I have been reconciling some monthly archive files of XBRL documents (April, May and June 2015) with the daily files that used to be available. I started using the daily XBRL data sometime in July when the first daily file listed was April 15th 2015. I had not appreciated that only a fixed number of daily files are shown/available on the CoHo web site so I deleted the daily archive files after extracting their contents: why waste the space if they are on the CoHo web site (oops).
As a result of this reconciliation process I discovered discrepancies: files missing from the monthly archives that used to be in the respective daily archive. An example is the document for 05681931_20141231. This document was present in Accounts_Bulk_Data-2015-04-21.zip but is not in the monthly archive for April. I understand from conversations with the front end support team these discrepancies arise when accounts have been removed from the register. You can see this is a problem because, for example, the accounts still exist on the Corefiling web site even though the accounts are no longer in the register. As a result, invalid data is being promulgated though not directly by Companies House.
So my question is to ask if there is a feature of the API that identifies accounts that have been removed from the register? Such a feature will make it possible and efficient for 3rd party consumers to update the Companies House information they hold.
Thanks
BIll Seddon
Bill,
We are currently working on an API which will stream events as they occur. One of these event streams will be changes to a Companies filing history, including the removal of a record. The general concept is that consumers can then subscribe to these streams and consume the events, which I think covers the requirement detailed in the post.
We do not have any firm release dates for the functionality but will publish on the forum once more concrete plans are developed.
Thanks,
Mark.
Thanks for your response. Is there a specification for the planned enhancements available for comment? It will be great to have access to information about items removed from the register. It may be an artefact of your your brief description of the planned enhancements but it sounds as if we consumers will still be left with invalid data if, for whatever reason, we were not listening or could not be contacted at the time the change event is broadcast.
Connection to any service over a public network is not going to be reliable. How will the proposed service enhancements ensure the fact of an event data is not missed by 3rd party consumers? Missing the event of a new set of accounts is not so bad because the same information will be able available later in an archive format . Missing the event of removing errored accounts from the register is more of a problem because this information is not currently available and your description does not suggest there is any plan to archive information about the removal event.
When I reconciled the the monthly archive with the daily archive the number of discrepancies (removals) was very low: less than 10 out of more than 100K in any of of the months. The only information about the removal of a set of accounts that I need (though I recognize others may have additional requirements) is access to the company number and filing date of removed accounts. This is a tiny piece of information (18 bytes) that it will be great to be able to access on-demand.