This fixes WCAG 2.0 compliance because we were already using HTML5 markup.
The strict XML headers setting never worked for production servers, developers
used browser validators for compliance testing. XHTML 5 option is relatively
similar to this obsolete option, but still it can not be used on production servers.
XHTML Strict 1.x was a standardised dead end, HTML5 is the new de-facto-standard
supported by all major browsers including latest versions of IE.
Please note nothing changes in our coding style because HTML5 is a superset of
several previous standards, it is recommended to use only features that are
already implemented in all our supported browsers.
It seems that sometimes trying to process an overdue quiz attempt can
throw an exception. In that case, we need to catch it and report it
nicely, and then carry on processing the rest of the attempts, rather
than just killing the whole of cron processing.
Also, there may be garbage quiz attempts where the associated quiz or
course has been deleted. Skip those too.
In the case where either a question_attempt had not steps, or a
question_usage had not question_attempts, the load_from_records methods
could get stuck in an infinite loop.
This fix ensures that does not happen, with unit tests to verify it. At
the same time, I noticed an error in the existing tests, which this
patch fixes.
These changes move the logic to the behaviour class, which is how things
should be.
It also makes it easier to re-use the code that displays the messages
like "Marks for this submission: 1.00/1.00. Accounting for previous
tries, this gives 0.33/1.00." in other places.
To try to make it clearer what is going on, I introduced a new class
qbehaviour_adaptive_mark_details to hold and document the data that the
behaviour needs to return, and the renderer can base its display on.
As far as my testing can tell (and there are new unit tests), this
commit does not change the existing behaviour.
This commit also replaces all the string concatenation that is going on,
which should help translators. At the moment, a few of the old strings
are still in the language file, and are used in the unit tests to verify
that the behaviour has not changed.
Thanks to Oleg Sychev for making a helpful suggestion about the API.
