[HTML5 Ruby] Testing Strategy of HTML5 Ruby algorithm.

Dear HTML5 Test planners,

Current HTML5 CR Ruby Specification [1] provides sophisticated
"segmentation and categorisation" algorithm, which satisfies all basic
needs of ruby use-cases, in a mannar consistent with modern HTML5 DOM
tree construction.

However, this algorithm consists of several tens of steps, meaning
that it is quite important to create the test to check whether a
web user agent satisfies this specification or not.

As of it, I and my collegues are planning to provide the test for this
algorithm in the following strategy.

Basically, this "segmentation and categorisation" algorithm splits
ruby markups into several segments and then categorise them to "ruby
bases" and "ruby texts". Therefore, in conventional left-to-right
writing mode, segmented box would be lined up from left to right, so
we can compare X-coordinates of each box to check if segmentations is
succeeded or not.

Likely, by default, ruby text is always attached above the ruby bases.
So, we can compare Y-coordinates of each box to check if each inline
element (text) box is properly categorised to Ruby Texts and Ruby
Bases.

Based on this idea, we are now preparing the test suite which we hope
to contribute to HTML5 working group. It would take several weeks to
test the test suites, but beforehand, I really appreicate if anyone
could comments on this strategy to test the HTML5 Ruby segmentation
and categorisation algorithm if there is a concern.

Sincerely,

[1] ...
http://www.w3.org/html/wg/drafts/html/CR/text-level-semantics.html#the-ruby-element


-- 
---------------------------------------------------------------------
KAWABATA, Taichi E-mail: kawabata.taichi@gmail.com