06f7a64edd
Even if this uses a secure connection, piping stuff returned by `curl` into an interpreter is bad practice and can lead to security problems. We now assume that `pip` and `setuptools` are installed on the target system, which is reasonable IMO. Signed-off-by: Felix Walter <felix.walter@d3tn.com>
2.2 KiB
Python 3 virtualenv Setup
Multiple functional tests as well as the integration test toolchain in the µD3TN project are using Python scripts to speed up development compared to pure C test implementations. The scripts are located in the tools/ directory. The integration test toolchain can be found in the test/integration/ directory. Both leverage a Python 3 implementation of several DTN protocols and convergence layers, provided in the pyd3tn/ directory.
A venv is used to isolate the
µD3TN Python environment from the system. The Makefile target virtualenv
creates a default Python virtualenv including all required Python packages.
make virtualenv
The location of the virtualenv directory can be controlled with the VENV
parameter. The default directory is .venv/.
make VENV=path/to/venv virtualenv
After the virtualenv is created the activate script has to be sourced to add
the virtualenv Python interpreter to the PATH variable. Alternatively,
environment switcher like direnv (see below) or virtualenvwrapper can be
used.
source .venv/bin/activate
The tools/ directory is added to the virtualenv site-packages allowing the
import of packages residing in tools/ like the pyd3tn package, e.g.
from pyd3tn.bundle7 import serialize_bundle7
serialize_bundle7("dtn://GS1/", "dtn://GS2/", b"Hello world!")
Use an own virtual environment handler
You may also use other tools or a manual approach for managing the virtual
environment, e.g., if you want it to use another path.
In this case, create and activate a new virtual environment using the method
of your choice and then install the dependencies:
make update-virtualenv
Optional: direnv support
direnv is an environment switcher loading environtal
variables depending on the current directory. It can be used to automatically
create and load Python virtualenvs. This is an example .envrc file making use
of the Makefile virtualenv target.
layout_virtualenv() {
local venv=$1
if [ ! -d "${venv}" ]; then
make virtualenv
fi
source ${venv}/bin/activate
}
layout virtualenv .venv