Mongo Aggregation Debugger


Mongo Aggregation Debugger helps debug MongoDb aggregation queries
by being able to visualize each stage of the pipeline

Why use it

It is pretty hard to understand why a specific aggregation query fails or doesn't output the
right results since it can be pretty complex and go through a lot of stages before returning values.

The Mongo Aggregation Debugger helps you understand what is going on by either:
- outputting in the console the results of each stage of the aggregation pipeline
- returning an array of results of each stage of teh aggregation pipeline for programmatic use
- running the query in a temporary database and outputting the results,
very useful for automated testing

How it works

You give the debugger access to your instance of mongodb, and it creates a temporary collection
in which it will run each stage of the aggregation query in series.
The temporary database is dropped after each debug.

Install

``
npm install mongo-aggregation-debugger
`

Instantiation

`
var mad = require('mongo-aggregation-debugger')();
`

You can provide an optional object as an argument to specify the mongodb connection information:

key | default value | description
------------ | ------------- | -------------
host | localhost | mongodb host name
port | 27017 | mongodb port number
username | null | (optional) username of the mongodb instance
password | null | (optional) password of the mongodb instance
options | {} | (optional) additional mongodb options

API

$3

This method outputs in the console the result of each stage of the aggregation pipeline.

#### Use
log(data, query[, options][, callback])

argument | type | values | description
------------ | ------------- | ------------- | -------------
data | array | | The data to run the query against
query | array | | The aggregation query
options | object | showQuery: boolean | Whether to show the query of the stage being run or not
callback | function(err) | | The callback returned when all stages were executed

#### Example:
`javascript
var mad = require('mongo-aggregation-debugger')();

var data = [{
foo: 'bar',
test: true,
array: [ 1, 2, 3 ]
}, {
foo: 'bar2',
test: false,
array: [ 10, 20 ]
}];

var query = [{
'$match': {
test: true
}
}, {
'$project': {
foo: 1,
array: 1
}
}, {
'$unwind': "$array"
}, {
'$group': {
_id: "$foo",
foo: { $first: "$foo" },
sum: { $sum: "$array" }
}
}];

mad.log(data, query, function (err) {
if (err) {
// do something
}

console.log('All done!');
});
`

Running the code above would output this in your console:


Example with the showQuery option:
`
mad.log(data, query, { showQuery: true }, function (err) {
if (err) {
// do something
}

console.log('All done!');
});
`


$3

This method returns the result of each stage of the aggregation pipeline for programmatic use.

#### Use
stages(data, query[, callback])

argument | type | description
------------ | ------------- | ------------- | -------------
data | array | The data to run the query against
query | array | The aggregation query
callback | function(err, results) | results is an array composed of as many objects as there are stages in the aggregation pipeline. Each object has a query attribute which is the query of the stage and a result attribute with the results of that query

#### Example:
`javascript
var util = require('util');
var mad = require('mongo-aggregation-debugger')();

var data = [{
foo: 'bar',
test: true,
array: [ 1, 2, 3 ]
}, {
foo: 'bar2',
test: false,
array: [ 10, 20 ]
}];

var query = [{
'$match': {
test: true
}
}, {
'$project': {
foo: 1,
array: 1
}
}, {
'$unwind': "$array"
}, {
'$group': {
_id: "$foo",
foo: { $first: "$foo" },
sum: { $sum: "$array" }
}
}];

mad.stages(data, query, function (err, results) {
if (err) {
// do something
}

console.log(util.inspect(results, { depth: null }));
});
`

The output is:
`javascript
[ { query: [ { '$match': { test: true } } ],
results:
[ { _id: 5599279a731b5aba47df6d97,
foo: 'bar',
test: true,
array: [ 1, 2, 3 ] } ] },
{ query:
[ { '$match': { test: true } },
{ '$project': { foo: 1, array: 1 } } ],
results: [ { _id: 5599279a731b5aba47df6d97, foo: 'bar', array: [ 1, 2, 3 ] } ] },
{ query:
[ { '$match': { test: true } },
{ '$project': { foo: 1, array: 1 } },
{ '$unwind': '$array' } ],
results:
[ { _id: 5599279a731b5aba47df6d97, foo: 'bar', array: 1 },
{ _id: 5599279a731b5aba47df6d97, foo: 'bar', array: 2 },
{ _id: 5599279a731b5aba47df6d97, foo: 'bar', array: 3 } ] },
{ query:
[ { '$match': { test: true } },
{ '$project': { foo: 1, array: 1 } },
{ '$unwind': '$array' },
{ '$group':
{ _id: '$foo',
foo: { '$first': '$foo' },
sum: { '$sum': '$array' } } } ],
results: [ { _id: 'bar', foo: 'bar', sum: 6 } ] } ]
`

$3

This method only runs the entire query passed, not all the stages seperately. It is useful for automated tests since it creates and drops a temporary database.

#### Use
exec(data, query[, callback])

argument | type | description
------------ | ------------- | ------------- | -------------
data | array | The data to run the query against
query | array | The aggregation query
callback | function(err, results) | results is the results of the query being run

#### Example:
`javascript
var util = require('util');
var mad = require('mongo-aggregation-debugger')();

var data = [{
foo: 'bar',
test: true,
array: [ 1, 2, 3 ]
}, {
foo: 'bar2',
test: false,
array: [ 10, 20 ]
}];

var query = [{
'$match': {
test: true
}
}, {
'$project': {
foo: 1,
array: 1
}
}, {
'$unwind': "$array"
}, {
'$group': {
_id: "$foo",
foo: { $first: "$foo" },
sum: { $sum: "$array" }
}
}];

mad.exec(data, query, function (err, results) {
if (err) {
// do something
}

console.log(util.inspect(results, { depth: null }));
});
`

The output is:
`javascript
[ { _id: 'bar', foo: 'bar', sum: 6 } ]
`

Unit tests


In order to test this lib you'll need to install mocha: npm install -g mocha.
Then just run the mocha` command at the root of the project.

More info

- MongoDb
- MongoDb Aggregation Framework
- MongoDb Native NodeJS Driver

Contribute

If you think it would make sense to add some features/methods don't hesitate to fork and
make pull requests.

You can contact the main contributor on Twitter

Licence

Distributed under the MIT License.