The _ARROW_SCALAR_IDS_TO_BQ mapping misses LargeStringArray type

[chelsea-lin]

Environment details

• OS type and version: Linux
• Python version: 3.11.7
• pip version: 23.3.1
• google-cloud-bigquery version: 3.16.0

Steps to reproduce

Call bqclient.load_table_from_dataframe with a dataframe containing a string type. Before pandas 2.2.0, the pyarrow.array would detect the type as pyarrow.lib.StringArray. After switching to pandas 2.2.0, the pyarrow.lib.LargeStringArray would be returned. But it misses the BQ type mapping.

Stack trace

The left results are in pandas 2.2.0 and the right result are from pandas 2.1.3