To build mappings, you must create a data design project in the Data Project Explorer view. The data design project contains the data models, the SQL scripts, the XML schemas, and the mapping models that you use to design and model enterprise relationships.
The Data Project Explorer view contains two kinds of SQL data models: those that are enabled with InfoSphere Information Server functions, and those that are non-federated SQL models. The mapping editor can use and deploy both types of models. In addition, the mapping editor can adjust the elements that are generated into the InfoSphere Information Server enabled SQL model to be specific to the InfoSphere Information Server environment.
The mappings that you create are stored in mapping models. Each mapping model is represented by one mapping specification language file (MSL). The data source can be any physical database model (DBM file) or logical data model (LDM file) in your data design project. The target can be a physical data model (DBM file), a logical data model (LDM file), or an XML schema (XSD file).
The mapping model can be used by tools that generate scripts to create scripts to transform and filter data from mapping model compliant sources to mapping model compliant targets. These script generation tools can capture the mapping semantics of the mapping model and render the results of the model into various types of runtime objects.
Script generation is not supported for logical data model mapping models. Mappings between logical data models and other sources or targets is used for publishing purposes only.
The mapping editor displays source and target data models in a tree format. You can identify elements or columns that have corresponding values between the source and target. The discover function can help you to find the matching elements automatically so that you do not have to specify them manually. The discover function can suggest relationships that might not be initially apparent or that might be difficult to see on the same screen. You can also manually identify elements between a source and target that match.
You can accept or reject mappings that the discover function proposes.
You can refine complex one-to-one and many-to-one relationships by providing transformation functions that modify source element values before they are applied to the target. Each transformation function applies to a single mapping. The function can optionally take one or more of the inputs (source elements) in the mapping but generates a single output (target element). Therefore, a transformation function is associated with only one target element. One-to-many type relationships must be specified as multiple one-to-one relationships, each with their own transformation function as necessary.