chemotion-converter-app

Development setup

First, ensure that basic Python development prerequisites are installed on the system. For Debian or Ubuntu this can be done using:

apt-get install build-essential python3-dev python3-pip python3-venv

After cloning the repository, we recommend to create a virtual environment for the application:

python3 -m venv env
source env/bin/activate  # this needs to be done in each new terminal session

Then, the application and its dependencies can be installed:

pip install -e .                     # installs the package in editable mode
pip install -r requirements/dev.txt  # only needed for the development setup

The application is configured using environment variables, which can be read from a .env file. The file .env.dev can be used as template. At least FLASK_APP=converter_app.app needs to be set.

The Flask development server can be started now:

flask run

Troubleshooting (when unsing Windows, updated 25.07.2024)

1. Please make sure that you are using a fresh Python3.10 virtual environment (without any packages installed, 3.12 and 3.9 are not supported)
2. Before installing the requirements pip install wheel setuptools pip pybind11 # always using pip of your virtual environment
3. After installing all requirements via pip install -e . and pip install -r requirements/dev.txt check if you have python-magic-bin and / or python-magic installed. If not pip install python-magic-bin
4. Now try starting it with the following env variables PYTHONUNBUFFERED=1;FLASK_APP=converter_app/app.py;FLASK_ENV=development;FLASK_DEBUG=1
5. If you are still getting ERRORS containing "libmagic" try uninstalling python-magic but keeping python-magic-bin

Production setup

For the production deployment, we recommend to create a dedicated user:

groupadd chemotion -g 2002
useradd chemotion -u 2002 -g 2002 -c Chemotion -m -d /srv/chemotion -s /bin/bash

Again we create a virtual environment, but now we install the application directly from GitHub:

python3 -m venv env
source env/bin/activate
pip install git+https://github.com/ComPlat/converter-app

In production, HTTP basic auth is used to seperate clients. First, create a htpasswd file using the sha1 hashing algorithm:

htpasswd -s -c htpasswd client  # create a new file with the client_id "client"
htpasswd -s htpasswd client     # update an existing file for the client_id "client"

The path to this file needs to be provided in .env as HTPASSWD_PATH.

The development server is not suited for a production deployment. Instead we recomend to use a reverse-proxy setup using gunicorn and NGINX.

Create the .env file in /srv/chemotion/.env, but use .env.prod as template, since it contains the variables we need to set for gunicorn.

In order to create the needed log and run directories create a tmpfiles.d config in /etc/tmpfiles.d/chemotion-converter.conf. A sample file can be found in etc/tmpfiles.d/chemotion-converter.conf. Create the directories using:

systemd-tmpfiles --create

Next, create a systemd service file in /etc/systemd/system/chemotion-converter.service. Again, a sample file can be found in etc/systemd/system/chemotion-converter.service. Reload systemd and start (and enable) the service:

systemctl daemon-reload
systemctl start chemotion-converter
systemctl enable chemotion-converter

If the service won't start, journalctl -xf might indicate what is wrong.

The guincorn server listens on the port given in the env file (default: 9000) on localhost. Using the following code in your NGINX configuration, it can be reversed proxied to the /api/v1 route:

    location /api/v1 {
        proxy_pass         http://127.0.0.1:9000/;
        proxy_redirect     off;

        proxy_set_header   Host                 $host;
        proxy_set_header   X-Real-IP            $remote_addr;
        proxy_set_header   X-Forwarded-For      $proxy_add_x_forwarded_for;
        proxy_set_header   X-Forwarded-Proto    $scheme;
    }

    client_max_body_size 64M;  # set this to the maximum file size allowed for upload

Acknowledgments

This project has been funded by the DFG.

Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under the National Research Data Infrastructure – NFDI4Chem – Projektnummer 441958208 since 2020.