nodemailer/mongo-cursor-pagination

Name: mongo-cursor-pagination

Owner: nodemailer

Description: Cursor-based pagination for Mongo

Forked from: mixmaxhq/mongo-cursor-pagination

Created: 2017-11-26 21:43:22.0

Updated: 2017-11-26 21:43:24.0

Pushed: 2017-11-26 21:52:39.0

Homepage: null

Size: 62

Language: JavaScript

GitHub Committers

UserMost Recent Commit# Commits

Other Committers

UserEmailMost Recent Commit# Commits

README

NB!

This module is a fork of (mongo-cursor-pagination)[https://github.com/mixmaxhq/mongo-cursor-pagination] by MixMax. The fork was needed to keep supporting Node.js v6 as the upstream version started using async..await.

Use the following command to install this module:

npm install mongo-cursor-pagination-node6 --save

And then in your code:

const MongoPaging = require('mongo-cursor-pagination-node6');

mongo-cursor-pagination

This module aids in implementing “cursor-based” pagination using Mongo range queries or relevancy-based search results. This module is currently used in production for the Mixmax API to return millions of results a day.

Background

See this blog post for background on why this library was built.

API Pagination is typically implemented one of two different ways:

  1. Offset-based paging. This is traditional paging where skip and limit parameters are passed on the url (or some variation such as page_num and count). The API would return the results and some indication of whether there is a next page, such as has_more on the response. An issue with this approach is that it assumes a static data set; if collection changes while querying, then results in pages will shift and the response will be wrong.

  2. Cursor-based paging. An improved way of paging where an API passes back a “cursor” (an opaque string) to tell the caller where to query the next or previous pages. The cursor is usually passed using query parameters next and previous. It's implementation is typically more performant that skip/limit because it can jump to any page without traversing all the records. It also handles records being added or removed because it doesn't use fixed offsets.

This module helps in implementing #2 - cursor based paging - by providing a method that make it easy to query within a Mongo collection. It also helps by returning a url-safe string that you can return with your HTTP response (see example below).

Here are some examples of cursor-based APIs:

Install

npm install mongo-cursor-pagination --save

Usage
find()

Find will return ordered and paged results based on a field (paginatedField) that you pass in.

Call find() with the following parameters:

erforms a find() query on a passed-in Mongo collection, using criteria you specify. The results
re ordered by the paginatedField.

param {MongoCollection} collection A collection object returned from the MongoDB library's
  `db.collection(<collectionName>)` method.
param {Object} params
  -query {Object} The find query.
  -limit {Number} The page size. Must be between 1 and `config.MAX_LIMIT`.
  -fields {Object} Fields to query in the Mongo object format, e.g. {_id: 1, timestamp :1}.
    The default is to query all fields.
  -paginatedField {String} The field name to query the range for. The field must be:
      1. Orderable. We must sort by this value. If duplicate values for paginatedField field
        exist, the results will be secondarily ordered by the _id.
      2. Indexed. For large collections, this should be indexed for query performance.
      3. Immutable. If the value changes between paged queries, it could appear twice.
      4. Complete. A value must exist for all documents.
    The default is to use the Mongo built-in '_id' field, which satisfies the above criteria.
    The only reason to NOT use the Mongo _id field is if you chose to implement your own ids.
  -sortAscending {Boolean} True to sort using paginatedField ascending (default is false - descending).
  -next {String} The value to start querying the page.
  -previous {String} The value to start querying previous page.
param {Function} done Node errback style function.

Example:

MongoClient = require('mongodb').MongoClient;
MongoPaging = require('mongo-cursor-pagination');

oClient.connect('mongodb://localhost:27017/mydb', (err, db) => {
db.collection('myobjects').insertMany(
    [
        {
            counter: 1
        },
        {
            counter: 2
        },
        {
            counter: 3
        },
        {
            counter: 4
        }
    ],
    err => {
        // Query the first page.
        MongoPaging.find(
            db.collection('myobjects'),
            {
                limit: 2
            },
            (err, result) => {
                console.log(result);

                // Query next page.
                MongoPaging.find(
                    db.collection('myobjects'),
                    {
                        limit: 2,
                        next: result.next // This queries the next page
                    },
                    (err, result) => {
                        console.log(result);
                    }
                );
            }
        );
    }
);

Output:

 1 { results:
 { _id: 580fd16aca2a6b271562d8bb, counter: 4 },
 { _id: 580fd16aca2a6b271562d8ba, counter: 3 } ],
xt: 'eyIkb2lkIjoiNTgwZmQxNmFjYTJhNmIyNzE1NjJkOGJhIn0',
sMore: true }
 2 { results:
 { _id: 580fd16aca2a6b271562d8b9, counter: 2 },
 { _id: 580fd16aca2a6b271562d8b8, counter: 1 } ],
evious: 'eyIkb2lkIjoiNTgwZmQxNmFjYTJhNmIyNzE1NjJkOGI5In0',
xt: 'eyIkb2lkIjoiNTgwZmQxNmFjYTJhNmIyNzE1NjJkOGI4In0',
sMore: false }
search()

Search uses Mongo's text search feature and will return paged results ordered by search relevancy. As such, and unlike find(), it does not take a paginatedField parameter.

erforms a search query on a Mongo collection and pages the results. This is different from
ind() in that the results are ordered by their relevancy, and as such, it does not take
 paginatedField parameter. Note that this is less performant than find() because it must
erform the full search on each call to this function. Also note that results might change

param {MongoCollection} collection A collection object returned from the MongoDB library's
  `db.collection(<collectionName>)` method. This MUST have a Mongo $text index on it.
  See https://docs.mongodb.com/manual/core/index-text/.
param {String} searchString String to search on.
param {Object} params
  -query {Object} The find query.
  -limit {Number} The page size. Must be between 1 and `config.MAX_LIMIT`.
  -fields {Object} Fields to query in the Mongo object format, e.g. {title :1}.
    The default is to query ONLY _id (note this is a difference from `find()`).
  -next {String} The value to start querying the page. Defaults to start at the beginning of
    the results.
param {Function} done Node errback style function.

Example:

MongoClient = require('mongodb').MongoClient;
MongoPaging = require('mongo-cursor-pagination');

oClient.connect('mongodb://localhost:27017/mydb', (err, db) => {
db.collection('myobjects').ensureIndex(
    {
        mytext: 'text'
    },
    err => {
        db.collection('myobjects').insertMany(
            [
                {
                    mytext: 'dogs'
                },
                {
                    mytext: 'dogs cats'
                },
                {
                    mytext: 'dogs cats pigs'
                }
            ],
            err => {
                // Query the first page.
                MongoPaging.search(
                    db.collection('myobjects'),
                    'dogs',
                    {
                        fields: {
                            mytext: 1
                        },
                        limit: 2
                    },
                    (err, result) => {
                        console.log(result);

                        // Query next page.
                        MongoPaging.search(
                            db.collection('myobjects'),
                            'dogs',
                            {
                                limit: 2,
                                next: result.next // This queries the next page
                            },
                            (err, result) => {
                                console.log(result);
                            }
                        );
                    }
                );
            }
        );
    }
);

Output:

 1  { results:
 { _id: 581668318c11596af22a62de, mytext: 'dogs', score: 1 },
 { _id: 581668318c11596af22a62df, mytext: 'dogs cats', score: 0.75 } ],
xt: 'WzAuNzUseyIkb2lkIjoiNTgxNjY4MzE4YzExNTk2YWYyMmE2MmRmIn1d' }
 2 { results:
 { _id: 581668318c11596af22a62e0, score: 0.6666666666666666 } ] }
Use with ExpressJS

A popular use of this module is with Express to implement a basic API. As a convenience for this use-case, this library exposes a findWithReq function that takes the request object from your Express middleware and returns results:

So this code using find():

er.get('/myobjects', (req, res, next) => {
MongoPaging.find(
    db.collection('myobjects'),
    {
        query: {
            userId: req.user._id
        },
        paginatedField: 'created',
        fields: {
            // Also need to read req.query.fields to use to filter these fields
            _id: 1,
            created: 1
        },
        limit: req.query.limit, // Also need to cap this to 25
        next: req.query.next,
        previous: req.query.previous
    },
    function(err, result) {
        if (err) {
            next(err);
        } else {
            res.json(result);
        }
    }
);

Is more elegant with findWithReq():

er.get('/myobjects', (req, res, next) => {
MongoPaging.findWithReq(
    req,
    db.collection('myobjects'),
    {
        query: {
            userId: req.user._id
        },
        paginatedField: 'created',
        fields: {
            _id: 1,
            created: 1
        },
        limit: 25 // Upper limit
    },
    function(err, result) {
        if (err) {
            next(err);
        } else {
            res.json(result);
        }
    }
);

findWithReq() also handles basic security such as making sure the limit and fields requested on the URL are within the allowed values you specify in params.

Number of results

If the limit parameter isn't passed, then this library will default to returning 50 results. This can be overridden by setting mongoPaging.config.DEFAULT_LIMIT = <new default limit>;. Regardless of the limit passed in, a maximum of 300 documents will be returned. This can be overridden by setting mongoPaging.config.MAX_LIMIT = <new max limit>;.

Indexes for sorting

mongo-cursor-pagination uses _id as a secondary sorting field when providing a paginatedField property. It is recommended that you have an index for optimal performance. Example:

oPaging.find(db.people, {
ery: {
name: 'John'

ginatedField: 'city'
mit: 25,

For the above query to be optimal, you should have an index like:

eople.ensureIndex({
me: 1,
ty: 1,
d: 1

Running tests

To run tests, you first must start a Mongo server on port 27017 and then run npm test.

Changelog
Future ideas

This work is supported by the National Institutes of Health's National Center for Advancing Translational Sciences, Grant Number U24TR002306. This work is solely the responsibility of the creators and does not necessarily represent the official views of the National Institutes of Health.