patentlyclear

Patently Clear is an API for the US patent office. Patently Clear provides easy access to the data through full text search and citation graphs. Check out our examples on tonic to see the library in action.

Register for an early access token here.

npm install patentlyclear

Usage

var PatentlyClear = require('patentlyclear');
// You can either pass in your API token into the constructor or have it set as an environment variable
// The library will look for process.env.PC_API_TOKEN if no api token is provided.
var pc = new PatentlyClear(/* your api token */);

Looking up by ID

getById(id, callback) Looks up a document by its id Supplying an id that points to multiple resources (such as an application number) will return all forms of that document.

Application ID

• api.getById('US20140000001') => {application: {...}}

Grant ID

• api.getById('US8877876') => {grant: {...}}

Application Number (both string and number formats are accepted)

• api.getById('13/538,394') => {application: {...}, grant: {...}}
• api.getById(13538394) => {application: {...}, grant: {...}}

pc.getById('US8046721', function(err, data) {
  // data => {grant: {id:..., title:..., abstract:...}}
});

Full text search

search(payload, callback) Search for a set of documents with a query payload.

There are several fields that can be searched on. They are:

• terms [string]: Perform a full text search with a set of keywords.
• inventor [string]: Search on inventor name.
• assignee [string]: Who owns the document.
• filing_date {before: Date, after: Date}: When the document was filed.
• publication_date {before: Date, after: Date}: When the document was published.
• type: [application || grant]: What sort of document it is.
• size int: Total number of documents to return. (Max = 100)

Fields are 'AND'-ed together for search queries, but the term field can support 'OR'-ing.

pc.search({terms: 'dna'}, function(err, data) {
  // data => {hits: [{grant: {}}, {application: {}}]}
});
 
// 'dna' AND 'Genentech'
 
{terms: 'dna', assignee: 'genentech'}
 
// 'dna' AND (filed during 2013)
{
  terms: 'dna',
  filing_date: {
    after: '2013-01-01',
    before: '2014-01-01'
  }
}
 
// 'dna' AND ('cat' OR 'dog' OR 'mouse')
{
  terms: [
    'dna',
    ['cat', 'dog', 'mouse']
  ]
}
 

Scrolling

Some result sets can be very large, so pagination is provided by simply re-running the previous search query with the scroll id provided by the returned data.

var query = {terms: 'dna'};
pc.search(query, function(err, data) {
  query.scroll_id = data.scroll_id;
  pc.search(query, function(err, nextData) {
    // another set of data
  });
});

Analysis

analyze(field, payload, callback) Fetches the top trends in a field for a search payload. The possible fields are:

• trends
• assingee
• filing_date

pc.analyze('trends', {terms: 'dna'}, function(err, data) {
  // data => {hits: [{grant: {}}, {application: {}}]}
});

Citation search

backwardCitation(id, payload, callback) Fetches the backward citations of id (eg. the documents that id cites.) that match a given search query. Optionally provide a search payload to filter documents that are retrieved.

pc.backwardCitation('US8046721', function(err, data) {
  // data => {hits: [{grant: {}}, {application: {}}]}
});
 
pc.backwardCitation('US8046721', {terms: 'password'}, function(err, data) {
  // data => {hits: [{grant: {}}, {application: {}}]}
});

forwardCitation(id, payload, callback) Fetches the forward citations of id (eg. the documents that cite id.) that match a given search query.

pc.forwardCitation('US8046721', function(err, data) {
  // data => {hits: [{grant: {}}, {application: {}}]}
});
 
pc.forwardCitation('US8046721', {terms: 'password'}, function(err, data) {
  // data => {hits: [{grant: {}}, {application: {}}]}
});

License

MIT