Objdictgen (odg)

This is python tools for working with Object Dictionary (OD) files for the CanFestival communication library. CanFestival is an open source implementation of the CANopen communication protocol.

This repo is located:

https://github.com/Laerdal/python-objdictgen

objdictgen includes tools to generate c code that works in tandem with a canfestival library. This tool has been built to work together with the Laerdal Medical fork for the canfestival library:

https://github.com/Laerdal/canfestival-laerdal

objdictgen is a tool to parse, view and manipulate files containing object dictionary (OD). An object dictionary is entries with data and configuration in CANopen devices. The odg executable is installed. It supports reading and writing OD files in .json/.jsonc format, in legacy XML .od and .eds files. It can generate c code for use with the canfestival library.

Install

To use this package Python3 must be installed. Use the package manager of choice to install the package in a virtual manager. We recommend the uv package manager.

Using uv (one of many methods)

$ uv venv
$ uv pip install objdictgen[ui]  # [ui] will install GUI tools
$ uv run odg

Using pip (for Windows)

$ py -3 -mvenv venv
$ venv/Scripts/pip install objdictgen[ui]

After this venv/Scripts/odg.exe (on Windows) or venv\bin\odg executable exists and can be called directly to run the command.

The objdictgen[ui] suffix will install the wx dependency needed for the UI odg edit. If no UI is needed, this suffix can be omitted.

odg command-line tool

odg is a command-line tool which is installed when the python package objdictgen is installed.

Invocation:

$ odg <options>
$ python -mobjdictgen <options>   # <-- If odg is unavailable

odg --help and odg <command> --help exists and shows the command options. The most useful commands are:

$ odg list <od-files...>              # List contents of the OD
$ odg convert <od-file1> <od-file2>   # Convert OD file
$ odg convert <od-file> <c-file>      # Convert OD to c code
$ odg diff <od-file1> <od-file2>      # Show differences between OD

Legacy commands

The legacy commands objdictgen and objdictedit are no longer available. The commands are available under odg gen and odg edit respectively.

JSON schema

src/objdictgen/schema/od.schema.json provides a JSON schema for the JSON OD format. This can be used for validation in editors.

To use the schema in VS Code, the following configuration must be added to your settings.json. After this is is installed, IntelliSense will show field descriptions, help with values and validate the file.

    "json.schemas": [
      {
        "fileMatch": [
          "**.jsonc"
        ],
        "url": "./src/objdictgen/schema/od.schema.json"
      }
    ],

Conversion

The recommended way to convert existing/legacy .od files to the new JSON format is:

$ odg generate <file.od> <file.jsonc> --fix --drop-unused [--nosort]

The --fix option might be necessary if the OD-file contains internal inconsistencies. It is safe to run this option as it will not delete any active parameters. The --drop-unused will remove any unused profile and DS-302 parameter that might be used in the file.

Motivation

The biggest improvement with the new tool over the original implementation is the introduction of a new .jsonc based format to store the object dictionary. The JSON format is well-known and easy to read. The tool use jsonc, allowing comments in the json file. odg will process the file in a repeatable manner, making it possible support diffing of the .jsonc file output. odg remains 100% compatible with the legacy .od format on both input and output.

The original objdictedit and objdictgen tool were written in legacy python 2 and and this is a port to python 3.

This tool is a fork from upstream canfestival-3-asc repo:

https://github.com/Laerdal/canfestival-3-asc

Making objdictedit excutable

To be able build an executable that can be run from anywhere:

$ pip install pyinstaller
$ pyinstaller packaging/objdictedit.spec

The file dist/objdictedit.exe can now be used anywhere. It does not need any pre-installed software.

License

Objdictgen has been based on the python tool included in CanFestival. This original work from CanFestival is:

Copyright (C): Edouard TISSERANT, Francis DUPIN and Laurent BESSARD

The Python 3 port and tool improvements have been implemented under

Copyright (C) 2022-2025 Svein Seldal, Laerdal Medical AS