<!--
SPDX-FileCopyrightText: 2024 Helmholtz-Zentrum hereon GmbH
SPDX-FileCopyrightText: 2024 Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research

SPDX-License-Identifier: CC-BY-4.0
-->

(installation)=
# Installation

```{warning}

This page has been automatically generated as has not yet been reviewed
by the authors of capice-compute-backend!
```

To install the _capice-compute-backend_ package, we
recommend that you install it from PyPi via

```bash
pip install capice-compute-backend
```

Or install it directly from [the source code repository on
Gitlab][source code repository]
via:

```bash
pip install git+https://git.rwth-aachen.de/nfdi4earth/pilotsincubatorlab/pilots/capice/capice-compute-backend.git
```

The latter should however only be done if you want to access the
development versions.

[source code repository]: https://git.rwth-aachen.de/nfdi4earth/pilotsincubatorlab/pilots/capice/capice-compute-backend

(install-develop)=
## Installation for development

Please head over to our
`contributing guide <contributing>`{.interpreted-text role="ref"} for
installation instruction for development in case you want to contribute back
to this repository.

To setup and test the DASF functionality of this package locally, you need to
run a DASF message broker. Everything for this is contained in the `django`
folder of the source code.

### Manual DASF setup

#### Setup the message broker
To manually setup a message broker and connect a backend module, run:

```bash
# create a virtual environment for development (note, on windows, you need
# to run these venv-install commands from the Makefile manually, or use the WSL)
make venv-install
source venv/bin/activate

# setup and run the django message broker
cd django
python manage.py migrate
python manage.py dasf_topic -n capice-compute --anonymous
python manage.py runserver
```

This will create a local file at `django/db.sqlite3` with the database and
spawn a server that listens to [http://localhost:8000](http://localhost:8000).

#### Connect a backend module
To connect a backend module to the message broker, run

```
python -m capice_compute.backend --websocket-url ws://localhost:8000/ws --topic capice-compute listen

```

#### Make Request via the autogenerated API
To make a request via the auto-generated `api` module, set the corresponding
environment variables for the websocket url and the topic, and call a function
from the `api` module. Like this for instance:

```bash
DE_BACKEND_WEBSOCKET_URL=ws://0.0.0.0:8000/ws \
  DE_BACKEND_TOPIC=capice-compute \
  python -c "from capice_compute.api import version_info; print(version_info())"
```