colSpList API

• 1 Content
• 2 Deploying the API with Heroku
  • 2.1 Creation of the local database
  • 2.2 Deployment in heroku
  • 2.3 Local deployment for development and test

---

The colSpList API (https://colsplist.herokuapp.com) provides utilities to determine whether species are alien/invasive, endemic and/or threatened on the Colombian territory.

The objective of this repository is to present both the API and its use (mostly example code in R and python).

1 Content

In the folder SQL, you will find the SQL code for initializing the database of the API

In the folder API, all the functions used by the API, and the definitions of classes, endpoints and routes for the API to work.

In the folder usage, you’ll find various folder containing examples on how to use the API The folder UsingPost contains examples concerning the Post endpoints and the API. More specifically, you will find the scripts that I used to feed the database with dataset concerning exotic, threatened, and endemic

2 Deploying the API with Heroku

In order to deploy the API locally and/or remotely, you will need:

1. to install heroku. (See https://devcenter.heroku.com/articles/getting-started-with-python#set-up).
2. git
3. an heroku account
4. python3
5. postgreSQL
6. psql

The file runtime.txt contains the python version used for development and deployment of the API

Then, clone the repository:

git clone https://github.com/marbotte/colSpList.git
cd colSpList

2.1 Creation of the local database

Whether you want to deploy the API locally or on heroku, the best solution for the database is to create one locally, and then to push it on heroku or use it locally (in particular, there is no other simple way to implement a base encoding of the postgres database in Unicode).

Therefore we need to create a local database called here “sp_list”.

In bash it may be done with:

createdb sp_list

Create a pgpass, in order not to worry about your credentials to connect to the local database

Then, the initialization of the database is done with:

psql sp_list -f SQL/init.sql

2.2 Deployment in heroku

You first need to login to heroku from the CLI:

heroku login

Then create the application in heroku:

heroku create

Push the application, using git, to the heroku system (Note that we push only the API folder, the rest of the repository containing files which are unnecessary for the API and might cause errors)

git subtree push --prefix API heroku master

Create the heroku database:

heroku addons:create heroku-postgresql:hobby-dev

Push the local database to heroku

heroku pg:push sp_list DATABASE_URL

Finally, deploy the web-based heroku system:

heroku ps:scale web=1
heroku open

More explanation here

2.3 Local deployment for development and test

First we will manage the python environment necessary to run the API. You do not have to use a particular python environment, but it may be helpful if you have other python projects which may present different requirements:

python3 -m colSpList_env venv
source colSpList_env/bin activate

Install all the dependencies required by the API

pip3 install -r API/requirements.txt

Export the database URL

export DATABASE_URL=postgres://localhost/sp_list

Run the API locally:

cd API
heroku local