fi-lter

Fi Lter

MongoDB and Mongoose aggregation filter and search helper.

This module is intended as a simple abstraction layer to perform basic but powerful matching with a concatenation of one or multiple model's fields.

Diacritic Matching

The diacritic insensitive functionality is handled by Fi Di Regex. If you find an issue with this functionality, please open an issue in the Fi Di Regex Issues page.

Important

Although this module's functionality is nearly complete the documentation is still in progress. In the meantime, the docs may help you a bit.

Installation

npm i fi-lter

Remember to use --save if you're using a NPM version less than 5.x.x.

Usage

const mongoose = require('mongoose');
const filter = require('fi-lter');

/* Let's assume this is the string the user provided */
const queryText = 'google';

/**
 * This are the fields you want to search in and will be used to generate a slug
 * field where the diacritic-insensitive matching will occur.
 *
 * IMPORTANT: This stage should be prebuilt unless you use dynamic values.
 */
const KEYWORDS_SLUG = filter.keywordsSlug([
  '$brand', '$model', '$color', '$serial', {
    $substrBytes: ['$year', 0, -1] // Convert Number to String
  }
]);

/**
 * This are the fields you want to keep after grouping the results by _id.
 *
 * They will be added using a $group stage with a $first match.
 *
 * IMPORTANT: This stage should be prebuilt unless you use dynamic values.
 */
const KEYWORDS_GROUP = filter.keywordsGroup([
  'brand', 'model', 'year', 'serial', 'price', 'createdAt', 'updatedAt'
]);

/* This example uses a device's information as source */
Device = mongoose.model('device', new mongoose.Schema({
  brand: String,
  model: String,
  color: String,
  serial: String,
  year: Number,
  price: Number
}, {
  timestamps: true
}));

/* Start by creating an aggregation query */
const query = Device.aggregate();

// Initial aggregation stages...
// E.g.: query.match({...});

/* Build and append the keyword filter stages */
query.append(filter.byKeywords(queryText, KEYWORDS_SLUG, KEYWORDS_GROUP));

// More aggregation stages...
// E.g.: query.project({...});

query.then(results => {
  /* Here, results should be an Array of Devices matching 'google' by their
   * brand, model, year, serial or color, sorted by their match score */
  console.log(results);
}).catch((err) => {
  console.error(err);
});