San Francisco MTC

Fetching data from San Francisco MTC with Airbyte and Python.


This Airbyte Connector uses the HTTP API provided by the San Francisco MTC in Python. With this connector, data is easily pulled from San Francisco MTC and loaded into the ViaWarn Postgres database, and are available in the ViaWarn app.

The connector is built on top of the Airbyte open-source platform, which provides a standardized way of integrating data sources and destinations. It is designed to be simple and easy to use. The required configuration is show in the the sections below. 

The pulled resources from the AFMTA API are the following:

Requirements and Installation

Find the requirements and steps to create the connector in the Airbyte documentation.

Customized files

After creating the Airbyte connector template, modify the next files to retrieve data from San Francisco MTC.


Location: source-viawarn-us-sf-mtc/spec.yaml
In this case, the connector password and the request limit are specified in the spec file.
connectionSpecification: $schema: 
    title: VIAWARN US SF MTC Spec 
    type: object 
                    type: string
                    airbyte_secret: true
                    description: Developer API Key obtained from
                    type: integer
                    description: Request limit as described here


Location: secrets/config.json
This file is ignore by git.


Location: integration_tests/configure_catalog.json
List of Streams that describes the data in the data store that a Source represents. It also declares the sync modes supported by each stream (full refresh or incremental).
        "stream": {
            "name": [STREAM_NAME],
            "json_schema": {},
            "supported_sync_modes": ["full_refresh"] 
        "sync_mode": "full_refresh", "destination_sync_mode": "overwrite" 

Location: source-viawarn-us-sf-mtc/

class ViawarnUsSfMtcStream()

Create a stream to represent the data that will be pull from the API. It contains all the common API functionalities:
  •  url_base

This field is the common part of the URL for all streams, if it does not exist, add it in each stream method.


  • __init__()

This is a constructor method used to initialize the class. It uses the ‘config’ object to define the ‘developer_api_key’ and ‘request_limit’ attributes.

return config[]

  • next_page_token()

If pagination is required, the response json would have two more elements: ‘pagination’ and ‘next_url’. Use them to set the ‘offset’ parameter. 

return dictionary if there pagination is required, or None otherwise.

  • request_params()

Create a dictionary. Keys:

‘developer_api_key’ and ‘limit’= request_limit. If pagination is available add ‘next_page_token’ key.

return a dictionary

  • parse_response()

It is a placeholder to store the request.response object and constructs the appropriate mapping object for each parsed element by each Stream class.

yields an empty dictionary.

  • backoff_time()

Time for a next tried in seconds. This is needed because the API states a request-rate limit of 60 requests per 3600s.

return number

class EVENT()

This method extends the ViawarnUsSfMtcStream() method to specified unique behavior for this stream.
  • primary_key

Unique field required in the stream associate to each output record.


  • path()

subdomain and paramters to pass into the url: traffic/events

return a string

  • parse_response()

Definition of how a response is parsed.

Set a definition to: primary key and ewkt fields. The response json is decoded using ‘ response.encoding=”‘utf-8-sig'” ‘

return an iterable containing each record in the response.

class SourceViawarnSfMtc()

  • check_connection()

Set up a test to verify the connection between the connector and the API. This test should return two variables: the first one is a boolean that indicates the status of the connection, while the second one shows a log message if the connection fails.

return True, None

  • streams()
Set up the stream in the source that will be parsed.
return List of streams


Location: source-viawarn-us-sf-mtc/schemas/event.json
Declare schema for event stream by describing the output in a json file.
Stream name: event