feat: Support use of is_between range predicate with IEJoin operations (join_where)

[alexander-beedie]

Ref: #18365

One of the most useful applications of join_where is range joins, and the is_between expression is a natural way to express them. This PR allows for use of such in the list of join predicates, streamlining these queries.

Also improved several related error messages, providing more detail if a given predicate is not suitable for use with join_where (such as actually including that predicate in the error).

Example

Setup:

import polars as pl

df1 = pl.DataFrame({
    "id": ["aa", "bb", "cc"],
    "start": [date(2020,1,1), date(2022,10,10), date(2024,7,5)],
    "end": [date(2022,12,31), date(2024,10,1), date(2024,12,31)],
})
df2 = pl.DataFrame({
    "id": ["aa", "cc", "bb"],
    "dt": [date(2022,12,31), date(2024,2,21), date(2024,8,8)],
    "price": [100, 200, 300],
})

The following two formulations are now equivalent:

df1.join_where(
    df2,
    pl.col("id") == pl.col("id_right"),
    pl.col("dt") >= pl.col("start"),
    pl.col("dt") <= pl.col("end")
)

df1.join_where(
    df2,
    pl.col("id") == pl.col("id_right"),
    pl.col("dt").is_between("start","end")
)

shape: (2, 5)
┌─────┬────────────┬────────────┬────────────┬───────┐
│ id  ┆ start      ┆ end        ┆ dt         ┆ price │
│ --- ┆ ---        ┆ ---        ┆ ---        ┆ ---   │
│ str ┆ date       ┆ date       ┆ date       ┆ i64   │
╞═════╪════════════╪════════════╪════════════╪═══════╡
│ aa  ┆ 2020-01-01 ┆ 2022-12-31 ┆ 2022-12-31 ┆ 100   │
│ bb  ┆ 2022-10-10 ┆ 2024-10-01 ┆ 2024-08-08 ┆ 300   │
└─────┴────────────┴────────────┴────────────┴───────┘