leesf commented on a change in pull request #1761: URL: https://github.com/apache/hudi/pull/1761#discussion_r446965016
########## File path: docs/_docs/2_2_writing_data.md ########## @@ -176,15 +176,49 @@ In some cases, you may want to migrate your existing table into Hudi beforehand. ## Datasource Writer -The `hudi-spark` module offers the DataSource API to write (and also read) any data frame into a Hudi table. -Following is how we can upsert a dataframe, while specifying the field names that need to be used -for `recordKey => _row_key`, `partitionPath => partition` and `precombineKey => timestamp` +The `hudi-spark` module offers the DataSource API to write (and read) a Spark DataFrame into a Hudi table. There are a number of options available: +**`HoodieWriteConfig`**: + +**TABLE_NAME** (Required)<br> + + +**`DataSourceWriteOptions`**: + +**RECORDKEY_FIELD_OPT_KEY** (Required): Primary key field(s). Nested fields can be specified using the dot notation eg: `a.b.c`. When using multiple columns as primary key use comma seperated notaion, eg: `"col1,col2,col3,etc"`. Single or multiple columns as primary key specified by `KEYGENERATOR_CLASS_OPT_KEY` property.<br> +Default value: `"uuid"`<br> + +**PARTITIONPATH_FIELD_OPT_KEY** (Required): Columns to be used for partitioning the table. To prevent partitioning, provide empty string as value eg: `""`. Specify paritioning/no partitioning using `HIVE_PARTITION_EXTRACTOR_CLASS_OPT_KEY`<br> +Default value: `"partitionpath"`<br> + +**PRECOMBINE_FIELD_OPT_KEY** (Required): When two records have the same key value, the record with the largest value from the field specified here will be choosen.<br> +Default value: `"ts"`<br> + +**OPERATION_OPT_KEY**: The [write operations](#write-operations) to use. Note: this cannot change across writes.<br> Review comment: the flag method is not only applicable when using DeltaStreamer, but also applicable when using HoodieWriteClient/Spark DataSource, in this way, the ingested batch record would consist of both new insert records and delete records, that means you would insert and delete records with one batch, but when setting to `EmptyHoodieRecordPayload`, the ingested batch record is considered to be all delete records. ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: us...@infra.apache.org