XML parser doesn't support namespace declarations

[m8pple]

The XML parser does not correctly process and ignore namespace declarations. e.g.:

<?xml version='1.0'?>
<Graphs xmlns="https://poets-project.org/schemas/virtual-graph-schema-v4" formatMinorVersion="0">
<GraphType xmlns="https://poets-project.org/schemas/virtual-graph-schema-v4" xmlns:ns0="https://poets-project.org/schemas/virtual-graph-schema-v3" id="clock_tree">
  <Properties><![CDATA[uint32_t max_ticks;
]]></Properties>
  <SharedCode><![CDATA[

where "ns0" is never used as a prefix. This pattern occasionally turns up when gluing together xml fragments in other languages, and a compliant XML parser would be expected to just ignore the prefix (well, unless ns0 is actually used, then it should process it).

I guess it is an artefact of rolling your own XML front-end - the IC and UoN tools never had a problem as they all used pre-existing XML parsers. But I'm not sure why it wasn't picked up as a parsing problem during parser development, as it appears in 100s of the v4 xml graphs included in the 2019-09-06 stress/benchmark set (though not all of them, as they come from different generators).

It can be worked around on the input side with filter programs, so it probably isn't worth the effort of fixing this in the Orchestrator - but still an issue, albeit a low-priority one.

[DrongoTheDog]

It was a deliberate decision not to include namespaces. The objective at the time was to support the minimum viable subset as fast as possible so we could move around/past/through the ADR contribution. I couldn't (and still can't) see how these would be useful in the POETS context, but it wold be quite easy to incorporate them if anyone tells me in what form they would be stored. (Compound names -> vectors of strings?)

[DrongoTheDog]

Or I could just replace the ':' with a '_' and turn the thing into a simple name? I don't really like the idea of embedding ':' in a string, 'cos it's an operator elsewhere in the grammar and there be dragons.

[mvousden]

Decision (DBT, ADB, GMB, MLV) to ignore xmlns entirely (at least for now).

[m8pple]

This one is still relevant: there is perfectly valid v3 XML that still exists and is still used, as well as v4 XML. Ignoring v4 XML is an appropriate strategy for the Orchestrator as there are tools for converting vX to v4, but XML that contains no xmlns cannot be parsed.

So the Orchestrator can continue to ignore xmlns. But any example files should still contain a valid xmlns so that other implementations can understand how to parse them.