openworm / org.geppetto.core

Geppetto core bundle

Geppetto Core Bundle

This project contains the Geppetto core bundle. Inside this module Geppetto keeps all the common interfaces, models, and utilities shared across every other bundle.

For information about how this fits into Geppetto refer to the umbrella project org.geppetto on GitHub.

org.geppetto.core's Issues

Create interface and abstract class for conversion service

• Create IConversionService in core
• Create AConversionService with common implementation if needed

Address security warnings re jackson libraries

See OpenSourceBrain/geppetto-osb#325 for details

Update Javascript API to allow for calling conversion services (writeModel)

• Make proposal for API methods
• Review proposal
• Add API methods and implementation

Change importValue interface in model interpreter

The actual interface is:
Value importValue(ImportValue importValue) throws ModelInterpreterException;

With this interface we are not able to handle multiple runtime projects simultaneously: we are able to find the path from the ImportValue (String path = ((Variable)importValue.eContainer().eContainer()).getPath();) but that path must refer to a model we have not access to, unless we have a reference to the model in the ModelInterpreter.

Proposal new interface:
The importType method has a reference to the geppetto model through a , GeppettoModelAccess parameter. We can add this parameter also to importValue:

Value importValue(ImportValue importValue, , GeppettoModelAccess modelAccess) throws ModelInterpreterException;

Generate new Common Library to go with new model 1.1.0

Add JSONType, SimpleArrayType, MetadataType to the Geppetto Common Library

Stopping and restarting a recording results in NullPointer

Add mechanism to specify authentication configuration for DB+S3

• Find the best place to keep the S3 and DB connection descriptor files

Improve GeppettoRecordingCreator

The quality of code in the GeppettoRecordingCreator can be improved. Also the interactions with the DAT to HDF5 conversion class. Also time is created for each simulationtree while it should only be added for the simulationtree that are part of an aspect for which there is a simulatorconfiguration defined since those are the only aspects that will be creating time.

Global projects state variables filter not working

As per title, the filter does not show up in the casper tests in the development branch.

build fails on running org.geppetto.core as maven install

I attempted to set up org.geppetto.core on my ubuntu 12.10 . Using the blue pill method mentioned here(https://docs.google.com/document/d/1UPfS5UbQ9z61EJ4Uf6saivSy8IR4JHoyQO38FY66ifE/edit) does not allow me to start eclipse as i get the error: binary file cannot be executed
Using the red pill method causes build failure during maven install, screenshot below:

How can i address this issue?
thank you

Add Unit Tests showing server side can read a recording

XXX

Model reader should throw an exception if there is any problem

Figure out why reading values from recordings doesn't start from the first step

Flexible version

Generate new Common Library to go with new model 1.2.0

Add Node and Edge Types to the Geppetto Common Library.

Particle mass should not be on org.geppetto.core

Move it to SPHConstants on org.geppetto.model.sph

Improve performance reading out SkeletonAnimationNode

Tests broken: cannot find HDF5 FileFormat on Windows

Building on windows the following error occurs:

Tests in error:
  HDF5ReaderTest.testExample1:38 » GeppettoExecution java.lang.NullPointerExcept...
  HDF5ReaderTest.testExample2:46 » GeppettoExecution java.lang.NullPointerExcept...
  HDF5ReaderTest.testExample3:54 » GeppettoExecution java.lang.NullPointerExcept...
  HDF5ReaderTest.test:30 » GeppettoExecution java.lang.NullPointerException
  TestConvertACNet2DATToRecordingClass.datToHDF5:136 » GeppettoExecution Cannot ...
  TestConvertBigDATToRecordingClass.datToHDF5:143 » GeppettoExecution Cannot fin...
  TestConvertDATToRecordingClass.datToHDF5:64 » GeppettoExecution Cannot find HD...
  TestConvertMuscleDATToRecordingClass.datToHDF5:136 » GeppettoExecution Cannot ...
  TestGeppettoRecordingsCreator.addSibling:68 » GeppettoExecution Cannot find HD...
  TestGeppettoRecordingsCreator.addSingleInteger:160 » GeppettoExecution Cannot ...
  TestGeppettoRecordingsCreator.addSetInteger:188 » GeppettoExecution Cannot fin...
  TestGeppettoRecordingsCreator.addGrandChildren:130 » GeppettoExecution Cannot ...
  TestGeppettoRecordingsCreator.addInitialVariable:39 » GeppettoExecution Cannot...

Tested on Windows 10, org.geppetto.core v1.0.0

Most likely the Java class unpacking the HDF5 is out of date and doesn't detect newer Windows systems

IWatchableVariableListFeature may need access to experiment results

It used to be that recordings (results) where inside aspects, now they are at the experiment level but in case of this implementation of the interface recordings are needed (as there is no other way to interrogate watchable variables if there's no simulator, we just have a recording). Since they are not at that level anymore, they are not found and it's not populating the simulation tree.

I think we should pass results down to the IWatchableVariableListFeature interface. It is not unreasonable that we may want to get the list of watchable variables from whatever is in the recording (which is what I need and what was happening in the development branch) in other cases too.

For example, take a NeuroML recording with no simulator defined in the experiment, where would we get the watchable variables if not from the recording? And so on for SPH etc.

Add HTMLMetaDataNote

• Add class server side
• Add class client side
• Add serialisation with escape for HTML characters

Update backend to use conversion service

• Create ServiceRegistry to manage what input/output format is connected to the different Services
• Enable backend code to use the conversion simulator if specified by the user (see schema extension) or if needed (the format returned by the ModelInterpreter doesn't match with the format accepted by the specified simulator, look for registered conversion service that can take care of this conversion or return an error if none are present)
• Add API method to the backend to getSupportedOutputs
• Add API mothed to the backend for writeModel

Extend schema to allow for a conversion service to be specified

• Extend geppetto simulation schema to allow for specifying a conversion service in the simulator block.
• Update geppetto simulation file documentation

See below for an example:

<tns:entity>
     <tns:id>hhcell</tns:id>
     <tns:aspect>
        <tns:id>electrical</tns:id>
        <tns:simulator>
            <tns:simulatorId>jLemsSimulator</tns:simulatorId>
    <tns:conversionServiceId>NeuroMLConversionService</tns:conversionServiceId>
         </tns:simulator>
         <tns:model>
            <tns:modelInterpreterId>lemsModelInterpreter</tns:modelInterpreterId>
            <tns:modelURL>https://raw.github.com/openworm/org.geppetto.samples/master/LEMS/SingleComponentHH/LEMS_NML2_Ex5_DetCell.xml</tns:modelURL>
         </tns:model>
     </tns:aspect> 
</tns:entity>

QueryChecker Class Bug

The QueryChecker.java class has a bug, comparing Type(s).

That if statemente fails, when variable 'type' is a CompositeType.

Replace travis with github actions