This is a plugin for the Taverna Workbench and Taverna Command Line which allows export of the provenance of a workflow run according to the W3C PROV-O standard.
This plugin is distributed under the GNU Lesser General Public License 2.1 (LGPL). The source code for this plugin is available at https://github.com/taverna/taverna-prov
Installation of an released version of this plugin does not require compilation, but uses Taverna's plugin mechanism.
You need:
Taverna-PROV is included in Taverna 2.5 and later.
To install any updates to the Taverna-PROV plugin:
Start Taverna workbench
In the menu, click Advanced -> Updates and plugins
If Taverna does not say “An update is available” , click Find updates
For Taverna-PROV databundle, click Update if shown.
Click Close
Exit and Restart Taverna workbench
In order for PROV export to be possible from the Taverna Workbench, click Preferences from the File or Taverna menu, then go to the Preferences for Data and Provenance. Ensure that Provenance capture is enabled (this is the default in 2.5).
If you would like Taverna to keep the provenance data between runs of the workbench (in order to export at a later time), then you need to also untick to disable In memory storage; note however that this may slow down executions of some workflows.
Open and run a workflow to export provenance for. As a quick example, try the very simple helloanyone.t2flow from myExperiment (included in the example/
folder of this source code).
To save a Taverna-PROV databundle containing the workflow run provenance:
.bundle.zip
The Taverna documentation has general information about how to use the Taverna Command line tool.
The Taverna-PROV command line does not support all the output options of the regular Taverna command line, output has to be saved using -outputdir
. The databundle will be saved in a filename which corresponds to the output directory, but with the extension .bundle.zip
.
To enable PROV export for the workflow run, add the parameter -provenance
, which requires the parameter -embedded
or -clientserver
.
Note: -embedded
allows only one concurrent execution at a time due to database locking. To support multiple concurrent runs, start executeworkflow -startdb
separately to start the database server, and use executeworkflow -provenance -clientserver
for the workflow executions.
Example:
stain@biggie-mint ~/src/taverna-prov/example $ executeworkflow -embedded -provbundle helloanyone.bundle.zip -inputvalue name fred helloanyone.t2flow Provenance bundle zip will be saved to: /home/stain/src/taverna-prov/example/helloanyone.bundle.zip stain@biggie-mint ~/src/taverna-prov/example $ mkdir helloanyone.bundle ; cd helloanyone.bundle stain@biggie-mint ~/src/taverna-prov/example/helloanyone.bundle $ unzip ../helloanyone.bundle.zip Archive: ../helloanyone.bundle.zip extracting: mimetype creating: inputs/ inflating: inputs/name.txt creating: outputs/ inflating: outputs/greeting.txt creating: intermediates/ creating: intermediates/3a/ inflating: intermediates/3a/3a82e39d-a537-40cf-91a0-2c89d4a2e62b.txt inflating: workflowrun.prov.ttl inflating: workflow.wfbundle creating: .ro/ creating: .ro/annotations/ inflating: .ro/annotations/workflow.wfdesc.ttl inflating: .ro/annotations/a2f03983-8836-4c36-bfb2-d713d9a1928f.ttl inflating: .ro/manifest.json
The .bundle.zip
file is a RO bundle, which species a structured ZIP file with a manifest (.ro/manifest.json
). You can explore the bundle by unzipping it or browse it with a program like 7-Zip.
This source includes an example bundle and unzipped bundle as a folder. This data bundle has been saved after running a simple hello world workflow.
The remaining text of this section describes the content of the RO bundle, as if it was unpacked to a folder. Note that many programming frameworks include support for working with ZIP files, and so complete unpacking might not be necessary for your application. For Java, the Data bundle API gives a programmating way to inspect and generate data bundles.
The folders inputs/
and outputs/
contain files and folders corresponding to the input and output values of the executed workflow. Ports with multiple values are stored as a folder with numbered outputs, starting from 0
. Values representing errors have extension .err
, other values have an extension guessed by inspecting the value structure, e.g. .png
. External references have the extension .url
- these files can often be opened as “Internet shortcut” or similar, depending on your operating system.
Example listing:
c:\Users\stain\workspace\taverna-prov\example\helloanyone.bundle>ls inputs intermediates mimetype outputs workflow.wfbundle workflowrun.prov.ttl c:\Users\stain\workspace\taverna-prov\example\helloanyone.bundle>ls outputs greeting.txt c:\Users\stain\workspace\taverna-prov\example\helloanyone.bundle>cat outputs/greeting.txt Hello, John Doe
The file workflowrun.prov.ttl
contains the PROV-O export of the workflow run provenance (including nested workflows) in RDF Turtle format.
This log details every intermediate processor invocation in the workflow execution, and relates them to inputs, outputs and intermediate values.
Example listing:
c:\Users\stain\workspace\taverna-prov\example\helloanyone.bundle>cat workflowrun.prov.ttl | head -n 40 | tail -n 8 <#taverna-prov-export> rdf:type prov:Activity ; prov:startedAtTime "2013-11-22T14:01:02.436Z"^^xsd:dateTime ; prov:qualifiedCommunication _:b1 ; prov:endedAtTime "2013-11-22T14:01:03.223Z"^^xsd:dateTime ; rdfs:label "taverna-prov export of workflow run provenance"@en ; prov:wasInformedBy <http://ns.taverna.org.uk/2011/run/385c794c-ba11-4007-a5b5-502ba8d14263/> ;
See the provenance graph for a complete example. The provenance uses the vocabularies PROV-O, wfprov and tavernaprov.
Intermediate values are stored in the intermediates/
folder and referenced from workflowrun.prov.ttl
Intermediate value from the example provenance:
<http://ns.taverna.org.uk/2011/data/385c794c-ba11-4007-a5b5-502ba8d14263/ref/d588f6ab-122e-4788-ab12-8b6b66a67354> tavernaprov:content <intermediates/d5/d588f6ab-122e-4788-ab12-8b6b66a67354.txt> ; wfprov:describedByParameter <http://ns.taverna.org.uk/2010/workflowBundle/01348671-5aaa-4cc2-84cc-477329b70b0d/workflow/Hello_Anyone/processor/Concatenate_two_strings/in/string1> ; wfprov:describedByParameter <http://ns.taverna.org.uk/2010/workflowBundle/01348671-5aaa-4cc2-84cc-477329b70b0d/workflow/Hello_Anyone/processor/hello/out/value> ; wfprov:wasOutputFrom <http://ns.taverna.org.uk/2011/run/385c794c-ba11-4007-a5b5-502ba8d14263/process/bbaedc02-896f-491e-88bc-8dd350fcc73b/> .
Here we see that the bundle file intermediates/d5/d588f6ab-122e-4788-ab12-8b6b66a67354.txt
contains the output from the “hello” processor, which was also the input to the “Concatenate_two_strings” processor. Details about processor, ports and parameters can be found in the workflow definition.
Example listing:
c:\Users\stain\workspace\taverna-prov\example\helloanyone.bundle>ls intermediates/d5 d588f6ab-122e-4788-ab12-8b6b66a67354.txt c:\Users\stain\workspace\taverna-prov\example\helloanyone.bundle>cat intermediates/d5/d58* Hello,
Note that “small” textual values are also included as cnt:chars
in the graph, while the referenced intermediate file within the workflow bundle is always present.
<intermediates/d5/d588f6ab-122e-4788-ab12-8b6b66a67354.txt> rdf:type cnt:ContentAsText ; cnt:characterEncoding "UTF-8"^^xsd:string ; cnt:chars "Hello, "^^xsd:string ; tavernaprov:byteCount "7"^^xsd:long ; tavernaprov:sha512 "cf83e1357eefb8bdf1542850d66d8007d620e4050b5715dc83f4a921d36ce9ce47d0d13c5d85f2b0ff8318d2877eec2f63b931bd47417a81a538327af927da3e"^^xsd:string ; tavernaprov:sha1 "f52ab57fa51dfa714505294444463ae5a009ae34"^^xsd:string ; rdf:type tavernaprov:Content .
The file workflow.wfbundle
is a copy of the executed workflow in SCUFL2 workflow bundle format. This is the format which will be used by The file workflow.wfbundle
contains the executed workflow in Taverna 3.
You can use the SCUFL2 API to inspect the workflow definition in detail, or unzip the workflow bundle to access the complete workflow definition files as RDF/XML.
Usually it is sufficient to inspect the simpler file .ro/annotations/workflow.wfdesc.ttl
, which contains the abstract structure (but not all the implementation details) of the executed workflow, in RDF Turtle according to the wfdesc ontology.
c:\Users\stain\workspace\taverna-prov\example\helloanyone.bundle>cat .ro/annotations/workflow.wfdesc.ttl | head -n 20 @base <http://ns.taverna.org.uk/2010/workflowBundle/01348671-5aaa-4cc2-84cc-477329b70b0d/workflow/Hello_Anyone/> . @prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> . @prefix xsd: <http://www.w3.org/2001/XMLSchema#> . @prefix owl: <http://www.w3.org/2002/07/owl#> . @prefix prov: <http://www.w3.org/ns/prov#> . @prefix wfdesc: <http://purl.org/wf4ever/wfdesc#> . @prefix wf4ever: <http://purl.org/wf4ever/wf4ever#> . @prefix roterms: <http://purl.org/wf4ever/roterms#> . @prefix dc: <http://purl.org/dc/elements/1.1/> . @prefix dcterms: <http://purl.org/dc/terms/> . @prefix comp: <http://purl.org/DP/components#> . @prefix dep: <http://scape.keep.pt/vocab/dependencies#> . @prefix biocat: <http://biocatalogue.org/attribute/> . @prefix : <#> . <processor/Concatenate_two_strings/> a wfdesc:Process , wfdesc:Description , owl:Thing , wf4ever:BeanshellScript ; rdfs:label "Concatenate_two_strings" ; wfdesc:hasInput <processor/Concatenate_two_strings/in/string1> , <processor/Concatenate_two_strings/in/string2> ; wfdesc:hasOutput <processor/Concatenate_two_strings/out/output> ; wf4ever:script "output = string1 + string2;" .
If the workflow was executed in Taverna 2 (which is the case when using this plugin), then the unzipped workflow.wfbundle
will contain a copy of the original .t2flow
within its history
folder:
stain@biggie-mint ~/src/taverna-prov/example/helloanyone.bundle $ mkdir workflow ; cd workflow ; unzip ../workflow.wfbundle Archive: ../workflow.wfbundle extracting: mimetype creating: META-INF/ inflating: META-INF/manifest.xml creating: annotation/ inflating: annotation/705d1f32-ea94-4a5b-b8fb-1cc1a5023427.ttl inflating: annotation/0f1641bd-49ef-41bd-9e4b-e20cc2bf37bb.ttl inflating: annotation/d007623d-cd94-4888-a583-e8d97953a159.ttl inflating: annotation/e9eb11fa-fc63-4ddc-92ca-ec6921af1667.ttl inflating: annotation/c5e91716-2f9f-4e7c-a692-8f58dba265e3.ttl creating: history/ inflating: history/01348671-5aaa-4cc2-84cc-477329b70b0d.t2flow creating: workflow/ inflating: workflow/Hello_Anyone.rdf creating: profile/ creating: profile/unspecified/ creating: profile/unspecified/configuration/ inflating: profile/unspecified/configuration/Concatenate_two_strings.json inflating: profile/unspecified/configuration/Concatenate_two_strings-proc.json inflating: profile/unspecified/configuration/hello.json inflating: profile/unspecified/configuration/hello-proc.json inflating: profile/unspecified.rdf inflating: workflowBundle.rdf inflating: META-INF/container.xml stain@biggie-mint ~/src/taverna-prov/example/helloanyone.bundle/workflow $ head history/01348671-5aaa-4cc2-84cc-477329b70b0d.t2flow <workflow xmlns="http://taverna.sf.net/2008/xml/t2flow" version="1" producedBy="unspecified"><dataflow id="01348671-5aaa-4cc2-84cc-477329b70b0d" role="top"><name>Hello_Anyone</name><inputPorts><port><name>name</name><depth>0</depth><granularDepth>0</granularDepth><annotations><annotation_chain encoding="xstream"><net.sf.taverna.t2.annotation.AnnotationChainImpl xmlns=""> <annotationAssertions> <net.sf.taverna.t2.annotation.AnnotationAssertionImpl>
The UUID in the t2flow filename (01348671-5aaa-4cc2-84cc-477329b70b0d
above) should match the executed wfbundle URI as seen in the provenance
http://ns.taverna.org.uk/2010/workflowBundle/01348671-5aaa-4cc2-84cc-477329b70b0d/
Not yet implemented: a specific link to the history/*.t2flow
file within the wfbundle.
Example SPARQL query from test.sparql:
PREFIX prov: <http://www.w3.org/ns/prov#> PREFIX wfdesc: <http://purl.org/wf4ever/wfdesc#> PREFIX wfprov: <http://purl.org/wf4ever/wfprov#> PREFIX tavernaprov: <http://ns.taverna.org.uk/2012/tavernaprov/> PREFIX cnt: <http://www.w3.org/2011/content#> PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#> PREFIX wf4ever: <http://purl.org/wf4ever/wf4ever#> SELECT DISTINCT ?paramName ?name ?value ?ended ?script WHERE { ?greeting tavernaprov:content <outputs/greeting.txt> . ?greeting prov:wasGeneratedBy ?concatenate . ?concatenate prov:endedAtTime ?ended ; wfprov:wasPartOfWorkflowRun ?run ; wfprov:describedByProcess ?plan . ?concatenate wfprov:usedInput ?input . ?input tavernaprov:content ?name . OPTIONAL { ?name cnt:chars ?value } . OPTIONAL { ?plan wf4ever:script ?script } . ?input wfprov:describedByParameter ?param . ?plan wfdesc:hasInput ?param . OPTIONAL { ?param rdfs:label ?paramName } . }
This query will be starting with the data ?greeting
which content is represented by the existing output file outputs/greeting.txt, and the remaining query tries to find which input or upstream values it has effectively been derived from.
To do this, we find the ?concatenate
process run that generated the greeting, and ask when it ?ended
. We also look up its ?plan
, which should match the process identifier within the workflow definition.
We then look at the ?input
s used by the ?concatenate
process run (this should give two matches as the “Concatenate string” processor takes two arguments). We look up their ?content
(a file witin this bundle), in addition to its textual ?value
(optionally, as this is only included in the graph for small non-binary content).
Now we do a lookup of the ?script
behind the defined ?plan
- this is optional because not all processors have scripts (it might be a web service). This information is extracted from the .ro/annotations/workflow.wfdesc.ttl file which must also be parsed before querying.
Lastly we look up the pararameters which describes ?input
, filtered by the ones bound to the processor ?plan
(thus avoiding the workflow inputs “name”), and include their ?paramName
- each result should therefore show the input port name and value.
Query using rdfproc (apt-get install redland-utils
):
stain@biggie-mint ~/src/taverna-prov/example/helloanyone.bundle $ rdfproc test parse workflowrun.prov.ttl turtle rdfproc: Parsing URI file:///home/stain/src/taverna-prov/example/helloanyone.bundle/workflowrun.prov.ttl with turtle parser stain@biggie-mint ~/src/taverna-prov/example/helloanyone.bundle $ rdfproc test parse .ro/annotations/workflow.wfdesc.ttl turtle rdfproc: Parsing URI file:///home/stain/src/taverna-prov/example/helloanyone.bundle/.ro/annotations/workflow.wfdesc.ttl with turtle parser stain@biggie-mint ~/src/taverna-prov/example/helloanyone.bundle $ rdfproc test2 query sparql - "$(cat ../test.sparql)" rdfproc: Warning - URI file:///home/stain/src/taverna-prov/example/helloanyone.bundle/:1: Variable run was bound but is unused in the query rdfproc: Query returned bindings results: result: [, paramName="string2", name=<file:///home/stain/src/taverna-prov/example/helloanyone.bundle/inputs/name.txt>, value="fred"^^<http://www.w3.org/2001/XMLSchema#string>, ended="2014-05-28T11:49:43.711+01:00"^^<http://www.w3.org/2001/XMLSchema#dateTime>, script="output = string1 + string2;"] result: [, paramName="string1", name=<file:///home/stain/src/taverna-prov/example/helloanyone.bundle/intermediates/3a/3a82e39d-a537-40cf-91a0-2c89d4a2e62b.txt>, value="Hello, "^^<http://www.w3.org/2001/XMLSchema#string>, ended="2014-05-28T11:49:43.711+01:00"^^<http://www.w3.org/2001/XMLSchema#dateTime>, script="output = string1 + string2;"] rdfproc: Query returned 2 results
This shows that the string2
input port was bound to value "fred"
(content of inputs/name.txt
) in the process run that generated outputs/greeting.txt
. Next, string
was bound to "Hello, "
(content of intermediates/3a/3a82..62b.txt
). The executed script was output = string1 + string2
.
Note - you do not normally need to build from source code; unless you are modifying the source code, the “Taverna Prov” plugin included in Taverna's default plugins is the preferred way of using Taverna-PROV.
You need:
To compile, run mvn clean install
On first run this will download various open-source third-party libraries from Maven repositories, including required modules of Taverna 2.4. These JARs, and the compiled JARs from this source code, are installed into $HOME/.m2/repository/
by default. Depending on your network connection, this first run might take about 5 minutes to complete.
Example compilation:
stain@vmint:~/src/taverna-prov$ mvn clean install [INFO] Scanning for projects... [INFO] ------------------------------------------------------------------------ [INFO] Reactor Build Order: [INFO] [INFO] Taverna PROV [INFO] Taverna PROV W3 Provenance ontology Elmo bindings [INFO] Taverna PROV export [INFO] Taverna PROV activity UI bindings [INFO] Taverna PROV plugin [INFO] PROV Taverna command line [INFO] [INFO] ------------------------------------------------------------------------ [INFO] Building Taverna PROV 2.0.0-SNAPSHOT [INFO] ------------------------------------------------------------------------ (..) Downloading: http://www.mygrid.org.uk/maven/repository/net/sf/taverna/t2/taverna-workbench/workbench-dev/2.5.0/workbench-dev-2.5.0.pom Downloaded: http://www.mygrid.org.uk/maven/repository/net/sf/taverna/t2/taverna-workbench/workbench-dev/2.5.0/workbench-dev-2.5.0.pom (3 KB at 5.4 KB/sec) (..) [INFO] Installing /home/stain/src/taverna-prov/prov-taverna-cmdline/target/prov-taverna-cmdline-2.0.0-SNAPSHOT-tests.jar to /home/stain/.m2/repository/org/purl/wf4ever/provtaverna/prov-taverna-cmdline/2.0.0-SNAPSHOT/prov-taverna-cmdline-2.0.0-SNAPSHOT-tests.jar [INFO] ------------------------------------------------------------------------ [INFO] Reactor Summary: [INFO] [INFO] Taverna PROV ...................................... SUCCESS [4.109s] [INFO] Taverna PROV W3 Provenance ontology Elmo bindings . SUCCESS [1.55.270s] [INFO] Taverna PROV export ............................... SUCCESS [45.873s] [INFO] Taverna PROV activity UI bindings ................. SUCCESS [1.56.572s] [INFO] Taverna PROV plugin ............................... SUCCESS [15.028s] [INFO] PROV Taverna command line ......................... SUCCESS [1:12.093s] [INFO] ------------------------------------------------------------------------ [INFO] BUILD SUCCESS [INFO] ------------------------------------------------------------------------ [INFO] Total time: 5:54.000s [INFO] Finished at: Wed Jul 18 12:08:13 BST 2012 [INFO] Final Memory: 72M/274M [INFO] ------------------------------------------------------------------------
Note that to work with Taverna 2's plugin system, the build is specific for a particular Taverna version (but works for any edition). To build this plugin for a different version of Taverna, modify the <parent>
section of pom.xml
to match the Taverna release, e.g. <version>2.4.0</version>
.
In order for Taverna to find your install of the plugin rather than download from the official plugin site, you will need to manually edit plugins/plugins.xml
of the Taverna home directory (run Taverna once to create the file) to include this section right before the final </plugins:plugins>
. The
Note that you will get two copies of the “Save provenance” button, one for the official version included from workbench-plugin
, and one for the one you built.
<plugin> <provider>org.purl.wf4ever</provider> <identifier>org.purl.wf4ever.provtaverna.prov-taverna-plugin</identifier> <version>2.0.0-SNAPSHOT</version> <name>Taverna PROV plugin</name> <description/> <enabled>true</enabled> <repositories> <repository>file:///home/johndoe/.m2/repository/</repository> <repository>http://www.mygrid.org.uk/maven/repository/</repository> <repository>http://uk.maven.org/maven2/</repository> <repository>http://repo.aduna-software.org/maven2/releases/</repository> <repository>http://repo.aduna-software.org/maven2/snapshots/</repository> <repository>http://repo.aduna-software.org/maven2/ext/</repository> <repository>http://repository.aduna-software.org/maven2/</repository> <repository>http://repository.aduna-software.org/maven2-snapshots/</repository> <repository>http://people.apache.org/repo/m2-snapshot-repository/</repository> <repository>http://nexus.codehaus.org/snapshots/</repository> <repository>http://repository.jboss.org/maven2/</repository> </repositories> <profile> <dependency> <groupId>org.purl.wf4ever.provtaverna</groupId> <artifactId>prov-taverna-ui</artifactId> <version>2.0.0-SNAPSHOT</version> </dependency> <dependency> <groupId>junit</groupId> <artifactId>junit</artifactId> <version>4.8.2</version> </dependency> </profile> <compatibility> <application> <version>2.5.0</version> </application> </compatibility> </plugin>
Modify the line:
<repository>file:///home/johndoe/.m2/repository/</repository>
to match your $HOME/.m2/repository. On Windows this path should look like:
<repository>file:///C:/Users/johndoe/.m2/repository/</repository>
Take care that the <version>
of the <plugin>
and prov-taverna-ui
matches the <version>
specified in the pom.xml
of this source code (typically 1.x-SNAPSHOT
).
If you are building for a different Taverna, check that the <application><version>
matches the Taverna version (compare with the existing plugins)
Start Taverna Workbench.
You only need to do this plugin installation once - if you later recompile the source code without changing the version numbers, next start of Taverna will use the newer JARs from mvn clean install
as Taverna will prefer accessing $HOME/.m2/repository
. You might however need to delete the repository/org/purl/wf4ever
folder of the
Taverna home directory
if Taverna has downloaded the SNAPSHOT versions from the myGrid repository instead (for instance because you used the wrong <repository>
path or did not update <version>
).
To make the plugin installation system-side (ie. multiple UNIX users or to make a customized Taverna distribution), copy plugins/
and repository
with overwrite onto the Taverna installation directory. (See Installation for Taverna Command Line above)