Distributed with Taverna is a utility script called
executeworkflow.sh or
executeworkflow.bat that allows a workflow to be executed
outside of the Taverna workbench.
In Mac OS X, this script is located within the Taverna application.
So if Taverna was installed to /Applications, you can
invoke the script as:
/Applications/Taverna.app/Contents/MacOS/executeworkflow.sh
The minimum requirement is to provide the name, or URL, to the
workflow scufl document. The default behaviour is to save the workflow
results into a directory with the same name of the workflow appended with
_output. Within this directory is store the progress
report of the workflow, and additional directories that contain the actual
results. These directories are named after the workflow output names. Be
aware that existing results are automatically overwritten if they
exist.
Supplying inputs and modifying the default behaviour can be achieved with the following options.
-help displays this list of options. |
-input name
filename
load a single value for the named input port from a file or URL. |
-inputdoc filename load inputs from a
Baclava XML document. |
-outputdoc filename save the outputs to a
new Baclava XML document. |
-output directory save outputs as files
in directory. |
-report filename save the progress
report as an XML document. |
If outputs are stored as baclava data documents using
-outputdoc, these results can be viewed graphically using
the data viewer utility. This is used using the
dataviewer script that can be found in the same location
as executeworkflow.
Example 1.1. Executeworkflow examples
executeworkflow myworkflow.xml executeworkflow http://somewhere.org/myworkflow.xml executeworkflow -input myinput myinput.txt myworkflow.xml executeworkflow -input myinput http://somewhere.org/myinput.txt myworkflow.xml executeworkflow -inputdoc inputs.xml myworkflow.xml executeworkflow -output result_run1 myworkflow.xml
Note that while the workflow is running you won't get much feedback
about the progress of the workflow. Stack traces might or might not indicate
an error, for instance a workflow with a processor that intentionally fails
would produce a stack trace from the enactor even if the workflow itself did
not fail. For more detailed log messages, explore
log/taverna.log in your Taverna home directory.
If you are interested in executing a workflow from within your own application you should also read the section Programmatically executing a workflow.