Add Parameterized Terms

[stasm]

Fix #176.

[Pike]

Can you please open a separate issue for the changes to function names?

I'm wondering if we could formulate the parametrized terms better. How about replacing TermReference in InlineExpression with TermExpression, and maybe factor the argument list a bit to include the full call expression postfix?

I picture this from a resolver implementation, and I expect the implementation to set up the environment for Terms to be rather different than to that of a call into the globals. So implementers might do the above anyway, but they'd need to piece that together?

This would also avoid the addition to valid.md? There'd be renames there, though.

[stasm]

Can you please open a separate issue for the changes to function names?

Sure: #209.

I'm wondering if we could formulate the parametrized terms better. How about replacing TermReference in InlineExpression with TermExpression, and maybe factor the argument list a bit to include the full call expression postfix?

That's an interesting approach. However, I see CallExpression as being useful for macros because of its generality. We might want to enable the -term.attr() syntax soon (#189). And later, dynamic references will likely require $var() (#80).

I picture this from a resolver implementation, and I expect the implementation to set up the environment for Terms to be rather different than to that of a call into the globals. So implementers might do the above anyway, but they'd need to piece that together?

Or they can check the type of the callee.

[Pike]

I'm not happy with the split of where term references are after this patch. But I also see that we might have { $animal(decl: "nominative") } at some point, where the point that we're have a TermReference as callee won't be visible to the parser.

I filed #213 to document what we're doing here.

[Pike]

Are positional arguments intended to be legal?

[stasm]

They're not used by the runtime to resolve the term so even if they're passed, they're just ignored. We could forbid them in abstract for now, in case we find a use for them in the future. How does that sound to you?

[Pike]

I wondered about abstact, too, but no idea if there's going to be a use or not. I'm not a friend of "for now", either, so, well. Yeah, I filed a documentation bug. No strong opinion.

[stasm]

Let's leave them as valid, then. Thanks for filing the issue about the docs.