quasi-polynomials and annotated functions

Nobody has claimed this yet.

I have a quasi-polynomial given to me as a string from an external program:
```m2
H(x,y) = ((1 + x) + (5/2 + x) * y + 3/2 * y^2) + (-1/2 + x) * floor((x)/3) - 3/2 * floor((x)/3)^2
```
Unfortunately M2 doesn't support defining this as a native polynomial because of the floor, though it would be great if it did. One approach is to store 3 different polynomials, or perhaps 1 common polynomial + 3 small correction terms.

Currently I've resorted to parsing that using regex and defining:
```m2
H = value "(x,y) -> ((1 + x) + (5/2 + x) * y + 3/2 * y^2) + (-1/2 + x) * floor((x)/3) - 3/2 * floor((x)/3)^2"
```
So I can use it to evaluate terms, but ideally I'd like to _see_ it like a polynomial, and `code H` doesn't work, because the source isn't in a file (I just see `FunctionClosure[currentString:1:0-1:82]`).

I resorted to something hacky, like this:
```m2
i62 : setAttribute(H, symbol ReverseDictionary, "((1 + x) + (5/2 + x) * y + 3/2 * y^2) + (-1/2 + x) * floor((x)/3) - 3/2 * floor((x)/3)^2");

i63 : H

o63 = ((1 + x) + (5/2 + x) * y + 3/2 * y^2) + (-1/2 + x) * floor((x)/3) - 3/2 * floor((x)/3)^2

o63 : FunctionClosure
```
But it feels like this might have some unintended consequences. (Sidenote: using the attribute `PrintNames` or `PrintNet` did not have any effect. Why do we have them?!)

Any suggestions?

Start with the value-created FunctionClosure and the code command, then inspect how ReverseDictionary, PrintNames, and PrintNet affect display. Determine whether the intended change is source-backed display or a native quasi-polynomial representation; done requires an agreed design and defined evaluation and printed-output behavior.

Written by the indexing model from the issue text.