I found that Teradata does not have the Dialect in spark.So I want to add the TeradataDialect.scala in package org.apache.spark.sql.jdbc to support Teradata database better.
I override three functions in TeradataDialect.scala:
1.The url: jdbc:teradata
2.The JDBCType:Teradata database does not support "TEXT",so we replace it with "VARCHAR(255)".Also we replace "BLOB" with "VARBYTE(4)".
3.Teradata database does not support sql like "LIMIT 1",so we replace it with "TOP 1".