leesf commented on a change in pull request #3754:
URL: https://github.com/apache/hudi/pull/3754#discussion_r724012463
##########
File path:
hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/HoodieSqlUtils.scala
##########
@@ -92,7 +93,45 @@ object HoodieSqlUtils extends SparkAdapterSupport {
properties.putAll((spark.sessionState.conf.getAllConfs ++
table.storage.properties).asJava)
HoodieMetadataConfig.newBuilder.fromProperties(properties).build()
}
- FSUtils.getAllPartitionPaths(sparkEngine, metadataConfig,
HoodieSqlUtils.getTableLocation(table, spark)).asScala
+ FSUtils.getAllPartitionPaths(sparkEngine, metadataConfig,
getTableLocation(table, spark)).asScala
+ }
+
+ /**
+ * This method is used to compatible with the old non-hive-styled partition
table.
+ * By default we enable the "hoodie.datasource.write.hive_style_partitioning"
+ * when writing data to hudi table by spark sql by default.
+ * If the exist table is a non-hive-styled partitioned table, we should
+ * disable the "hoodie.datasource.write.hive_style_partitioning" when
+ * merge or update the table. Or else, we will get an incorrect merge result
+ * as the partition path mismatch.
+ */
+ def isNotHiveStyledPartitionTable(partitionPaths: Seq[String], table:
CatalogTable): Boolean = {
Review comment:
would rename to `isHiveStylePartitionPartitioning`
##########
File path:
hudi-spark-datasource/hudi-spark/src/main/scala/org/apache/spark/sql/hudi/HoodieSqlUtils.scala
##########
@@ -92,7 +93,45 @@ object HoodieSqlUtils extends SparkAdapterSupport {
properties.putAll((spark.sessionState.conf.getAllConfs ++
table.storage.properties).asJava)
HoodieMetadataConfig.newBuilder.fromProperties(properties).build()
}
- FSUtils.getAllPartitionPaths(sparkEngine, metadataConfig,
HoodieSqlUtils.getTableLocation(table, spark)).asScala
+ FSUtils.getAllPartitionPaths(sparkEngine, metadataConfig,
getTableLocation(table, spark)).asScala
+ }
+
+ /**
+ * This method is used to compatible with the old non-hive-styled partition
table.
+ * By default we enable the "hoodie.datasource.write.hive_style_partitioning"
+ * when writing data to hudi table by spark sql by default.
+ * If the exist table is a non-hive-styled partitioned table, we should
+ * disable the "hoodie.datasource.write.hive_style_partitioning" when
+ * merge or update the table. Or else, we will get an incorrect merge result
+ * as the partition path mismatch.
+ */
+ def isNotHiveStyledPartitionTable(partitionPaths: Seq[String], table:
CatalogTable): Boolean = {
+ if (table.partitionColumnNames.nonEmpty) {
+ val isHiveStylePartitionPath = (path: String) => {
+ val fragments = path.split("/")
+ if (fragments.size != table.partitionColumnNames.size) {
+ false
+ } else {
+ fragments.zip(table.partitionColumnNames).forall {
+ case (pathFragment, partitionColumn) =>
pathFragment.startsWith(s"$partitionColumn=")
+ }
+ }
+ }
+ !partitionPaths.forall(isHiveStylePartitionPath)
+ } else {
+ false
+ }
+ }
+
+ /**
+ * If this table has disable the url encode, spark sql should also disable
it when writing to the table.
+ */
+ def isUrlEncodeDisable(partitionPaths: Seq[String], table: CatalogTable):
Boolean = {
Review comment:
`isUrlEncodeEnabled` and fix the description please
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@hudi.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
