Fix annotation and namespace issue (continued)

[jombr]

Continue the work of #19 to fix #16

Fixes #16, and a variety of other issues hit in processing my existing grammars to extract documentation

• Support documentation (or any other annotations) on groups and references to a named define
• use a: as the xmlns prefix for http://relaxng.org/ns/compatibility/annotations/1.0, but don't assume that the user has already defined it, and make sure to qualify the case where they defined namespace a=... to be something else too.
• Follow the spec for whitespace handling in a:documentation
• Fix combining |=, &= productions in the presence of annotations.
• Fix handling of datatag and literals with possibly-namespaced type references.

Additional changes:

• Re-based onto latest master
• Resolved #19 review style nit comments https://github.com/djc/rnc2rng/pull/19#pullrequestreview-293522561
• Implemented grammar level datatypeLibrary solution from https://github.com/djc/rnc2rng/pull/19#issuecomment-535790411

grammar level datatypeLibrary solution

• a grammar-level datatypeLibrary="http://www.w3.org/2001/XMLSchema-datatypes" gets printed if xsd:* was ever referenced
• Any datatypes != http://www.w3.org/2001/XMLSchema-datatypes written explicitly on <data> or <value>
• Include grammar root assert removed (one that tries to verify the grammar level datatypes are the same)

Changes in addition to discussed solution

• Capture datatypes from included grammar so the prefix can be looked up and included on explicit written custom datatypes

[djc]

Do you want to take over maintenance?

https://github.com/djc/rnc2rng/issues/41

[jombr]

Do you want to take over maintenance?

41

I'm willing to but there may be quite a bit with the code base, Continuous Integration, and Deployment to PiPy that I'm not familiar with. That and I started learning Python this year while only visiting relaxng off and on.

That being said I'm interested, enjoy learning, and the package seems pretty stable. If your comfortable with that and can be available for questions then I'll do it.

[djc]

I've invited you as a collaborator on this repository. Let's see how this works out? Since you're indicating that you're pretty new to things I'll withhold PyPI permissions for now -- feel free to ask questions.

[jombr]

I've invited you as a collaborator on this repository. Let's see how this works out? Since you're indicating that you're pretty new to things I'll withhold PyPI permissions for now -- feel free to ask questions.

Alright! I have some reading to do.

Let me know if this PR works for you (If you think its aligned with what @puetzk had in mind).

If this PR checks out I'll probably re-base it after #45 lands.

[coveralls]

coverage: 89.527% (+0.6%) from 88.963% when pulling 12073d925a2b7d984aa1e095f329717dc3838942 on josephbredahl:annotations into 813bc863f29ce13e005ec95ab3d3bd3b7beef411 on djc:main.

[djc]

Sorry, I won't have time to thoroughly review this PR -- happy to answer any specific questions you have, though!

[jombr]

No worries, figured it could be rude if I pushed without asking for a review. I'll review it again before merging.

[djc]

Not rude, I'm considering you the primary maintainer from here on, though I'll still want to check before we publish to PyPI.