KeplerWorkflows
The keplerworflows project under GForge is used for exchanging
Kepler workflows among ITM users.
Under this project, ITM developers should store the XML file describing the
Kepler workflow.
The workflow should not be dependent on any other file. To be
reproducible, public ITM workflows must use only ITM actors from the public
kepleractors repository with a unique naming convention.
Input datasets used by the workflow must be copied from the private to the public ITM database (email to Fréderic Imbeaux for the moment).
Structure of the workflows repository
The keplerworkflows repository is under SVN under the Gforge
project KeplerWorkflows.
It is recommended to use the tags directory for release versions
and trunk for development versions: only public versions of
trunk/datastructure_version/ should be tagged as tags/datastructure_version/.
To check out the repository please do
svn checkout https://gforge6.eufus.eu/svn/keplerworkflows target_dir
The workflow is organised in the following way:
keplerworkflows/trunk/datastructure_version/IMP/physical_topic/workflowname
Currently the keplerworkflows repository has the following subdirectories for the UAL release versions:
- 4.07b - 4.07c - 4.08a - 4.08b - 4.09a
Each UAL release version hosts the following subdirectories for the ITM projects:
- amns - edrg - isip - imp12 - imp3 - imp4 - imp5 - ism
Below these the following physics topics are currently defined:
imp12:
- fixed_boundary_equilibrium
- free_boundary_equilibrium
- linear_MHD
Procedure to put a workflow in the workflows repository
The script put_workflow allows to commit a workflow (in fact, the xml file created by Kepler when the workflow is saved) in the svn repository https://gforge6.eufus.eu/svn/keplerworkflows.Pre-requisites
- Have a copy of Kepler installed in your environment
- Run the script ITMv1 to specify the working kepler directory (private) and set the environment variables (for the UAL)
- Go into a directory in which you have write permission
How to commit a workflow in the repository
Copy into the current directory or locate the xml file describing the workflow. This xml file is created by Kepler when you save a designed workflow. The file does not have to be in your own directory. Only read permission is needed.
Usage :
put_workflow [options] –d svnpath [path/]workflownamesvnpath is mandatory after -d to point out the target location of the workflow in the svn repository.
path is only necessary if the xml file of the workflow is not located in the current directory.
Example :
put_workflow –d trunk/4.08b/isip/examples/loopexample ./loopexample/loopexample.xmlsvnpath is isip/examples/loopexample
path is ./loopexample
workflowname is loopexample.xml
The script put_workflow copies the workflow xml file into the folder svnpath under the KeplerWorklows SVN repository.
NB :The subdirectory of trunk or tags corresponding to the data structure version must match the current $DATAVERSION environment variable
Additional options
- -h : display usage information
- -c : create the target directory in the svn repository keplerworkflows if it does not exist
- -m : message for svn import
Procedure to get a workflow from the workflows repository
The script import_svn_workflow allows you to copy a workflow from the svn repository https://gforge6.eufus.eu/svn/keplerworkflows in your workspace.Pre-requisites
- Have a copy of Kepler installed in your environment
- Run the script ITMv1 to specify the working kepler directory (private) and set the environment variables (for the UAL)
- Go into a directory in which you have write permission
How to import workflow from svn repository
- You do not know the location of the workflow in the repository
Usage:
import_svn_workflow -R workflowname
Example:
import_svn_workflow –R loopexample.xml
The script displays the location(s) of the workflow in the svn repository (in trunk or tags subdirectories) and the user can choose the item he wants to import. - You know the location of the workflow in the repository
Usage:
import_svn_workflow -d path_to_the_workflow_in_the_svn_repository workflowname
Example:
import_svn_workflow -d trunk/4.08b/isip/examples/loopexample loopexample.xml
path_to_the_workflow_in_the_svn_repository is trunk/4.08b/isip/examples/loopexample
workflowname is loopexample.xml
-
Import the actors used by the workflow
If the option -a is added to the previous command lines, the script import_svn_workflow also imports the actors used by the workflow, if they exist in the KeplerActors SVN repository.newline/> For each actor, the script import_svn_workflow launch the command
import_actor -R actor_name
and search in the Kepleractors SVN repository if the actor exists, and if exists, import it in the private copy of Kepler.For more details about the script import_actor, see the section How to import an actor from svn repository.
NB :The subdirectory of trunk or tags corresponding to the data structure version must match the current $DATAVERSION environment variable
Additional options
- -h : display usage information
- -a : import the actors used by the workflow from the svn repository
- -f : ignored if -a not present; force import of an already present actor
- -p : ignored if -a not present; partial import: copy files but do not compile Kepler"
- -s : ignored if -a not present; skip import if the actor is already present
last update: 2019-01-31 by g2dpc