- rename aFlagsToUpdate to aWhichFlags.
- rename a couple of bit flags in mPresentationData.flags.
- map 'false' to '0' in FindAttrValueIn to ease the mapping in the brain when reading that code.

Comment 6

[Robert O'Callahan (:roc) (email my personal email if necessary)]

+  static nsIContent::AttrValuesArray strings[] =
+    {&nsMathMLAtoms::_false, &nsMathMLAtoms::_true, nsnull};
+  switch (mContent->FindAttrValueIn(kNameSpaceID_None, nsMathMLAtoms::displaystyle_,
+                                    strings, eCaseMatters)) {
+    case 0:
+      mPresentationData.flags &= ~NS_MATHML_DISPLAYSTYLE;
+      mPresentationData.flags |= NS_MATHML_EXPLICIT_DISPLAYSTYLE;
+      break;
+    case 1:
+      mPresentationData.flags |= NS_MATHML_DISPLAYSTYLE;
+      mPresentationData.flags |= NS_MATHML_EXPLICIT_DISPLAYSTYLE;
+      break;

This code could be shared somehow, right?

Otherwise looks good.

Comment 7

[rbs]

Attachment: patch - iteration to share the indicated snippet

Comment 8

[rbs]

Checked in.

Comment 9

[Jesse Ruderman]

I checked in a reftest based on the original testcase.  Should rbs's extended testcase also be a series of reftests?  (Note that the "refdyn" tool is great for finding bugs in dynamic rendering, as long as there is a reftest or crashtest file that covers the features in question.)