Effective in version 10.4.1, you can use the Amazon S3 and Azure Data Lake Store resources to identify and publish horizontally partitioned files under the same directory, and files organized in hierarchical Hive-style directory structures as a single partitioned file.