compynieshouse

compynieshouse - a convenient wrapper for the UK Companies House REST API

Overview

The UK's Companies House publishes the financial returns, lists of officers etc. of many of the companies registered with it. They now provide some of that information conveniently over a REST API, which is a fantastic resource.

While the API and its documentation are fairly intuitive, this module seeks to make it more "turnkey" for the user, avoiding headaches over basic auth, the requests library etc.

Requirements:

1. Python 3
2. datagrab
3. A Companies House API key

Getting it

To download datagrab, either fork this github repo or simply use Pypi via pip.

Usage

Retrieve the key information about a specific company

Let's say you're a BI developer and you want an easy way of refreshing your the company key information in your CRM system. You'd first need a way of easily checking the Companies House as a single source of truth.

Here's the docstring for the main class, CHCompany:

Companies House Company

A way of retrieving data about a target company using the Companies house
REST API

------------------------
args:

    appKey: string
    Your companies house API key

    company_query_string: string
    if by=="id" then this should be the companies house company ID of the
    target company, else if by=="friendly_string" then it's the long-form
    name of the target company.
------------------------
kwargs:

  by: string
  Accepts: "id" if you are using the companies house ID or "friendly_string"
  if you are using the name of the target company

  zero_results_suppression: bool
  Accepts: True or False
  Default: False
  When set to true, this prints some troubleshooting tips if you run a
  company search which returns zero results. Setting to True is not
  recommended for production code.

So you can see, you just need your Companies House API Key and the company number.

>>> from compynieshouse.company_query import CHCompany
>>> ch = CHCompany("<my_companies_house_API_key",
                company_query_string="02557590",   # What we ask for
                by="id", # Search type - can be by "friendly_string", or "id"
                )

>>> ch.jsonDict

              {'sic_codes': ['82990'],
                 'jurisdiction': 'england-wales',
                 'date_of_creation': '1990-11-12',
                 'type': 'ltd',
                 'undeliverable_registered_office_address': False,
                 'last_full_members_list_date': '2015-11-12',
                 'registered_office_address': {'postal_code': 'CB1 9NJ',
                  'address_line_2': 'Cambridge',
                  'address_line_1': '110 Fulbourn Road',
                  'locality': 'Cambridgeshire'},
                 'accounts': {'overdue': False,
                  'next_made_up_to': '2020-03-31',
                  'next_accounts': {'overdue': False,
                   'due_on': '2020-12-31',
                   'period_end_on': '2020-03-31',
                   'period_start_on': '2019-04-01'},
                  'accounting_reference_date': {'day': '31', 'month': '03'},
                  'last_accounts': {'period_start_on': '2018-04-01',
                   'period_end_on': '2019-03-31',
                   'made_up_to': '2019-03-31',
                   'type': 'group'},
                  'next_due': '2020-12-31'},
                 'company_number': '02557590',
                 'has_been_liquidated': False,
                 'company_name': 'ARM LIMITED',
                 'has_insolvency_history': False,
                 'etag': '7011282471135667318564d3ba8a2c3942359264',
                 'company_status': 'active',
                 'has_charges': True,
                 'previous_company_names': [{'effective_from': '1990-12-03',
                   'ceased_on': '1998-05-21',
                   'name': 'ADVANCED RISC MACHINES LIMITED'},
                  {'effective_from': '1990-11-12',
                   'ceased_on': '1990-12-03',
                   'name': 'WIDELOGIC LIMITED'}],
                 'confirmation_statement': {'next_made_up_to': '2020-11-14',
                  'overdue': False,
                  'last_made_up_to': '2019-11-14',
                  'next_due': '2020-11-28'},
                 'links': {'self': '/company/02557590',
                  'filing_history': '/company/02557590/filing-history',
                  'officers': '/company/02557590/officers',
                  'charges': '/company/02557590/charges',
                  'persons_with_significant_control': '/company/02557590/persons-with-significant-control',
                  'registers': '/company/02557590/registers'},
                 'registered_office_is_in_dispute': False,
                 'can_file': True}

CHCompany inherits JsonResponseInterpreter class, which is part of the datagrab library. So we can access attributes like jsonDict and the json_tree_traverse method.

This is pretty handy if, say, we want to find the city of the registered address.

>>>>ch.json_tree_traverse(["registered_office_address","locality"])         
'Cambridgeshire'

If we're new to the Companies House schema, we can use ch.visualize_json() to see the structure.

Root
├── accounts
│   ├── accounting_reference_date
│   │   ├── day
│   │   └── month
│   ├── last_accounts
│   │   ├── made_up_to
│   │   ├── period_end_on
│   │   ├── period_start_on
│   │   └── type
│   ├── next_accounts
│   │   ├── due_on
│   │   ├── overdue
│   │   ├── period_end_on
│   │   └── period_start_on
│   ├── next_due
│   ├── next_made_up_to
│   └── overdue
├── can_file
├── company_name
├── company_number
├── company_status
├── confirmation_statement
│   ├── last_made_up_to
│   ├── next_due
│   ├── next_made_up_to
│   └── overdue
├── date_of_creation
├── etag
├── has_been_liquidated
├── has_charges
├── has_insolvency_history
├── jurisdiction
├── last_full_members_list_date
├── links
│   ├── charges
│   ├── filing_history
│   ├── officers
│   ├── persons_with_significant_control
│   ├── registers
│   └── self
├── previous_company_names
├── registered_office_address
│   ├── address_line_1
│   ├── address_line_2
│   ├── locality
│   └── postal_code
├── registered_office_is_in_dispute
├── sic_codes
├── type
└── undeliverable_registered_office_address

Let's imagine now that we want to see if a company name is available. In that case, we'd need to search by a particular substring.

In that case, set the "by" keyword argument to "friendly_string":

>>> ch = CHCompany("<my_companies_house_API_key>",
                company_query_string="ARM",   # What we ask for
                by="friendly_string", # Search type
                  )

>>> ch.json_tree_traverse(["items"])[:2]
[{'snippet': '',
  'company_number': '02557590',
  'description': '02557590 - Incorporated on 12 November 1990',
  'company_type': 'ltd',
  'title': 'ARM LIMITED',
  'date_of_creation': '1990-11-12',
  'company_status': 'active',
  'address_snippet': '110 Fulbourn Road, Cambridge, Cambridgeshire, CB1 9NJ',
  'address': {'premises': '110',
   'postal_code': 'CB1 9NJ',
   'address_line_2': 'Cambridge',
   'address_line_1': 'Fulbourn Road',
   'locality': 'Cambridgeshire'},
  'matches': {'title': [1, 3], 'snippet': []},
  'kind': 'searchresults#company',
  'links': {'self': '/company/02557590'},
  'description_identifier': ['incorporated-on']},
 {'links': {'self': '/company/11551941'},
  'description_identifier': ['incorporated-on'],
  'kind': 'searchresults#company',
  'matches': {'snippet': [1, 3]},
  'address': {'premises': '3000a',
   'country': 'United Kingdom',
   'postal_code': 'PO15 7FX',
   'address_line_1': 'Parkway Whiteley',
   'locality': 'Hampshire'},
  'address_snippet': '3000a Parkway Whiteley, Hampshire, United Kingdom, PO15 7FX',
  'company_status': 'active',
  'date_of_creation': '2018-09-04',
  'title': 'AMR CYBERSECURITY LTD',
  'company_type': 'ltd',
  'description': '11551941 - Incorporated on  4 September 2018',
  'company_number': '11551941',
  'snippet': 'ARM CYBERSECURITY '}]

It's also useful for identifying the company number of your target entity. So in this case, if you were looking for the chip designer who revolutionised computing, you would be looking for 02557590.

Now, you've probably noticed that the main payload within the jsonDict is under the "items" key. These items are contained in a list - because it's the list of items returned by the search. That means that our method ch.visualize_json() produces an AttributeError, because visualize_json assumes a dictionary.

So visualize_json is available as a standalone function

>>> from datagrab.interpret_response.json_visualize import visualize_json
>>> visualize_json(ch.json_tree_traverse(["items"])[0])
Root
├── address
│   ├── address_line_1
│   ├── address_line_2
│   ├── locality
│   ├── postal_code
│   └── premises
├── address_snippet
├── company_number
├── company_status
├── company_type
├── date_of_creation
├── description
├── description_identifier
├── kind
├── links
│   └── self
├── matches
│   ├── snippet
│   └── title
├── snippet
└── title

Who has a stake in this company?

Companies house provides an endpoint to find persons (natural or corporate) with significant control.

Here's how to access that:

>>> from compynieshouse.significant_control_query import SignificantControlQuery

>>> scq = SignificantControlQuery(
    appKey="<my_companies_house_API_key>",
    company_code="02557590")

>>> scq.visualize_json()
Root
├── active_count
├── ceased_count
├── items
├── items_per_page
├── links
│   └── self
├── start_index
└── total_results

Note again that the key payload is under the "items", which is effectively a list of dictionaries.

This is what such a record looks like:

>>> visualize_json(scq.json_tree_traverse(["items"])[0])
Root
├── address
│   ├── address_line_1
│   ├── country
│   ├── locality
│   ├── postal_code
│   └── premises
├── etag
├── identification
│   ├── country_registered
│   ├── legal_authority
│   ├── legal_form
│   ├── place_registered
│   └── registration_number
├── kind
├── links
│   └── self
├── name
├── natures_of_control
└── notified_on

Who are the company officers?

>>> from compynieshouse.officer_query import CompanyOfficers
>>> from datagrab.interpret_response.json_visualize import visualize_json

>>> cos = CompanyOfficers("<my_companies_house_API_key>",
                    "02557590",
                    )
>>> cos.visualize_json()
Root
├── active_count
├── etag
├── inactive_count
├── items
├── items_per_page
├── kind
├── links
│   └── self
├── resigned_count
├── start_index
└── total_results


>>> officers_records = cos.json_tree_traverse(["items"])[0]

>>> visualize_json(officers_records)
Root
├── address
│   ├── address_line_1
│   ├── address_line_2
│   ├── locality
│   └── postal_code
├── appointed_on
├── links
│   └── officer
│       └── appointments
├── name
└── officer_role

What other appointments does this officer have?

First, we need the ID of the officer:

>>> officers_records["links"]
{'officer': {'appointments': '/officers/4yYi8Ok5MbG3QNg8t05GUZF-u-U/appointments'}}

>>> from compynieshouse.officer_appointments import OfficerAppointments

>>> oa = OfficerAppointments("<my_companies_house_API_key>",
                    "4yYi8Ok5MbG3QNg8t05GUZF-u-U", # officer ID
                    )
>>> oa.visualize_json()
Root
├── etag
├── is_corporate_officer
├── items
├── items_per_page
├── kind
├── links
│   └── self
├── name
├── start_index
└── total_results

>>> oa.jsonDict["name"]
'Carolyn HERZOG'

How many appointments does Carolyn Herzog have?

>>> len(oa.json_tree_traverse(["items"]))
1

Just the one. What details about that appointment does Companies House provide us?

>>> visualize_json(oa.json_tree_traverse(["items"])[0])
Root
├── address
│   ├── address_line_1
│   ├── address_line_2
│   ├── locality
│   └── postal_code
├── appointed_on
├── appointed_to
│   ├── company_name
│   ├── company_number
│   └── company_status
├── links
│   └── company
├── name
├── name_elements
│   ├── forename
│   ├── surname
│   └── title
└── officer_role