- From: Jim Allan <jimallan@tsbvi.edu>
- Date: Tue, 3 Aug 2010 15:56:23 -0500
- To: "'Doug Schepers'" <schepers@w3.org>, <w3c-wai-ua@w3.org>
- Cc: "'Jeanne Spellman'" <jeanne@w3.org>
Doug, Most helpful. And just the right amount of info. Jim > -----Original Message----- > From: Doug Schepers [mailto:schepers@w3.org] > Sent: Tuesday, August 03, 2010 3:50 PM > To: jimallan@tsbvi.edu > Cc: 'Jeanne Spellman' > Subject: Re: MATHML DOM question > > Hi, Jim, Jeanne- > > Jim Allan wrote (on 8/3/10 4:26 PM): > > See below > > > >> -----Original Message----- > >> From: Jeanne Spellman [mailto:jeanne@w3.org] > >> Sent: Tuesday, August 03, 2010 3:10 PM > >> To: Jim Allan; Doug Schepers > >> Subject: MATHML DOM question > >> > >> Doug meet Jim, Jim meet Doug, > >> > >> The User Agent Accessibility Working Group is meeting to write a > >> technical support document. A question came up about the DOM and > accessibility. > >> Here is Jim's question (as paraphrased by me, you may get more info > >> from Jim directly). > >> > >> With MathML: Now that MathML is rendered natively in say, Firefox, > >> does it still have two DOMs - one for HTML processing and one for > >> MATHML processing? > > > > And, if there are 2 separate DOMs does the user agent expose the > > MATHML DOM to the accessibility API? > > > Short answer: there is a single DOM, and it should be exposed to the > accessibility API. > > > Longer, more technical answer: > > First, you have to break the idea of a DOM into 2 aspects: > 1) the memory representation > 2) the interfaces defined for each element (and attribute), i.e. methods > and properties > > MathML is an easy case (as far as compound documents go), because as far > I know, they don't describe any special interfaces specifically for > MathML; it just uses the regular DOM, with few if any extensions. > > And you also have to look at two usage scenarios: > a) inline content: a single document containing a mixture of HTML, > MathML, SVG, etc. > b) referenced content: multiple documents integrated via referencing > elements (such as <img>, <object>, <embed>, etc.) > > Let's assume you're talking about (b), since that is the most natural > way for an author to include it. This means a single document with both > HTML and MathML (see snippet [1]... note that it could also be > SVG+MathML, the principle is the same). That document would result in a > single memory representation; that is, the user agent would have a > single 'document' object with a bunch of child nodes, like a 'body', > 'p', 'mfrac', etc., from different namespaces (languages), but all > readable from memory in the same way, and all as children of the same > document. If MathML were to have special methods and properties, this > document would have all the shared interfaces, and each element (child > node) in it would have the element-specific methods, as well. > > For standalone MathML documents (not mixed with HTML or SVG), of course, > the memory representation would just be the MathML document, and would > only have those interfaces. > > > Okay, all that said, an accessibility API would have to know how to read > MathML in order to take advantage of this fact. And an AT really should > have special ways of presenting MathML that are more helpful to e.g. the > visually impaired than treating it as simple text; it should allow the > user to explore the equation or math snippet (for myself, simply hearing > or reading the text string in sequence is rarely good enough, and I'm > sighted). I have no idea what the state of AT support for MathML is. > > > I hope that answers your question. Let me know if you need any more (or > less). > > > [1] An inline HTML+MathML snippet: > <html><body> > <p>The solution for quadratic equations is > <math> > <mi>x</mi> > <mo>=</mo> > <mfrac> > <mrow> > <mo>−</mo> > <mi>b</mi> > > <mo>±</mo> > <msqrt> > <msup> > <mi>b</mi> > <mn>2</mn> > </msup> > <mo>−</mo> > <mn>4</mn> > > <mo>⁢</mo> > <mi>a</mi> > <mo>⁢</mo> > <mi>c</mi> > </msqrt> > </mrow> > <mrow> > <mn>2</mn> > > <mo>⁢</mo> > <mi>a</mi> > </mrow> > </mfrac> > </math>.</p> > <p><b>Warning:</b> Remember that ± means that there are two > solutions!</p> > </body></html> > > Regards- > -Doug Schepers > W3C Team Contact, SVG and WebApps WGs
Received on Tuesday, 3 August 2010 20:57:02 UTC