fix: add signature for read_excel

[universalmind303]

closes #2373

> select function_name, parameters from glare_catalog.functions WHERE function_name = 'read_excel';
┌───────────────┬─────────────────────────────────────────────────────────────────────────┐
│ function_name │ parameters                                                              │
│ ──            │ ──                                                                      │
│ Utf8          │ List<Utf8>                                                              │
╞═══════════════╪═════════════════════════════════════════════════════════════════════════╡
│ read_excel    │ [Utf8, Utf8, sheet_name: Utf8, infer_rows: UInt64, has_header: Boolean] │
└───────────────┴─────────────────────────────────────────────────────────────────────────┘

[greyscaled]

Quick Q:

In the PR description, why is there two Utf8's in a row?

[Utf8, Utf8, sheet_name: Utf8, infer_rows: UInt64, has_header: Boolean]

The first seems to be for required path arg, but what's the second?

[tychoish]

From is implemented for Vec<Field> and Fields, so I think FromIterator handlers, which might make this simpler?

To be clear, this is fine, but I definitely stopped and read this three times because it looks wrong. "why are we making a vector, converting it to an iterator and collecting it back into a vector without doing anything else?" and while it compiles and does the thing, mds

If it's easy enough to change I think that would make it better and less likely to confuse other people in the future.

[universalmind303]

Quick Q:

In the PR description, why is there two UTF8's in a row?

Datafusion doesn't really give us that good formatting out of the box for the signatures.

so the first signature is Utf8 -> read_excel('path'), the second is Utf8, sheet_name: Utf8, infer_rows: UInt64, has_header: Boolean -> read_excel('path', options)

@greyscaled I agree that the formatting choices are not great though.

[greyscaled]

Quick Q:
In the PR description, why is there two UTF8's in a row?

Datafusion doesn't really give us that good formatting out of the box for the signatures.

so the first signature is Utf8 -> read_excel('path'), the second is Utf8, sheet_name: Utf8, infer_rows: UInt64, has_header: Boolean -> read_excel('path', options)

@greyscaled I agree that the formatting choices are not great though.

all good, I appreciate the explanation