…ne for RELAXNG, one for DTD) with a new function, tei:norMinMax(); should be no change to output; passes both Test/ and Test2/.

… both sets of tests.

… untested because of #631. This comit is *not* ready for prime time.

In this commit I intended to comment out some debugging code (which renders output kinda useless, is it sticks the results of each pass into the output as RELAXNG <div> elements, rather than as comments) and update some tests. But I found to my horror that (despite message of previous 3 commits) tests were not passing. So lots of test updates, too. NB: One of the things that makes this ineligible for merging into dev is that the constraint for the <choice> element has changed from the incorrect, but workable '( model.choicePart | choice )+' to the incorrect and *not* workable '( model.choicePart | choice )+'. At the moment I have no idea why, but it does not really matter, as attacking minOccurs= and maxOccurs= on <alternate> and <sequence> is next.

…nterleave>), too. Passes tests, so *may* be eligible to merge into dev, but not really ready, yet, as need to work on repeatability of <alternate>.

 * Removed some debugging code
 * Added tests (in Test2/testPure1) for sequence/@m[ia][nx]Occurs and alternate/@m[ia][nx]Occurs, including valid instance, but not invalid
 * Updated tests (in Test/)

1) Apparently there are times when the 'c' variable in the template for elementRef|classRef|macroRef boils down to an empty sequence.
2) Handle the case when there is no <datatype> child of the <attDef> being processed.

…23-10-15 on PR #633.

…e Guidelines test process, and the Guidelines build OK.

…ake it work after said merge.