[BUG] Translate mssql to tsql in read_sql scan

[colin-ho]

When running read_sql against SQL server using a SQL Alchemy connection, for example:

connection_url = sqlalchemy.engine.URL.create(
    "mssql+pyodbc",
    username=user,
    password=password,
    host=host,
    port=1433,
    database=database,
    query={
        "driver": "ODBC Driver 18 for SQL Server",
    },
)
def create_conn():
    return sqlalchemy.create_engine(connection_url).connect()

df = daft.read_sql("SELECT * FROM test_data", create_conn)

The query errors with Unsupported dialect: mssql, please refer to the documentation for supported dialects.

This is because SQLGlot, the library that read_sql uses for query construction, does not recognize mssql as a dialect, it instead recognizes tsql, which is the name of the SQL dialect for Microsoft SQL Server: https://learn.microsoft.com/en-us/sql/t-sql/language-reference?view=sql-server-ver16

This PR adds a translation step during sql query construction to fix this issue.

NOTE:

• This PR was tested locally against a Docker instance of Azure SQL Edge.