Message class should be slightly amended

[fabrizio-turchi]

In the init method there is the call self.append_indexed_items included in base.py but the code is useless even though it can be ignored, no issues are raised by the case_validate.

def append_indexed_items(self, *args):
        """
        :param args: A single/tuple of ObservableObjects
        """

        for key in ["olo:length", "olo:slot"]:
            if self.get(key) is None:
                self[key] = None

        if len(args) == 1 and not args[0]:  # True if no objects to append provided
            pass
        else:
            if len(args) and self["olo:slot"] is None:
                self["olo:slot"] = list()
...

[ajnelson-nist]

My understanding is this method was built on an example pre-dating UCO's adoption of the Collections Ontology (done in Issue 389).

If there's a desire for this method, it should:

1. Replace Ordered List Ontology (olo:) references with Collections Ontology (co:).
2. Move to MessageThreadFacet, or to a yet-unwritten class specifically for uco-types:Thread management.

I think the right direction is to remove this function, and code up a demo of the thread illustrated in the ASCII art in UCO Issue 393, Competency 1 ("Suppose an email message set is found..."). Threads being non-linear, diverging, and converging, will open up some questions on class design.

[fabrizio-turchi]

Every time I use the Message class the following code is included:

 {
            "@id": "kb:...",
            "@type": "uco-observable:Message",
            "olo:length": null,
            "olo:slot": null,
            "uco-core:hasFacet": [
            ]
}

It's a harmless but useless code.

[ajnelson-nist]

I've removed the affect from uco-observable:Message in #77 .

I haven't looked into addressing the longer-term questions, though.

[fabrizio-turchi]

@ajnelson-nist thanks! It's up to you to close it or leave it open.

[ajnelson-nist]

There's at least one more PR I'll push later next week to convert olo to co, because I saw that the append_indexed_items function is being used in the drafting module. So, let's leave this open for now, but by my understanding, it's no longer a blocker for you, or generating distracting data.