This repository includes:
- Specifications and documentation
- Programs and tools
...for the the Interface Exchange framework (IFEX).
IFEX is a general interface description and transformation technology which started in the Vehicle Service Catalog (VSC) project. The technology (IFEX) is now developed under a separate name to better describe its purpose and that it is widely and generally applicable. (This repository was previously named vsc-tools)
Please refer to the documentation for more information.
Some of the programs implement the tools for reading/writing/translating interface descriptions and code-generation. Other programs generate the IFEX interface model/language-specification and other documentation.
The implementations are primarily written in python and using some preferred technologies, such as the Jinja2 templating language ((alt. link)).
- Python >=3.10 installed (exact version might vary - the best definition of what works is likely the automated workflow files or tox.ini)
- Dependencies installed according to instructions below
As an alternative to installation instructions below, all the installations can also be hidden in a container. Refer to the README in the docker/ directory for running the tools using containers instead.
NOTE: Pyenv is not quite the same as other virtual environment handlers. Its most important function is to download, compile, and install a particular python version from source code. If your system python version is one that is not supported by this project and you are not able to install the right python version using another method, then this can be used if you are not able to install the right python version using another method. It can also be used in combination with virtual-environment handlers, to get access to different python versions.
If pyenv
shell command is not installed, use its installer to get it:
curl https://pyenv.run | bash # download and install (YOU are responsible to check the script content)
exec $SHELL # restart your shell using the new $PATH
Once you have an appropriate version of python, a virtual environment is recommended to avoid any particulars in the main system installation.
Go to project directory and then:
python -m venv venv
source venv/bin/activate
NEXT: Go to Installing packages
Go directly to Installing packages
pipenv is a tool that manages a virtual environment and install the package and its dependencies, making the process much simpler and predictable, since the Pipfile
states the dependencies, while Pipfile.lock
freezes the exact version in use.
Install this project and its dependencies in the local .venv
folder in this project, then use it (pipenv shell
):
export PIPENV_VENV_IN_PROJECT=1 # will create a local `.venv` in the project, otherwise uses global location
pipenv install --dev # install the development dependencies as well
NEXT: Go to Installing packages
You can then run:
pipenv shell # starts a shell configured to use the virtual environment
Activate a version in the current environment
pyenv local 3.10.6
IMPORTANT: Follow the pyenv instructions to make sure that pyenv environment
setup is added to .bashrc
so that the binaries can be found every time a
shell is started. Something like:
eval ($pyenv init)
should be run before anything else.
NEXT: Go to Installing packages
Tox is another way to set up the working environment. It is primarily used to test the program using multiple python versions.
- Install tox
- Install pyenv (most likely needed, use it if additional python versions are required)
- (optional) edit the provided tox.ini file
- Run tox -- this will execute pytest for all stated python versions
Here we provide a script that will check which versions are requested by tox.ini
, and install all of those versions first, using pyenv:
pip install "tox>=4"
scripts/pyenv_install_for_tox.sh
tox
Note: In this case, tox takes care of calling pip
and setup.py
to install the required packages.
(for any or none, virtual-environment -- but not needed if using tox)
Regardless of which type of virtual environment (if any) you use, it is required to install the IFEX package into your python environment, and to install needed dependencies with pip.
- Install dependencies:
pip install -r requirements.txt
- Install the IFEX provided modules into your virtual environment)
python setup.py develop
Installing the IFEX tools using setup.py
creates some convenient
executable shims, e.g. ifexgen
, ifexgen_dbus
, ifexconv_protobuf
, ...
If those commands are not in your environment, try setting up python virtual environment and make sure setup.py runs correctly. After that, they should be in the $PATH
variable and possible to run.
To run a generic code generator and specify an output template:
usage: ifexgen [-h] -d templates-dir-name ifex-input-file
To get some test IFEX files, clone the VSC repo:
git clone https://github.com/COVESA/vehicle_service_catalog
# Using a template
ifexgen vehicle_service_catalog/comfort-service.yml -d dtdl
# D-Bus:
ifexgen_dbus vehicle_service_catalog/comfort-service.yml
To test some -to-IFEX conversion, for example gRPC/protobuf:
git clone https://github.com/COVESA/uservices
ifexconv_protobuf uservices/src/main/proto/vehicle/propulsion/engine/v1/engine_service.proto >engine_service.ifex
To try the D-Bus XML generator:
usage: ifexgen_dbus input_ifex.yaml
The project uses pytest to define unit tests. A starting point is in the tests directory and more can be added.
To run tests, just run pytest in the root directory, (optionally specify the tests directory).
pytest -v tests
Propose changes using the GitHub Issues or Pull Request.
Make sure you have read the Specification Then please refer to the developer documentation
Please refer to GitHub tickets (Feel free to make a proposal or ask a question)
Please refer to GitHub tickets with the label "bug"
Various tips to consider:
- If the installation (pip install) is executed behind a (corporate) proxy, the following environments variables must be set:
http_proxy
andhttps_proxy
(including authentication e.g.,http://${proxy_username):$(proxy_password)@yourproxy.yourdomain
) - If you use a custom pip installation directory, set the
PYTHONPATH
environment variable to the directory that you set in thepip.ini
file. - If you do not run with administration rights, you may need to configure pip target path to write to your user home directory or consider using one of the
virtual environment
methods.