Skip to content

Latest commit

 

History

History
51 lines (35 loc) · 2.27 KB

README.md

File metadata and controls

51 lines (35 loc) · 2.27 KB

MPServer

Server Implementation for the ModelPolisher 2.1.

The corresponding Python client library stub can be found here. Information on how to generate it is here.

See the Releases Page for a download of the standalone JAR. You can run it with

java -jar model-polisher-server-2.1.0-standalone.jar --config-file path/to/server-config.edn

The server-config.edn determines server behaviour and where to find systems the server depends on.

A config.json can be used to determine default behaviour for all options that the client does not provide values for.

Note:

  • Currently, it is intended for this project to run on Java 17. However, this is untested. It is known to run on Java 21.

Example Usage

curl -v -F "parameters=<config.json;type=application/json" \
        -F "model-file=@e_coli_core.xml" \
        localhost:3000/submit/file -o result.json

Note: This assumes that in your current directory, you have config.json and e_coli_core.xml. Further note the difference between < and @ in curl:

  • < inserts the content of a file
  • @ includes the actual file

This is intended. The python wrapper will of course cover this up.

The result.json (or whatever you name it) should contain

  • a runId UUID which should be reflected in the server logs
  • a diff which reflects the changes to the model
  • modelFile which is the base64 encoded output file.

Development

Docker

The container schmirgel/mpserver:development is configured to build ModelPolisher and MPServer from VCS on startup.

Python Client

The Python client library can be generated using the codegen script. This requires to be the swagger-codegen-cli jar to be in the directory. It can be downloaded from Maven Central.

Swagger UI

While the server is running, you can find the SwaggerUI at http://localhost:3000/api/docs/index.html. Note that this is not completely up to date.