Airtable.py: Python API wrapper for Airtable

Python interface to the Airtable's REST API - https://airtable.com -

For javascript enthusiasts: https://github.com/Airtable/airtable.js

Installation

Airtable Python uses Requests.py: make sure you have it installed by running

$ pip install requests

Getting started

Once you have created a new base and a new table through the Web interface, you're ready to start using Airtable Python.

import airtable
at = airtable.Airtable('BASE_ID', 'ACCESS_TOKEN')
at.get('TABLE_NAME')

Here's an example of response from the Restaurant's example base

{u'records': [
  {u'fields': {u'Diet': u'Kosher or Halal',
    u'Friendly Restaurants': [u'recr0ITqq9C1I92FL', u'recGeAJLw0ZkbwdXZ'],
    u'Icon': [{u'filename': u'no-pig.jpg',
      u'id': u'attzKGOBbjndOx0FU',
      u'size': 34006,
      u'thumbnails': {u'large': {u'height': 202,
        u'url': u'https://dl.airtable.com/trmtq3BaRoa0sWnyffWZ_large_no-pig.jpg',
        u'width': 256},
       u'small': {u'height': 36,
        u'url': u'https://dl.airtable.com/yzuRv5CyRs2PVH4fDvCe_small_no-pig.jpg',
        u'width': 46}},
      u'type': u'image/jpeg',
      u'url': u'https://dl.airtable.com/DyGOjAASze6AIkQxFpDv_no-pig.jpg'}],
    u'People': [u'Annie', u'Maryam']},
   u'id': u'rec5sD6mBBd0SaXof'},
   ...

API Reference

The available methods closely mimic the REST API:

Get

Given a table name, fetched one or multiple records.

at.get(table_name, record_id=None, limit=0, offset=None,
       filter_by_formula=None, view=None, max_records=0, fields=[])

where

table_name (required) is a string representing the table name
record_id (optional) is a string, which fetches a specific item by id. If not specified, all items are fetched
limit (optional) is an integer, and it can only be specified if record_id is not present, and limits the number of items fetched (see pageSize in the AirTable documentation)
offset is a string representing the record id from which we start the offset
filter_by_formula (optional) is a string to filter the retrieving records (see filterByFormula in the AirTable documentation)
max_records (optional) is the total number of records that will be returned (see maxRecords in the AirTable documentation)
fields (optional) is a list of strings with the field names to be returned

Iterate

Given a table name, fetched all records.

at.iterate(table_name, batch_size=0, filter_by_formula=None,
       view=None, max_records=0, fields=[])

where

table_name (required) is a string representing the table name
batch_size (optional) is an integer specifying the number of records to fetch per request. The default (0) uses the API default, which is (as of 2016-09) 100.
limit (optional) is an integer, and it can only be specified if record_id is not present, and limits the number of items fetched (see pageSize in the AirTable documentation)
filter_by_formula (optional) is a string to filter the retrieving records (see filterByFormula in the AirTable documentation)
view  (optional) is a string representing the name or ID of a view in the table. If set, only the records in that view will be returned. The records will be sorted according to the order of the view.
max_records (optional) is the total number of records that will be returned (see maxRecords in the AirTable documentation)
fields (optional) is a list of strings with the field names to be returned

Note: this returns a generator instead, which you can use to loop each record:

# example with similar results of at.get
result = { "records": [] }
for r in self.at.iterate(self.table, fields=fields):
    result["records"].append(r)

Create

Creates a new entry in a table, and returns the newly created entry with its new ID.

at.create(table_name, data)

where

table_name (required) is a string representing the table name
data (required) is a dictionary containing the fields and the respective values

Update

Updates some fields in a specific entry in the table. Fields which are not explicitly included will not get updated

at.update(table_name, record_id, data)

where

table_name (required) is a string representing the table name
record_id (required) is a string representing the item to update
data (required) is a dictionary containing the fields (and the respective values) to be updated

Update All

Like the previous method, but updates all fields, clearing the ones that are not included in the request.

at.update_all(table_name, record_id, data)

Delete

Delete a specific record from the table

at.delete(table_name, record_id)

where

table_name (required) is a string representing the table name
record_id (required) is a string representing the item to update

Release

To release, tag the Git repo with a new version number, push that tag to GitHub then Travis CI will do the rest.