Namespace Meta tag

[SteveLLamb]

Example (using 429-16) of one (or more) definitions of new meta tag for defined namespace, could be either validate and/or injected from XSD as additional element. (defined as "qualifier; value" in content field).

<meta itemprop="xmlns" content="xs; http://www.w3.org/2001/XMLSchema">
<meta itemprop="xmlns" content="cpl; http://www.smpte-ra.org/schemas/429-7/2006/CPL">
<meta itemprop="xmlns" content="meta; http://www.smpte-ra.org/schemas/429-16/2014/CPL-Metadata">

This is for better data extraction on the HTML.

[palemieux]

I would prefer the notation be inline, where the namespace is defined, so that the header does not become out of sync.

[SteveLLamb]

Then do we need to define how that is handled outside of just being in the XSD as an additional element? Right now some docs define via a table and some via a code block and some via just prose. It's all over the place and nearly impossible to parse without the XSD (assuming it's even properly attached, which isn't defined either).

[palemieux]

Yes, there should be a way to label an element as specifying a namespace that is defined in the document.

For instance we could do:

<span itemscope itemtype="http://smpte.org/standards/documents" itemprop="xmlns">http://www.smpte-ra.org/schemas/2067-3/2016</span>

Why is the namespace prefix important? It is completely arbitrary.

[SteveLLamb]

Well like in the case of 429-16 example, it makes it easier to map and validate against the schema that uses the same ns qualifiers, and then also quickly see which are root elements (such as CPL, and defined elsewhere) and which are extensions that are being defined in this doc.

Noting this is via machine reading not by a person.

[palemieux]

The machine should not care about the prefixes: XML validators automatically elide them.

[SteveLLamb]

I'm not speaking about a xml or content validator specifically. I mean about a machine parsing the doc (only) to know what namespace(s) are defined in the doc being read and which are only used as reference and being able to map a schema to a specific doc that defines it. In this case specifically - my standards document registry. It's not going to be able to decipher that specific context via only prose. Even a flag/attribute that defines that context would suffice such as:

'span itemscope itemtype="http://smpte.org/standards/documents"
itemprop="xmlns"
content="http://www.smpte-ra.org/schemas/2067-3/2016"
data-definition="local">
</span'

This would also enable a possible validation system to be able to reverse that logic and know which document to point someone to in the case of schema issues.

[palemieux]

content="http://www.smpte-ra.org/schemas/2067-3/2016"
data-definition="local"

Why do you need the content attribute if the contents of the span is the namespace?

[SteveLLamb]

Riiiiiiight. I was thinking for all cases including non-SMPTE ones. Suppose I could parse against the doc info and deduce it's internal vs anything else is assumed external (and then parse refs as needed). Thanks!