The default view of the Workflow Explorer shows the workflow as a hierarchy of entities that it consists of. This includes all service services, workflow inputs and outputs, data connections and coordination links. Some items may be expanded to show properties of child items such as input and output ports on individual processors.
All workflow components from workflow input/output ports, services, data and control links may have additional information regarding their use and validity. This may be obtained by right-clicking on a given component and selecting Show details or Show validation report.
Right-clicking on a workflow component provides access to various configuration options. In general, the right-click menu on an entity will expose at least 'delete from workflow' functionality and may provide more in terms of configuration of the selected component:
Right-clicking on the top level node of the workflow, that is denoted by the icon, will provide options for inserting new input/output ports and different types of processors:
New instances of workflow inputs and outputs can be generated by right-clicking on these nodes:
New inputs and outputs are made by providing a name of the new workflow input or output and defining whether it is a single value or a list of values. In contrast, the options available by right-clicking on an actual workflow input or output include renaming the port and its deletion.
Right-clicking on a processor provides access to a number of configurable options. The running of all processors can be configured so that it can be parallelised, invoked repeatedly until a certain condition is met and retried if the tool it is invoking fails.
In addition, the configurable options available to a processor is also dependent on the type of processor. Information on how different types of processors may be configured can be found in the Service types section of this manual.
Each data link (between workflow inputs, processors and workflow outputs) in the workflow appears under 'Data links' node in the Workflow Explorer. These data links correspond to the flow of data that occurs between workflow components which are represented in the workflow diagram as arrows. The data links display is largely informational although data links may also be deleted from the right click context menu. Note that any data links connected to a processor are deleted if that processor is also deleted. There are no 'hanging links' in a workflow model.
Naming for the data links is in the form of the processor name and port name for the output or data source, separated by a colon, then an arrow and the same syntax for the sink or input port it links to. In the case of workflow inputs and outputs there is no processor name, just the name of the output or input.
As with the data links these are primarily for information but also allow removal of individual control links. The naming syntax consists of the processor acting as the gate condition, two colons then the name of the processor being controlled by it. So, 'foo::bar' represents a condition stating that 'bar' may not start until 'foo' has finished. Control links are represented in the workflow diagram by a symbol.
For ease of use, we suggest that you build your workflows using the interactive workflow diagram window. However, it is also possible to construct workflows using the Workflow Explorer by making use of the options to connect workflow input and output ports with those emanating from workflow processors.
In the following example, we will build a simple workflow which concatenates two strings and consists of one processor, two input ports and one output port . To begin, select the "Concatenate two strings" local Java processor from the Service Panel. You can find this processor by typing "concatenate" into the search box. Simply drag this processor into the workflow diagram to use it in a workflow.
Right click on the Workflow input ports node and add two new workflow inputs called str1 and str2. You can leave the type as single value. Now create a workflow output and name it output by right clicking on the Workflow outputs node. Your Taverna workbench should now look as follows:
The Workflow Explorer can now be used to connect these four components together to create a workflow. Firstly, right-click on the str1 workflow input and connect it to the string1 input port of the Concatenate_two_strings processor. Secondly, right-click on the str2 workflow input and connect it to the string2 input port of the Concatenate_two_strings processor. Your Taverna Workbench should now look as follows:
To complete the workflow, you need to connect the output port of the Concatenate_two_strings processor with the workflow output port. Again, you can achieve this using the Workflow Explorer panel by right-clicking on the output port of the Concatenate_two_strings processor and connecting it to the workflow output:
The completed workflow should now look like the one below with the ports of the processor displayed: