Add missing bits to SAWScript's Cryptol import syntax

[sauclovian-g]

Closes #2605.

Note that it now accepts import submodule and stuffs the existence of the word submodule into the Import record, but doesn't do anything with it; @mtullsen I expect you have places to hook that up to in your code.

[RyanGlScott]

Surely we should have a test case that exercises this new bit of the parser?

[sauclovian-g]

We should, but I left it off for now because it's going to interact with Mark's branches. There's basically three options:

1. Merge this now, rebase all of Mark's code on it, add code there to recognize the submodule tag when processing the import, add a test there.

2. Merge Mark's code first, rebase this on it, add code here to recognize the submodule tag when processing the import, add a test here.

3. Merge both independently, then do another branch that adds the code to recognize the submodule tag when processing the import along with a test.

For various reasons (3) seems preferable...

[sauclovian-g]

(Contributing factor: nothing new should be needed on Mark's side to be able to handle the qualified names; that was purely a parser limitation)

[RyanGlScott]

Ah, I see. Should I understand that this just makes the parser recognize the submodule syntax, but doesn't actually do anything semantically meaningful with it?

[sauclovian-g]

Yes.

[sauclovian-g]

(rebased on master since we merged #2562)

@mtullsen do you want to attend to making the import submodule syntax DTRT or do you want me to? (in the latter case you'll have to tell me what to do though)

[sauclovian-g]

(this force-push to fix test_sawscript_builtins)