3.1.0

This minor release provides the following enhancements and bug fixes:

• Incremental write, as introduced in the MarkLogic Java Client 8.1.0 release, is now available through a set of options prefixed with spark.marklogic.write.incremental.
• To support incremental write, the option spark.marklogic.write.logSkippedDocuments can be used to control how often messages are logged when documents are skipped.
• The spark.marklogic.streamTransformBinaryExtensions option allows for defining a comma-delimited list of URI extensions that controls which documents are sent to a REST transform when streaming documents during import. This supports a use case of streaming documents into MarkLogic when a subset of the documents must be sent to a REST transform, typically to alter the document type - e.g. ensuring that a JSON or XML document is loaded as a binary. This allows for only passing a subset of documents through a REST transform, as using a REST transform requires loaded the document into memory in MarkLogic which runs counter to the purpose of streaming.
• To support exporting and importing archives where JSON and/or XML documents are stored in MarkLogic as binaries, the name of a metadata entry in an archive file will now include the document type - e.g. "JSON", "XML", "TEXT", or "BINARY". The document type is not required to be present in an archive file - i.e. archive files created with previous version of the connector will still be imported correctly.
• Fixed a bug where empty documents would cause an error when written to an archive file.