feat(python): add "pyxlsb" engine support to read_excel (for excel binary workbook files)

[alexander-beedie]

Closes #11181 (and also closes #11184 by adding a .. versionadded tag to the docs).

• Adds support for the pyxlsb engine so that we can also read Excel Binary Workbook files (those with an ".xlsb" extension, which are not compatible with any of the existing ".xlsx" engines). Note that this engine does not currently autodetect datetime/date columns (it reads them in as Excel's native offset-Julian float), and therefore requires the use of schema_overrides to load them correctly (I have added a note in the docstring about this).

• Also: improves Date parsing for OpenOffice files via read_ods).

• Also: slightly updates show_versions() docs/example with the latest libs.

---

Support for spreadsheet data is improving nicely; with this update read_excel can now read all of the major Excel formats (".xlsx", ".xlsm", "xlsb"), and we can handle the OpenOffice ".ods" format via read_ods.

Example

Reading ".xlsb" files:

pl.read_excel(
    source = "~/test.xlsb",
    sheet_name="misc_data",
    schema_overrides={"dtm":pl.Datetime, "dt":pl.Date},
)

Before

xlsx2csv.XlsxException: Sheet 'misc_data' not found

After

shape: (2, 3)
┌─────────────────────┬────────────┬──────┐
│ dtm                 ┆ dt         ┆ val  │
│ ---                 ┆ ---        ┆ ---  │
│ datetime[μs]        ┆ date       ┆ f64  │
╞═════════════════════╪════════════╪══════╡
│ 1999-12-31 10:30:45 ┆ 2024-01-01 ┆ 1.5  │
│ 2010-10-11 12:13:14 ┆ 2018-08-07 ┆ -0.5 │
└─────────────────────┴────────────┴──────┘

---

FYI: @SaelKimberly has been experimenting with writing a potentially superior ".xlsb" reading engine; if & when this is ready (and well tested and available from pypi ;) we can look at including that as the default .xlsb engine instead - looking forward to it 👍

[stinodego]

Great stuff as always! Some minor remarks/questions.

[stinodego]

All right, good to go!