package installation: the usage of a gh token is probably not needed
The 0.1.4 package documentation explains that a github auth token must be pass to install the package.
I don't understand why we need the token as all resources used by the package are public. If this is required, please document why.
as we are using the Java API and not the CLI, we should add a try catch bock when calling the 'layout generation from csv' method and at least print the stack trace
java.lang.NullPointerException
at io.process.analytics.tools.bpmn.generator.input.CSVtoBPMN.assignIncomingAndOutgoingReferences(CSVtoBPMN.java:113)
at io.process.analytics.tools.bpmn.generator.input.CSVtoBPMN.readFromCSV(CSVtoBPMN.java:51)
at io.process.analytics.tools.bpmn.generator.BPMNLayoutGenerator.generateLayoutFromCSV(BPMNLayoutGenerator.java:48)
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
at java.base/java.lang.reflect.Method.invoke(Method.java:566)
at RJavaTools.invokeMethod(RJavaTools.java:386)
CSVtoBPMN.readFromCSV received string parameters as
edges --> the issue is here, from and to id must be number, not string, because in nodes, they are defined as number. Both must match.
We need to add robustness to the Java implementation, when the referenced id in from/to doesn't exist in nodes. See #95
package installation: the usage of a gh token is probably not needed
The
0.1.4
package documentation explains that a github auth token must be pass to install the package. I don't understand why we need the token as all resources used by the package are public. If this is required, please document why.package installation failure
devtools::install_github("process-analytics/bpmn-layout-generators", ref="bpmnLayoutGeneratoR-0.1.4", subdir="R/bpmnLayoutGeneratoR")
The package requires JDK to be installed on the machine. Otherwise the
rjava
package installation failedExample on Ubuntu 20.04
rJava installation: http://rforge.net/rJava/ Java installation
Once JDK is installed, don't forget to run
R CMD javareconf
. See https://cran.r-project.org/doc/manuals/R-admin.html#Java-supportUsage example in the readme --> NullPointerException
Note: confirmed on Ubuntu 20.04 by @tbouffard and on Windows 10 Pro by @oanesini
With package 0.1.4, R 3.6.3, AdoptOpenJDK 11.0.11
Same issue when running
diagram <- bpmnLayoutGeneratoR::generateBpmnLayout(flow_node, sequence_flow, "BPMN")
diagram <- bpmnLayoutGeneratoR::generateBpmnLayout(flow_node, sequence_flow)
as we are using the Java API and not the CLI, we should add a try catch bock when calling the 'layout generation from csv' method and at least print the stack trace
http://rforge.net/doc/packages/rJava/Exceptions.html
When using the stack trace, we got
CSVtoBPMN.readFromCSV
received string parameters asnodes
edges --> the issue is here, from and to id must be number, not string, because in nodes, they are defined as number. Both must match. We need to add robustness to the Java implementation, when the referenced id in from/to doesn't exist in nodes. See #95