sitn_applications

Apps and custom webservices from SITN. This app can run in two different contexts: internet or intranet. Depending on the context, some apps will not be installed when deploying.

Requirements

• PostgreSQL >= 11 + PostGIS
• Python >= 3.10
• GDAL, GEOS (if running without docker on Windows, install OSGEO4W and add its bin directory to the Path)
• GNU gettext https://www.gnu.org/software/gettext/

Getting started

Fork and clone this repository. Make a copy of the .env.sample file and adapt it to your environment settings:

cp .env.sample .env

and configure the different variables.

The paths related to GDAL et GEOS can either point to existing installations on the system, e.g.:

# Path to the folder containing all GDAL/GEOS DLLs
GDAL_PATH="C:/Program Files/GDAL/"

# Full path to the GDAL DLL
GDAL_LIBRARY_PATH="C:/Program Files/GDAL/gdal.dll"

# Full path to GEOS C API DLL
GEOS_LIBRARY_PATH="C:/Program Files/PostgreSQL/16/bin/libgeos_c.dll"

Alternatively, you can install GDAL (and its GEOS and PROJ dependencies), using the geospatial python wheels provided here:

• geospatial-wheels

Download the appropriate version, then install it in the venv using pip, e.g.:

pip install .\gdal-3.11.4-cp314-cp314-win_amd64.whl

Then configure the GDAL paths in the .env file to point to the ressources in the venv, e.g.:

# Path to the folder containing all GDAL/GEOS DLLs
GDAL_PATH="C:/Projects/sitn_applications/venv/Lib/site-packages/osgeo/"

# Full path to the GDAL DLL
GDAL_LIBRARY_PATH="C:/Projects/sitn_applications/venv/Lib/site-packages/osgeo/gdal.dll"

# Full path to GEOS C API DLL
GEOS_LIBRARY_PATH="C:/Projects/sitn_applications/venv/Lib/site-packages/osgeo/geos_c.dll"

You should also have PostgreSQL with PostGIS installed, and the path to PostgreSQL's bin directory should be defined in the Path variable in your Windows system environment variables (e.g. C:\Program Files\PostgreSQL\18\bin).

Finally, you need to have gettext installed:

• gettext for Windows

Running in development mode, without docker

First, create a copy of the .env.sample file called .env:

cp .env.sample .env

Install and activate the virtual environment:

python -m venv venv
./venv/Scripts/activate
python -m pip install --upgrade pip
pip install -r requirements-lock.txt

Create a local database:

python manage.py createdb

Dump some fresh data from prod database

python manage.py getdata

Run migrations

python manage.py migrate

You're now ready to go:

python manage.py collectstatic
python manage.py compilemessages --locale=fr
python manage.py runserver

Switching between the Internet & Intranet instances

If developping on the Internet instance, then you should define a new environment variable in the .env.dev. file:

IS_INTRANET=False

Running locally on Docker

First, create a copy of the env.sample file called env.<context>.local:

cp .env.sample .env.intranet.local

Then you'll be able to deploy locally on docker:

python deploy intranet local

Deploying on production

Deployments are based on shared .env files. If your project requires migrations, the default user will not be able to create tables, so you must temporarily switch to a more privileged user.

Then you'll be able to deploy your instance with python deploy <context> <instance>

Example:

python deploy intranet prepub

Tests

Running tests will require a database for testing.

python manage.py testdb

then you can run tests

python manage.py test --keepdb

Translations

Prepare translation files

django-admin makemessages -l fr --no-location --ignore venv

then compile them

python manage.py compilemessages --locale=fr

Starting a new app

python manage.py startapp myapp

Then:

1. Put something in urls.py in your app folder
2. Import your urls in sitn/urls.py
3. Install the app in settings.py

Upgrading packages

In your venv:

1. Update the versions manually in requirements.in if necessary
2. pip-compile requirements.in --upgrade --output-file=requirements-lock.txt
3. pip install -r requirements-lock.txt