Is there a reason there’s HTML in the search results. In particular, searching for a company by name doesn’t seem to return the name as plain text but only formatted:
{
“kind”: “searchresults#company”,
“number”: “03331571”,
“title”: “03331571: BERKELEY GOLDMAN LIMITED: active”,
“snippet”: “The Courtyard, 75A Odsal Rad, Bradford, West Yorkshire BD6 1PN”,
“_score”: 0.019182215,
“reg”: “”,
“name”: “BERKELEY GOLDMAN LIMITED”,
“diss”: “”,
“status”: “active”,
“formattedUrl”: “/company/03331571”,
“date”: “11 March 1997”,
“liquidation_ind”: “”,
“live”: 1
}
Similarly searching by number doesn’t return the company number in an unformatted way, e.g.
“items”: [
{
“formattedUrl”: “/company/07444723”,
“title”: “07444723: CHRINON LTD: active”,
“liquidation_ind”: “”,
“date”: “18 November 2010”,
“_score”: 0.18652505,
“snippet”: “Aston House, Cornwall Avenue, London N3 1LF”,
“live”: “”,
“name”: “CHRINON LTD”,
“number”: “07444723”,
“kind”: “searchresults#company”,
“status”: “active”,
“diss”: “”,
“reg”: 1
}
This means using this programmatically involves parsing, for example, the URL, or stripping out tags from the name field in the first case. That’s not difficult, but I think it’s important that the results should be raw data first, with formatting as an add on in fields that are clearly marked as such.