About Inputs and Outputs
Each step in a process has information that flows into and out of it. Information flowing into a step is called input, and information flowing out of a step is called output. A process itself can also have inputs and outputs, such as when calling a process from a call activity step.
Process data assigned in Designer to flow in and out of steps needs to be mapped to physical data that the underlying services require in order for the process to execute.
Step inputs and outputs are used to define flow signatures, branching and looping logic in the process, data logging for examination at run time, and KPIs. See
About Inputs and Outputs and
About KPIs.
Step inputs and outputs are used to generate the signatures for the generated services that implement the process. If the underlying implementation of the step requires different physical data than this process data, the data must be mapped in the generated flows.
Process data follows a pipeline model, where all data that is input to a step must be output upstream in the process from that step.
Data can therefore enter the process in two ways:
In a receive step, a subscription document can trigger or join the process, and output data for that step and into the pipeline
In an activity step, the step can output new process data into the pipeline
While you can add new inputs to any step, the process will not be valid (e.g., ready to be built) until all step inputs are first selected as outputs of an upstream receive or activity step.
Designer can automatically map inputs and outputs in the following circumstances:
Step A is linked to step B, and the output of step A has the same name as the input of step B
An activity step input name is the same as the document or service input name
A Receive Task output document is the same type as its incoming document type
In all but these cases, you must manually map step input and output data. On the Inputs / Outputs page of the Properties view of the step whose data you want to map, select Edit Data Mapping. Alternatively, right-click the step and select Edit Data Mapping from the context menu.
Note:
By default, when you add a step to a process model and then
Edit Data Mapping, the step in the model is saved if the
Save before building/uploading processes Build and Upload preference is enabled. If it is not enabled,
Designer prompts you to save the step. See
Configuring Build and Upload
Preferences.
When you edit the data mapping for a service,
Designer saves the process definition. It then creates a generated mapping service on the
Integration Server that uses the step inputs and outputs as the new service inputs and outputs. It also puts the invoked service (specified in the
Service property) inside this mapping service, and adds a flow map step.
Designer then opens the Pipeline view to allow you to map the step inputs and outputs to the invoked service inputs and outputs. See
What Does the Pipeline View Contain? in
webMethods Service Development Help.
By default,
Designer does not display generated flow services in the Package Navigator view. You can view them by editing the Package Navigator filter. Click
Filter contents of Navigator and clear the
Hide generated flow services check box in the Choose Elements to Display window. You can also set a preference for this behavior in
Window > Preferences > Software AG > Service Development > Package Navigator. See
Package Navigator Preferences in
webMethods Service Development Help.
After a mapping service has been created, you can right-click its step and select Edit Mapping Service to open it in the flow service editor and update the generated service signature to match the step's current inputs and outputs. Another method is to select Edit Data Mapping on the Inputs/ Outputs page of the Properties view of the step whose data you want to map.
Note:
Mapping services are not intended to contain business logic. The mapping service editor does not prevent you from adding other flows inside a mapping service. However, all reusable business logic should be placed in the invoked service, not in the mapping service.
When you build and upload a process, Designer automatically updates the mapping services of all steps to include the current step signatures, and generates mapping services for all steps that do not yet have them defined.
For more information about data mapping in
Designer, see
Mapping Data in Flow Services in
webMethods Service Development Help.
Related Topics