gavagai-node

Node.js helper library for Gavagai API

Install

$ npm install gavagai

Api key

Get your own api key for free at Gavagai Developer Portal.

Use

See Gavagai API documentation for details about available API resources.

Lexicon

Lookup a word in Gavagai living lexicon.

var gavagai = require('gavagai');
var client = gavagai('GAVAGAI_APIKEY');

// getting all info about 'good'
client.lexicon('good', function (err, data) {
    console.log('lexicon returned:', data);
});

Stories

Given a set of texts, return summaries as a set of stories.

var gavagai = require('gavagai');
var client = gavagai('GAVAGAI_APIKEY');

var texts = [
    'Stayed here for 3 nights at the beginning of a trip of California. Could not say enough good things about the hotel Monaco. Amazing staff, amazing rooms and the location is brilliant! First stay at a Kimpton hotel, but definitely not the last!!!',
    'I did a lot of research looking for a hotel suite for our family vacation in San Francisco. The Hotel Monaco was a perfect choice. What friendly and delightful staff. I will miss the Grand Cafe, but I will make sure to come back to see their new offerings.',
    'My partner and I spent four nights here over New Years and loved it. Super staff; lovely, quiet room; excellent location within easy walking to much of Downtown and an overall experience that was perfect.'
];

client.stories(texts, function(err, data) {
    if (err) {
        console.error('error:', err);
    }
    console.log('stories returned:', data);
});

Topics

Given a set of texts, return significant topics and keywords.

client.topics(texts, function(err, data) {
    if (err) {
       console.error('error:', err);
    }
    console.log('stories returned:', data);
});

Keywords

Extract salient concepts from a collection of texts. Order by number of occurrences.

client.keywords(texts, function(err, data) {
    if (err) {
       console.error('error:', err);
    }
    console.log('keywords returned:', data);
});

Tonality

For each text, return their tonality based on lexical analysis in multiple dimensions. This method accepts either a single string, or an array of texts as input.

client.tonality("it’s beautiful and makes me want to cry", function(err, data) {
    if (err) {
       console.error('error:', err);
    }
    console.log('stories returned:', data);
});

Options

All methods accepts options as a second paramter. Options are corresponding to API parameters. E.g. if I have a set of french texts that I want to analyze:

client.topics(texts, { language: 'fr' }, function(err, data) {
    if (err) {
       console.error('error:', err);
    }
    console.log('stories returned:', data);
});

Document IDs

In some situations documents have a specific identifiers. In that case, you can use an array of objects that have properties id and body.

var texts = [
    {
      id: '1',
      body: 'Stayed here for 3 nights at the beginning of a trip of California. Could not say enough good things about the hotel Monaco. Amazing staff, amazing rooms and the location is brilliant! First stay at a Kimpton hotel, but definitely not the last!!!',
    },
    {
      id: '2',
      body: 'I did a lot of research looking for a hotel suite for our family vacation in San Francisco. The Hotel Monaco was a perfect choice. What friendly and delightful staff. I will miss the Grand Cafe, but I will make sure to come back to see their new offerings.',
    },
    {
      id: '3',
      body: 'My partner and I spent four nights here over New Years and loved it. Super staff; lovely, quiet room; excellent location within easy walking to much of Downtown and an overall experience that was perfect.'
    }
];

Promises

One method for for handling the common problem with deeply nested callbacks, is to use promises. Gavagai-node uses the Q library to provide an alternative to callbacks.

See example in promises.js for how to make an api call with promises.

Running tests

Start by install grunt-cli:

$ npm install -g grunt-cli

Execute specs:

$ grunt

Execute smoke tests against live API:.

$ grunt smoke