feat(polars): add ArrayDistinct operation

[IndexSeek]

Description of changes

Implements ArrayUnion with pl.Series.list.unique.

The test marker

Similar to what I was experiencing when implementing ArrayUnion, I'm getting nan/floats when there is a None in the array<int64> column. Here is some of the code I was using to test this behavior:

In [1]: import polars as pl
   ...: from ibis.interactive import *

In [2]: d_con = ibis.connect("duckdb://")
   ...: p_con = ibis.connect("polars://")

In [3]: df = pl.DataFrame({"a": [[1, 3, 3], [], [42, 42], [], [None], None]})
   ...: t = ibis.memtable(df)

In [4]: expr = t.select("a", uniqued=_.a.unique())

In [5]: d_con.execute(expr)
Out[5]: 
           a uniqued
0  [1, 3, 3]  [3, 1]
1         []      []
2   [42, 42]    [42]
3         []      []
4     [None]  [None]
5       None    None

In [6]: p_con.execute(expr)
Out[6]: 
                 a     uniqued
0  [1.0, 3.0, 3.0]  [1.0, 3.0]
1               []          []
2     [42.0, 42.0]      [42.0]
3               []          []
4            [nan]       [nan]
5             None        None

In [7]: df.select(pl.col("a"), pl.col("a").list.unique().alias("uniqued"))
Out[7]: 
shape: (6, 2)
┌───────────┬───────────┐
│ a         ┆ uniqued   │
│ ---       ┆ ---       │
│ list[i64] ┆ list[i64] │
╞═══════════╪═══════════╡
│ [1, 3, 3] ┆ [1, 3]    │
│ []        ┆ []        │
│ [42, 42]  ┆ [42]      │
│ []        ┆ []        │
│ [null]    ┆ [null]    │
│ null      ┆ null      │
└───────────┴───────────┘