pl.scan_parquet('gs://...') breaks when GOOGLE_APPLICATION_CREDENTIALS env variable is set

[Vincenthays]

Checks

• I have checked that this issue has not already been reported.

• I have confirmed this bug exists on the latest version of Polars.

Reproducible example

import polars as pl
pl.scan_parquet('gs://<bucket>/<file>.parquet', storage_options={'google_service_account': '/some/google-cloud-creds.json'}) # working

import os
os.environ['GOOGLE_APPLICATION_CREDENTIALS'] = '/some/google-cloud-creds.json'
pl.scan_parquet('gs://<bucket>/<file>.parquet') # not working (err log bellow)

pl.scan_parquet('gs://<bucket>/<file>.parquet', storage_options={'google_service_account': '/some/google-cloud-creds.json'}) # not working (same error bellow)

Log output

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/local/lib/python3.11/site-packages/polars/io/parquet/functions.py", line 268, in scan_parquet
    return pl.LazyFrame._scan_parquet(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.11/site-packages/polars/lazyframe/frame.py", line 453, in _scan_parquet
    self._ldf = PyLazyFrame.new_from_parquet(
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
polars.exceptions.ComputeError: Generic GCS error: Unsupported ApplicationCredentials type: service_account

Issue description

pl.scan_parquet doesn't work when GOOGLE_APPLICATION_CREDENTIALS env variable is set since 0.19.6

Expected behavior

pl.scan_parquet works with GOOGLE_APPLICATION_CREDENTIALS env variable

Installed versions

--------Version info---------
Polars:              0.19.12
Index type:          UInt32
Platform:            Linux-5.4.0-104-generic-x86_64-with-glibc2.36
Python:              3.11.6 (main, Nov  1 2023, 13:45:43) [GCC 12.2.0]

----Optional dependencies----
adbc_driver_sqlite:  <not installed>
cloudpickle:         <not installed>
connectorx:          <not installed>
deltalake:           <not installed>
fsspec:              2023.10.0
gevent:              <not installed>
matplotlib:          <not installed>
numpy:               1.26.1
openpyxl:            <not installed>
pandas:              2.1.2
pyarrow:             14.0.0
pydantic:            <not installed>
pyiceberg:           <not installed>
pyxlsb:              <not installed>
sqlalchemy:          2.0.22
xlsx2csv:            <not installed>
xlsxwriter:          <not installed>

[tustvold]

This should have been fixed by the latest object_store release, in particular apache/arrow-rs#4926

[Vincenthays]

I check on main and it was working just fine, I think you are right, the new version of object_store fixes it