This api allows you to upload an csv with links to images. The images will be cached on the server for further downloading. All values in the csvs are escaped and thus safe for insetion into html. No formatting is done on the data, as it is expected that the front-end will handle any formatting.

There is no restriction on the names and number of columns.

Images can be loaded in different sizes using the size query parameter.

This is a Python 3.5 app, and has not been tested with any other versions of Python.

Below is the basic installation instructions for developping on the app.

The app uses a Postgresql database and it is assumed this is installed and running. It is also assumed that the pg_hba.conf file is set as to allow local database users to connect over IPv4 using a md5 password.

Clone repository:

$ git clone https://github.com/inytar/elements_assignment.git
$ cd elements_assignment

Create a virtual environment and activate:

$ pyvenv .venv
$ source .venv/bin/activate

Install dependencies:

$ pip install -r requirements.txt
$ pip install -r dev_requirements.txt

Create new database and database user:

$ sudo -u postgres psql -c "CREATE USER elements WITH PASSWORD 'elements' CREATEDB"
$ sudo -u postgres psql -c "CREATE DATBASE elements OWNER elements"

Run the migrations:

$ python manage.py migrate

Create a superuser;

$ python manage.py createsuperuser

To run the application in development:

$ DEBUG=True python manage.py runserver

The api can be found under http://127.0.0.1:8000/api. API documentation can be found under http://127.0.0.1:8000/docs.

Uploading a CSV file using the Django html form gives problems. Using curl does not give this problem.

To test run:

$ python manage.py test

For deployment only the requirements in requirements.txt need to be installed. The .env-example should be renamed .env and edited to set the correct deployment setting. It is assumed the app will be run behind a server as described in: https://docs.djangoproject.com/en/1.10/howto/deployment/.

It is not possible to search through the csv files. This could be solved by instead of saving the csv files to disk saving a representation of the files in a Postgres JSONB column.

As we allow uploading files we are vulnerable to: https://docs.djangoproject.com/en/1.10/topics/security/#user-uploaded-content. To migate this problem only authenticated users are allowed to upload a new CSV file.

It is not possible to add, delete or edit a row in a csv file, if a csv file has beeen edited it must be uploaded again.