compat: Ensure that pandas dtype matches dask when loading data from parquet

[hoxbro]

It makes it so that the following will give the same types before it returned (string[pyarrow], object).

import dask
import spatialpandas.io as sio

dask.config.set({"dataframe.convert-string": True})

# http://s3.amazonaws.com/datashader-data/nyc_buildings.parq.zip
ddf = sio.read_parquet_dask("./data/nyc_buildings.parq")
ddf["type"].dtype, ddf["type"].compute().dtype

Together with https://github.com/holoviz/holoviews/pull/6362 should make it possible to run the NYC Buildings example.

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 93.10345% with 2 lines in your changes missing coverage. Please review.

Project coverage is 77.72%. Comparing base (3a1b6eb) to head (6c0f0e1). Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
spatialpandas/io/parquet.py	75.00%	2 Missing :warning:

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #156 +/- ## ========================================== + Coverage 77.47% 77.72% +0.24% ========================================== Files 50 50 Lines 4844 4871 +27 ========================================== + Hits 3753 3786 +33 + Misses 1091 1085 -6 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.