ETL Metadata Injection

ETL Metadata Injection is a transform that allows you to modify the properties of a specified "template" pipeline. The transforms that will be executed are chosen, but their properties and settings are "injected" at runtime.

This transform uses a mapping between information that the transform reads from source transforms in the form of one or more rows of data in source fields. These fields can then be mapped to target transform fields in the specified "template" pipeline.

How can you specify how a field to be injected shows up in the ETL Metadata Injection dialog? You can use the @HopMetadataProperty annotation. For more example on how this is used to serialize metadata, see the Metadata Serialization page.

By default, if a field is annotated, it’s going to be visible. You can specify that a field has a specific injectionKey for backward compatibility with older forms or annotations.

For lists, you can use the injectionGroupKey.

Avoid giving a File object in a list of files a key or injectionKey value of name if you would do the same in a Field class in a list of fields.

To allow for meaningful descriptions of the injectable fields to show up in the ETL Metadata Injection transform dialog you an specify properties injectionKeyDescription and injectionGroupDescription. The specified strings reference an entry to the message bundles in the transform, e.g. messages_en_US.properties and their localized variants.

The injection system will try to use the already specified methods for XML/JSON serialization but there are also these options: