Github user sureshthalamati commented on a diff in the pull request:
https://github.com/apache/spark/pull/16209#discussion_r107562605
--- Diff:
sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/jdbc/JdbcUtils.scala
---
@@ -680,19 +681,63 @@ object JdbcUtils extends Logging {
/**
* Compute the schema string for this RDD.
*/
- def schemaString(schema: StructType, url: String): String = {
+ def schemaString(
+ schema: StructType,
+ url: String,
+ createTableColumnTypes: Option[String] = None): String = {
val sb = new StringBuilder()
val dialect = JdbcDialects.get(url)
+ val userSpecifiedColTypesMap = createTableColumnTypes
+ .map(parseUserSpecifiedCreateTableColumnTypes(schema, _))
+ .getOrElse(Map.empty[String, String])
schema.fields foreach { field =>
val name = dialect.quoteIdentifier(field.name)
- val typ: String = getJdbcType(field.dataType,
dialect).databaseTypeDefinition
+ val typ: String = userSpecifiedColTypesMap.get(field.name)
+ .getOrElse(getJdbcType(field.dataType,
dialect).databaseTypeDefinition)
val nullable = if (field.nullable) "" else "NOT NULL"
sb.append(s", $name $typ $nullable")
}
if (sb.length < 2) "" else sb.substring(2)
}
/**
+ * Parses the user specified createTableColumnTypes option value string
specified in the same
+ * format as create table ddl column types, and returns Map of field
name and the data type to
+ * use in-place of the default data type.
+ */
+ private def parseUserSpecifiedCreateTableColumnTypes(schema: StructType,
+ createTableColumnTypes: String): Map[String, String] = {
+ val userSchema =
CatalystSqlParser.parseTableSchema(createTableColumnTypes)
+ val userColNames = userSchema.fieldNames
+ // check duplicate columns in the user specified column types.
+ if (userColNames.distinct.length != userColNames.length) {
+ val duplicates = userColNames.groupBy(identity).collect {
+ case (x, ys) if ys.length > 1 => x
+ }.mkString(", ")
+ throw new AnalysisException(
+ s"Found duplicate column(s) in createTableColumnTypes option
value: $duplicates")
+ }
+ // check user specified column names exists in the data frame schema.
+ val commonNames = userColNames.intersect(schema.fieldNames)
--- End diff --
Thank you for the review. Good question., updated the PR with
case-sensitive handling. Now column names from user specified schema are
matched with data frame schema based on the SQLConf.CASE_SENSITIVE flag.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
