factual Wrapper for the Factual HTTP API
Download

factual Ranking & Summary

factual Tags

Python library Python api

factual Description

factual is a Python module that wraps the Factual v3 API. Its query style is SQLAlchemy-inspired and designed to make it easy to build "read" requests by chaining filter calls together.There is also limited support for v2 API, including actions other than "read," under the "v2" subpackage.Note that there are minor API inconsistencies between the v2 and v3 versions, both on the server side and in this wrapper. The v3 version of the wrapper is now the default.API accessTo get started, you'll need to register for Factual API OAuth credentials.UsageFirst, create a Session using your v3 API OAuth consumer key and consumer secret:import factualsession = factual.Session( consumer_key="myOAuthConsumerKey", consumer_secret="myOAuthConsumerSecret" )Now, build a query using the read action on the "global" table:query = session.read("global")You can apply as many filters as you'd like to the query. Filters on a query are cumulative, and can be chained:query.search("coffee")query.filter({"locality": "Albany"}).filter({"region":"NY"})When you're done, run() the query and retrieve the results using the records() method:from pprint import pprintdata = query.run().records()pprint(data)Geographic queriesYou probably want to search for places near a point. Use the within(latitude, longitude, radius_in_meters) helper method of read. within() chains and applies like any other filter, except that the last call will overwrite earlier geo filters. The underlying Factual filter API has changed between v2 and v3, but this will work for both:query = session.read("global").within(40.7353,-73.9912,1000).search("coffee")PaginationGet more results using count() and page():query.count(30).page(2) # 30 results per request, second page of resultsNote that Factual's API instead uses limit and offset, so I should probably change the wrapper to match.CategoriesFactual provides categories as hierarchical strings. That is, any place marked "Food & Beverage > Bakeries" is in the "Bakeries" subcategory of "Food & Beverage."It's possible to query for either specific subcategories or parent categories using the $bw ("begins with") filter operator. You can then search across multiple of these $bw filters by chaining them together with $or.Because this can get pretty lengthy, the category_helpers module has a make_category_filter function. make_category_filter takes a list of category strings and combines them into a $bw/or filter. Since $bw will always include all subcategories of an supercategory listed, make_category_filter also dedupes the provided categories to generate the shortest list possible. This may mean that it will include more subcategories than you indended; but if you want to get in to $and $not $bw tangles, you're on your own.Pass blank = True as a kwarg to make_category_filter if you also want results without categories set.from factual import category_helpersmy_categories = my_filters = category_helpers.make_category_filter(my_categories, blank=True)# {'$or': ({'category': {'$bw': 'Food & Beverage'}}, {'category': {'$bw': 'Shopping'}}, {'category': {'$blank': True}})}query = s.read("global").filter(my_filters)Non-OAuth requestsIt's possible to skip OAuth for the v3 API, if, for instance, there's some trouble with signing requests. Creating a Session without a consumer_secret will authenticate requests via the KEY query string parameter rather than OAuth:non_oauth_session = factual.Session(consumer_key="myOAuthConsumerKey")Note that Factual discourages falling back to the KEY parameter, and intends it for debugging use only, so use OAuth if possible.Examplesfrom factual import *s = Session(consumer_key="deadbeef", consumer_secret="foobar")my_place = s.read("global").search("coffee").run().records()Building requests one piece at a time:query = s.read("global")query.filter({"name": "Foobar"})if my_address != None: query.filter({"address": my_address})response = query.run()records = response.records()Limiting categories and using the filter helper functions:from factual.filter_helpers import *q = s.read("global").filter( or_( bw_("category", "Food"), bw_("category", "Arts") ) ).search("Foobar")records = q.run().records()Geo queries:# lat, lon, radius in meterscoffee_places = s.read(places).search("coffee").within(40.7353,-73.9912,1000).run().records()Factual v2 "Server" APITo use the v2 API, instantiate a factual.v2.Session object instead of a factual.Session object:import factualv2_session = factual.v2.Session(api_key="deadbeef")...Note that you'll need to provide a v2 API key using the api_key argument. Factual issues v2 API keys via a different process than for v3 API credentials.In the v2 API, you can also modify a record in the Playpen:v2_session = factual.v2.Session(api_key="deadbeef")p = v2_session.read(USLocalPlaypen).search("coffee").count(1).run().records()p += "/Foobar"v2_session.input(USLocalPlaypen).values(p).comment("Silly update test").run()See also the Python documentation for session.Session and requests.Read and Factual's developer documentation.Product's homepage

factual Related Software