Flashtext.js

JavaScript (ES6+ source and zero dependencies) port of the amazing Python package flashtext

This module can be used to replace keywords in sentences or extract keywords from sentences.

Note - This is a quick experimental port. Please post issues for anything that doesn't work or can be improved

Installation

    $ npm install flashtext.js --save

or

    $ yarn add flashtext.js

Usage

Extract Keywords

const processor = new KeywordProcessor();
 
processor.addKeyword('Big Apple', 'New York');
processor.addKeyword('Bay Area');
 
const keywordsFound = processor.extractKeywords(
    'I love Big Apple and Bay Area.'
);
console.log(keywordsFound);
// output - ['New York', 'Bay Area']

Replace Keywords

const processor = new KeywordProcessor();
 
processor.addKeyword('Big Apple', 'New York');
processor.addKeyword('New Delhi', 'NCR Region');
 
const newSentence = processor.replaceKeywords(
    'I love Big Apple and new delhi.'
);
console.log(newSentence);
// output - 'I love New York and NCR Region.'

Case Sensitive Example

const processor = new KeywordProcessor(true);
 
processor.addKeyword('Big Apple', 'New York');
processor.addKeyword('Bay Area');
 
const keywordsFound = processor.extractKeywords(
    'I love big Apple and Bay Area.'
);
console.log(keywordsFound);
// output - ['Bay Area']

Multiple Keywords Entry using Object

const processor = new KeywordProcessor();
 
processor.addKeywordsFromObject({
    java: ['java_2e', 'java programming'],
    python: ['python2.7', 'python programming'],
});
 
const keywordsFound = processor.extractKeywords(
    'I love java_2e and python programming.'
);
console.log(keywordsFound);
// output - ['java', 'python']
 
const newSentence = processor.replaceKeywords(
    'I prefer python over java programming.'
);
console.log(newSentence);
// output - 'I prefer python over java.'

Multiple Keywords Entry using Arrays

const processor = new KeywordProcessor();
 
processor.addKeywordsFromArray(['java', 'product manager']);
 
const keywordsFound = processor.extractKeywords(
    'I am a product manager for a java platform'
);
 
console.log(keywordsFound);
// output - ['product manager', 'java']

Remove Keywords (One/Multiple)

const processor = new KeywordProcessor();
 
processor.addKeywordsFromArray(['react', 'angular', 'vue', 'javascript']);
processor.removeKeywordsFromArray(['angular', 'vue']);
processor.removeKeyword('javascript');
 
const keywordsFound = processor.extractKeywords(
    'I know React, Angular and Vue. I love JavaScript.'
);
console.log(keywordsFound);
// output - ['react'];

Add characters as a part of word

const processor = new KeywordProcessor();
 
processor.addKeyword('Big Apple');
 
const before = processor.extractKeywords('I love Big Apple/Bay Area.');
console.log(before);
// output - ['Big Apple']
 
processor.addNonWordBoundaries('/');
const after = processor.extractKeywords('I love Big Apple/Bay Area.');
console.log(after);
// output - []

References

The original paper published on FlashText algorithm

The article published on Medium freeCodeCamp

License

This code is under MIT license.