Fix handling of allowed-values enumeration in JSON Schemas

[aj-stein-nist]

Committer Notes

Closes #240.

To not clutter PR, sample JSON schema resulting from this change after update from previous PR review:

https://gist.github.com/aj-stein-nist/17c21528df4c12a33a49aff665b6e8e8/

All Submissions:

• Have you followed the guidelines in our Contributing document?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?
• Have you squashed any non-relevant commits and commit messages? [instructions]
• Do all automated CI/CD checks pass?

Changes to Core Features:

• Have you added an explanation of what your changes do and why you'd like us to include them?
• Have you written new tests for your core changes, as applicable?
• Have you included examples of how to use your new feature(s)?
• Have you updated all website](https://pages.nist.gov/metaschema) and readme documentation affected by the changes you made? Changes to the website can be made in the website/content directory of your branch.

[aj-stein-nist]

@wendellpiez let me know how you feel about the cleaned up tests and how we ought to handle ongoing considerations, in the code to resolve this issue or elsewhere. Should we do a walk-through review to discuss changes and improvements (like the overlapping @target discussion)? Let me know what you think.

[aj-stein-nist]

@david-waltermire-nist and @wendellpiez, this is ready for re-review.

[aj-stein-nist]

There does seem to be weirdness with how using the nm:compose-metaschema(/) function call works and fails the test as opposed to taking the resulting output from it, via the stylesheet below, and transforming that.

<!-- return-compose-metaschema.xsl file -->
<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
    xmlns:xs="http://www.w3.org/2001/XMLSchema"
    xmlns:math="http://www.w3.org/2005/xpath-functions/math" exclude-result-prefixes="xs math"
    xmlns:nm="http://csrc.nist.gov/ns/metaschema"
    version="3.0" xmlns="http://www.w3.org/2005/xpath-functions"
    xpath-default-namespace="http://csrc.nist.gov/ns/oscal/metaschema/1.0" expand-text="true">

    <!-- Purpose: Produce an XPath-JSON document representing JSON Schema declarations from Metaschema source data.
        The results are conformant to the rules for the XPath 3.1 definition of an XML format capable of being cast
        (using the xml-to-json() function) into JSON. -->
        
    <!-- Note: this XSLT will only be used on its own for development and debugging.
        It is however imported by `produce-json-converter.xsl` and possibly other stylesheets. -->
    
    <xsl:strip-space elements="METASCHEMA define-assembly define-field model"/>
    
    <xsl:output indent="yes" method="xml"/>
    <xsl:import href="../nist-metaschema-COMPOSE.xsl"/>
    
    <xsl:template match="/">
        <xsl:sequence select="nm:compose-metaschema(/)"/>
    </xsl:template>

</xsl:stylesheet>

This has left me scratching my head. Wendell believes, and there is a good chance, this difference has something to do with how XSpec builds context to make unit testing possible even with @run-as="external" something might be off. I need to explore that further. I will either resolve this hurdle in the next day or talk with @david-waltermire-nist about just adding became the "pre-composed metaschema" for the test and backing this out into a follow-on issue to figure out before it expends any other time.

[aj-stein-nist]

Touched up a rebase for now to come back to this in the evening or tomorrow morning.

[aj-stein-nist]

@wendellpiez recommended this morning an alternate approach and just using the shell (aka nist-metaschema-MAKE-JSON-SCHEMA.xsl and adjusting the tests accordingly. Experimenting with that now.

[aj-stein-nist]

I need to clean up base (weirdly git pull -r origin develop isn't doing what I'd expect), but I figured it out and this is ready for review.

Hat tip @liamquin, @martin-honnen, and of course @wendellpiez for helping me getting through what should have been an easy refactor.

[david-waltermire]

The test vectors look good to me. I cannot weigh in on the XSLT. I'll rely on the review from @wendellpiez for that.

[aj-stein-nist]

OK, @david-waltermire-nist and @wendellpiez I rebased to be clean against recent develop from yesterday afternoon, I have also created the requested tech debt issue I can begin working shortly.

usnistgov/metaschema-xslt#25

Please let me know if you need anything for finalizing a review and rolling this into develop and move onto the next issue.

[wendellpiez]

Looking good - so far!