Some services require the negotiation of an access token for each session when accessing an API.
This process typically consists of
a first call to the API with a user ID and passphrase/secret/password,
retreiving a token,
using the token in subsequent queries to the API.
BarentsWatch¶
The following is an example based on BartensWatch’s API.
client_id is the combination of an email address and a client name (see BarentsWatch API tutorial).
client_secret is a password connected to the client name.
# This file contains both the client_id and client_secret for the BarentsWatch API.
BW = open('../../../No_sync/BarentsWatch','r').read()
client_id = BW.split('\n')[0]
client_secret = BW.split('\n')[1]
config = {
'client_id': client_id,
'client_secret': client_secret,
'token_url': 'https://id.barentswatch.no/connect/token',
'api_base_url': 'https://www.barentswatch.no/bwapi'
}Convenience functions¶
We create two functions for easily requesting a token and data from the API (see API_Tokens.py).
These are based on the GitHub repository of BarentsWatch
import API_Tokens as AT
token = AT.get_token(config)Token request successful
data_2023_1 = AT.get_request('/v1/geodata/fishhealth/locality/2023/1', token, config)
data_2023_2 = AT.get_request('/v1/geodata/fishhealth/locality/2023/2', token, config)# Structure of the data
data_2023_1['localities'][0]['hasSalmonoids']False# Extract hasSalmonoids for all localities (at one timepoint)
hasSalmonoids = []
for i in range(len(data_2023_1['localities'])):
hasSalmonoids.append(data_2023_1['localities'][i]['hasSalmonoids'])
# Make a pie chart of the data
import matplotlib.pyplot as plt
plt.pie([hasSalmonoids.count(False), hasSalmonoids.count(True)], labels = ['No salmonoids', 'Salmonoids'])
plt.show()