Skip to content

Usage

Jump to bottom
Baud Rémy edited this page May 12, 2021 · 1 revision

Usage

HTTP API

The HTTP API features 2 API paths:

  • /api/v1/jobs to interact with scheduled graph packaging jobs.
  • /api/v1/users to interact with the user store, i.e. creating/listing/removing users.

Both paths have GET, POST and DELETE endpoints to list, create and remove entries, respectively.

Authentication

Some endpoints expect valid authentication in the form of Basic Authentication of registered users. These are:

  • POST /api/v1/jobs
  • DELETE /api/v1/jobs
  • POST /api/v1/users: needs even admin privileges, i.e. only the ADMIN_EMAIL account can create new users
  • DELETE /api/v1/users

Note, the configured admin user is created on API startup and can start administrating the API right away.

The authentication credentials are best placed in the header of the request and is expected to be a base64 encoded string of user_email:user_pass, e.g.

from base64 import b64encode
import requests
auth = f'Basic {b64encode("[email protected]:admin").decode()}'
print(requests.post('http://localhost:5000/api/v1/users', json={"email": "[email protected]", "password": "bla"}, headers={'Authorization': auth}).json())

Swagger

Documentation

The API documentation (via Swagger) is accessible under /api/v1 (e.g. http://<IP/domain/localhost>:\<port\>/api/v1), which is also the base URL for the whole API.

The documentation is grouped by API paths showing their endpoints and allowed methods with details on how to access them. Each endpoint has additional information on expected parameters, example response values and possible return status codes and their meaning.

E.g. users has 4 endpoints:

  • POST /users: corresponds to /api/v1/users, expects the email and password in the JSON payload and responds with either 200 (success), 401 for invalid authorization, 409 (already existing user) and a generic 500 (unknown error). E.g. curl -XPOST 'http://localhost:5000/api/v1/users' -d '{"email": "[email protected]", "password": "secpass"}' --header 'Authorization: Basic bmlsc0BnaXMtb3BzLmNvbTp'
  • GET /users: corresponds to /api/v1/users, expects no parameters and responds either with 200 (success) or 500 (unknown error). E.g.
curl -XGET 'http://localhost:5000/api/v1/users'
  • and so on

Interactive playground

Apart from documentation you can use Swagger to fire requests against the HTTP API.

When you click on an endpoint, you'll see a Try it out button on the right. When you click that, you'll be able to fill out the endpoint parameters (if any) and firing the request by hitting Execute. The response will be showing up in the section right below.

Note, that some endpoints expect authenticated requests. In Swagger you'll see a small lock on the endpoint if it expects authentication. When you try to request your browser will ask you for user name and password. To make it easier, you can also set the credentials globally by hitting the Authorize button on top of the documentation.

Schemas

Most of the schemas described in the Swagger documentation should be straight-forward. However, some endpoints need some more information.

POST /api/v1/jobs

The table below gives some insights. Note, a lot of fields are auto-generated so they only show up in the response.

Field Type Only response Description
name str The name you want to give the entry. Will determine part of the dataset path, where the full path is $DATA_DIR/<router>/<router>_<provider>_<name>.<compression>.
description str The description for the dataset.
provider str The data provider, one of ["osm", "tomtom", "here"].
router str The routing engine, one of ["valhalla", "osrm", "ors", "graphhopper"]
bbox str The bounding box to crop the graph package to in format minx,miny,maxx,maxy, e.g. 1.531906,42.559908,1.6325,42.577608
interval str The update interval for the graph package, one of ["once", "daily", "weekly", "monthly"]
compression str The compression method to use, one of ["zip", "tar"]
id int yes The database ID
user_id int yes The user's database ID
status int yes The current processing status of the job, one of ["Queued", "Extracting", "Tiling", "Failed", "Deleted", "Completed"]
job_id str yes The Redis ID of the job. Will be populated during processing.
container_id str yes The docker container ID of the routing engine. Useful for debugging purposes if something goes wrong. Will be populated during processing.
last_started str yes The UTC timestamp of the last time this job was started.
last_finished str yes The UTC timestamp of the last time this job was finished. Will be populated after processing.
path str yes The full path to the graph package.
pbf_path str yes The full path to the PBF file used to generate the graph.
Clone this wiki locally