Change default TINYINT conversion type

df = spark.read.format("jdbc").option("url", jdbc_url).option("dbtable", f"({query}) as tmp").options(**jdbc_conn_properties).load()
df.printSchema()

When checking how Azure SQL Database types are read in Fabric using JDBC INT, BIGINT and SMALLINT are transformed into logically sized types. TINYINT is read as an integer by default, which is not efficient in storage use.