MongoQueryEngine

Restful, free text and filter based, API and backend using MongoDB

The MongoQueryEngine (MQE) is library you can use to add middleware to your express application. The MQE provides a query layer for items stored in MongoDB. The Rest interface provides both filter based expressions as well as free text search. The interface also provides pagination. Finally there is a small helper jQuery library that can be loaded to use with your fontend. The library helps with query creation and parsing as well as app routing. Finally, the MQE adds SEO support.

Starting MQE Application

To add the MQE middleware to your app simply add app the following to your express app.

var express = require('express');
var app = express();
var http = require('http');
var mqeLib = require('MongoQueryEngine');

mqeLib.init({
    config: require('/path/to/config.js'), // see below
    app: app,
    express: express,
  }, function(){
  /**
   * Setup contains express, app, database, collection and mqe
   */
    var setup = mqeLib.getSetup();

    var server = http.createServer(app);
    var server.listen(3000);
  }
);

config

The MQE uses an config.js file to specify various options for the MQE application. Theres parameters include, database configuration and express configuration.

Here is on overview:

{
    db : { // config for MongoDB
      // connection string for the database
      url             : "mongodb://localhost:27017/wcga",

      // collection where the queryable items are stored
      mainCollection  : "myDataCollection",

      // primary filters.  these filters will have indexes created for theme as well as have remaining
      // counts show in queries.   These attributes are probably your 'suggestions' for further filtering.
      indexedFilters  : ['keywords','organization','anotherAttribute'],  

      // currently MQE only allows one sort option, place the attribute you wish to sort on here
      sortBy : 'title',

      // attributes that will be used in the text search.  Mongo only allows for one text index.  This list of
      // attributes will be combined into that index and ensured on start.
      textIndexes       : ['title','description','organization','anotherAttribute'],

      // attributes that should not be returned in the response objects
      blacklist : ['_id', 'md5','secret'],

      // mark this flag if you mainCollection is the result of a MapReduce opperation.  MQE will handle the marshalling
      // of the 'value' namespace.
      isMapReduce     : true
    },
    server : {
      // server host url
      host : "localhost",

      // port outside world goes to.  most likely 80
      remoteport : 80,

      // local port on machine
      localport : 3003
    },
  logging = {
  	dir : "/var/log/myapp",

  	// max log size
  	maxsize : 10485760
  }
}

Post Process Responses

You post process the 'get' and 'query' responses by adding additional config to mqeLib.init(). Example:

mqeLib.init({
    config: config,
    app: app,
    express: express,
    process : {
      get : function(params, item, callback) {
        // do stuff here
        callback(item);
      },
      query : function(params, items, callback) {
        // do stuff here
        callback(items);
      }
    }
});

As you see in the example above, each process function will be handed the query parameters from the request, the item or items in the response and a callback for when you are finished.