552044 - (mathcolor-background) [MathML 3] Add support for mathcolor/mathbackground on any presentation MathML element

Status: RESOLVED FIXED

(Core :: MathML, defect)

Description

[Frédéric Wang (:fredw)]

In MathML 2, mathcolor/mathbackground were valid on most token elements and <mstyle/>:

http://www.w3.org/TR/MathML2/chapter3.html#presm.commatt

In MathML 3, they are now valid on all presentation MathML elements:

http://www.w3.org/TR/MathML3/chapter3.html#presm.presatt

The MathML 3 testsuite includes several new tests for mathcolor and mathbackground:

http://www.w3.org/Math/testsuite/build/mathml3/overview.html

Comment 1

[Frédéric Wang (:fredw)]

This looks like a regression. 

Compare "or use color" in

http://www.mozilla.org/projects/mathml/demo/basics.xhtml

and

http://www.mozilla.org/projects/mathml/screenshots/basics.png

Comment 2

[Frédéric Wang (:fredw)]

Attachment: Patch

Comment 3

[Karl Tomlinson (:karlt)]

I landed this but backed it out due to failure of layout/reftests/bugs/393760-1.xml.

Some differences were expected changes in behavior, but iiuc <math> is not a presentation element and so the attributes should not have an effect there.

Comment 4

[Karl Tomlinson (:karlt)]

Some other strange issues only on the Linux reftests:

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1271823646.1271828181.15405.gz

REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/bugs/393760-1.xml |
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/mathml/mathcolor-1.mml | timed out waiting for onload to fire
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/mathml/mathcolor-2.mml | timed out waiting for onload to fire
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/mathml/mathcolor-3.mml | timed out waiting for onload to fire
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/mathml/mathcolor-4.mml | timed out waiting for onload to fire
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/mathml/mathbackground-1.mml | timed out waiting for onload to fire
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/mathml/mathbackground-2.mml | timed out waiting for onload to fire
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/mathml/mathbackground-3.mml | timed out waiting for onload to fire
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/mathml/mathbackground-4.mml | timed out waiting for onload to fire
REFTEST TEST-UNEXPECTED-FAIL | file:///builds/slave/mozilla-central-linux-debug-unittest-reftest/build/reftest/tests/layout/reftests/selection/pseudo-element-of-native-anonymous.html | (!=)
PROCESS-CRASH | automation.py | application crashed (minidump found)
Thread 1 (crashed)
PROCESS-CRASH | automation.py | application crashed (minidump found)
Thread 1 (crashed)
TEST-UNEXPECTED-FAIL | automationutils.processLeakLog() | leaked 786043 bytes during test execution

Don't know what's up there, but I might use the tryserver on the next patch.

Comment 5

[Frédéric Wang (:fredw)]

I'm going to ask the Math WG for these attributes on the <math/> element (as well as <semantics/>.

For the strange errors, I suspect this is because the reftests need bug 152001 and bug 124709 to work.

Comment 6

[Frédéric Wang (:fredw)]

> For the strange errors, I suspect this is because the reftests need bug 152001
> and bug 124709 to work.

There is a reftest for the support of application/mathml+xml but not for the extension *.mml (the system may map the extension to another mime type). I'll try to use *.xml file instead for the next patch.

Comment 7

[Karl Tomlinson (:karlt)]

Hmm.  My Gentoo machine has text/mathml mml in /etc/mime.types.
Though Firefox seems to be getting the mime type from a different system database.

Comment 8

[Frédéric Wang (:fredw)]

(In reply to comment #7)
> Hmm.  My Gentoo machine has text/mathml mml in /etc/mime.types.
> Though Firefox seems to be getting the mime type from a different system
> database.

Same for me. Actually if I remember well, the patch for 152001 didn't change anything: once bug 124709 was fixed, my system was able to open *.mml file. I don't know how the system determine the MIME type, but according to Boris *.mml may be mapped to another one.

Comment 9

[Frédéric Wang (:fredw)]

So finally, does the patch do the correct thing with mathcolor/mathbackground?
I simply need to modify layout/reftests/bugs/393760-1.xml and change extensions .mml, right?

Comment 10

[Karl Tomlinson (:karlt)]

Yes, I think so.

The draft has been modified so that the section for <mstyle> includes:
"As a presentation element, mstyle directly accepts the mathcolor and mathbackground  attributes. Thus, the mathbackground specifies the color to fill the bounding box of the mstyle element itself; it does not  specify the default background color. This is an incompatible change from MathML 2, but we feel it is more useful and intuitive. Since the default for mathcolor  is inherited, this is no change in its behaviour."

and the section for <math> includes:
"Also, it accepts the mathbackground attribute in the same sense as mstyle and other presentation elements to set the background color of the bounding box, rather than specifying a default for the attribute."

Comment 11

[Frédéric Wang (:fredw)]

Attachment: Patch v.2

Comment 12

[Frédéric Wang (:fredw)]

Attachment: Patch v.3

Update the patch to fix conflicts.

The errors above are no longer present in the results of the TryServer.
I've some new errors, but I'm not sure it is related to my patch since they seem already present in previous tryserver builds:

http://tinderbox.mozilla.org/showlog.cgi?log=MozillaTry/1273675112.1273689466.9393.gz

Comment 13

[Karl Tomlinson (:karlt)]

Yes, those new errors look unrelated.

Comment 14

[Karl Tomlinson (:karlt)]

http://hg.mozilla.org/mozilla-central/rev/5f53a3240101