SchemaMapper

Converts the existing schema (data model) of features to a new structure, based on mappings defined in an external lookup table.

This technique is very useful when the mappings are potentially complex or when they need to be maintained by someone who is not familiar with FME. Using an external lookup table to define these mappings simplifies the completion of these tasks.

The schema mapping lookup table, used by the SchemaMapper transformer, defines a series of conditions that are to be met (filters), and a series of actions that will be executed when the conditions are met.

The lookup table may come in different formats such as: a comma-separated or plain text file; a spreadsheet (Excel or Google); or a database such as Oracle, PostGRES, Informix, SQL Server, etc.

If an incoming feature matches the rules in any row of the table, the potential actions of the transformer on that feature are listed by type as follows:

Map Feature Types:

Feature types – as defined by fme_feature_type – are mapped from their existing value, to one which defines the new feature type. For example:

Roads → Center Lines

Map Attributes:

One or more attributes on the feature are mapped by renaming their existing attribute name (s) to a new one. For example:

Name_of_Road → RoadName

Set New Attributes:

A new attribute is created whose name and value are defined in a lookup table.

Output Ports

Features that have an action carried out on them are output via the Mapped port. Otherwise, they are output via the Unmapped port.

In addition to the feature type, the geometry of each feature is left untouched.

Features output through the Mapped port will receive a new attribute that specifies which rows from the schema mapping table that were applied to the feature. The rows are specified as a comma-separated string (for example, '3,4,5'). This attribute is specified by the Selected Row Attribute parameter, and can only be changed from within the Navigator pane. Setting this parameter to a blank value will prevent the attribute from being added to features.

SchemaMapper Uses

You can use the SchemaMapper for:

Domain Mapping: where attributes values can be remapped according to a well-defined domain or lookup table. For example:

• Primary Route → US Highway
• Secondary Route → Interstate Highway
• Primary Route → County Road

Dynamic Translations: where workspaces are created to handle any data structure

Automated Schema Mapping: where the manual connections between source and destination schemas are done automatically using an external lookup table.

Example: Using the SchemaMapper transformer

The following example describes a typical scenario

Feature Type Mapping

Using an external lookup table (.csv file), you can map feature types to simplify a schema by mapping the old feature type to a new feature type:

Old Feature Type	New Feature Type
River	Water
Lake	Water
Canal	Water
Road	Transportation
Railway	Transportation
Airport	Transportation

Attribute Mapping

Using an external lookup table, you can map attributes to new names:

Old Attribute	New Attribute
River_Name	WaterName
River_Alt_Name	AltWaterName
River_Country	WaterCountry
River_Length	WaterSize
River_Owner	WaterOwner

Filters

You can use filters to set your mapping rules. For example, if you want to map an “old pipe type” to a “new pipe type” You can set the rule (filter) according to the size of the pipes:

Old Pipe Type	Pipe Size Attribute	Pipe Size	New Pipe Type
Gas	PipeSize	6	GasSmall
Gas	PipeSize	12	GasMedium
Gas	PipeSize	24	GasLarge
Water	Diameter	6	WaterSmall
Water	Diameter	12	WaterMedium
Water	Diameter	24	WaterLarge
Sewage	PipeDiam	6	SewageSmall
Sewage	PipeDiam	12	SewageMedium
Sewage	PipeDiam	24	SewageLarge

Blank or Null Attribute Value Handling

When creating a filter, the Blank Attribute Value parameter determines how blank or null attribute values are handled:

Ignore	Blank or null attribute values will be ignored, and will not be assigned new values by the filter.
Don't Ignore	Blank or null attribute values will be handled like any other attribute value. They may be referenced in either the source or destination values as an empty string ("").

Expressions

Expressions are supported in Filter Features (FilterValue) and Set New Attributes (TargetValue) actions.

These are defined in the external lookup table, and are identified with the keyword prefix fmeexpression=.

Filtering

This example lookup table will filter features based on the value of the attribute attr. If:

• less than or equal to 15, or

• greater than or equal to 450

the feature is considered mapped and is passed to the next Action (or output via the Mapped port, if there are no further actions).

Adding Attributes

Using the same example lookup table, if the action is Set New Attributes, features will receive two new attributes - attr_scaled10 and attr_scaled100, both with values calculated by expressions.

Example Using Feature Type Mapping and Filtering

You can use the SchemaMapper wizard to define the filters and the type of mapping that you need. You accomplish this task by creating actions to be executed on the source schema.

To open the SchemaMapper wizard, click the Properties button of the transformer. The SchemaMapper Parameters dialog opens.

1. Specify the format and location of the schema mapping table (lookup table) to be used. You may want to edit format parameters but you can use the defaults that FME Workbench provides. Click Next.
2. If a dataset has more than one table, specify the table to be used. Click Next.
3. On the Create Actions pane, click Add and select the action type. For this example, we'll select Feature Type Map.

4. Select the Source and Destination Feature Type Fields. You can show/hide the table columns, and refresh the contents.

5. Click OK to return to the Create Actions pane, which now shows the Feature Type Map action you just set.

6. To add another action, select from the Add drop-down menu. In this example, we'll select Filter to display the Create Filter dialog.

7. Select the table columns in the Attribute Name and Value Fields. By default, Blank Attribute Values are ignored but you can choose to include them.
8. Click OK to return to the Create Actions dialog. The Action column now shows two actions: Map Feature Types and Filter Features.

9. Click Finish to set the SchemaMapper parameters.

More Information

• For a good basic introduction to the SchemaMapper, see Crouching Schema Hidden Dragon.
• This FME Community article contains descriptions and examples.

Editing Transformer Parameters

Using a set of menu options, transformer parameters can be assigned by referencing other elements in the workspace. More advanced functions, such as an advanced editor and an arithmetic editor, are also available in some transformers. To access a menu of these options, click beside the applicable parameter. For more information, see Transformer Parameter Menu Options.

Defining Values

There are several ways to define a value for use in a Transformer. The simplest is to simply type in a value or string, which can include functions of various types such as attribute references, math and string functions, and workspace parameters. There are a number of tools and shortcuts that can assist in constructing values, generally available from the drop-down context menu adjacent to the value field.

Dialog Options - Tables

Transformers with table-style parameters have additional tools for populating and manipulating values.

FME Community

The FME Community is the place for demos, how-tos, articles, FAQs, and more. Get answers to your questions, learn from other users, and suggest, vote, and comment on new features.

Search for all results about the SchemaMapper on the FME Community.