mongoose-query-parser
!NPM versionnpm-imagenpm-url
!Downloadsdownload-badgenpm-url
Convert url query string to MongooseJs friendly query object including advanced filtering, sorting, population, string template, type casting and many more...
The library is built highly inspired by api-query-params
Features
- Supports the most of MongoDB operators (
$in
,$regexp
,$exists
) and features including skip, sort, limit & MongooseJS population
- Auto type casting of
Number
,RegExp
,Date
,Boolean
andnull
- String templates/predefined queries (i.e.
firstName=${my_vip_list}
)
- Allows customization of keys and options in query string
Installation
```sh npm install mongoose-query-parser --save ```Usage
API
```js import { MongooseQueryParser } from 'mongoose-query-parser'; const parser = new MongooseQueryParser(options?: ParserOptions) parser.parse(query: string, predefined: any) : QueryOptions ```Arguments
ParserOptions
: object for advanced options (See below) optional
query
: query string part of the requested API URL (ie,firstName=John&limit=10
). Works with already parsed object too (ie,{status: 'success'}
) required
predefined
: object for predefined queries/string templates optional
Returns
QueryOptions
: object contains the following properties:
- `filter` which contains the query criteria
- `populate` which contains the query population. Please see [Mongoose Populate](http://mongoosejs.com/docs/populate.html) for more details
- `select` which contains the query projection
- `sort`, `skip`, `limit` which contains the cursor modifiers for paging purpose
Example
```js import { MongooseQueryParser } from 'mongoose-query-parser'; const parser = new MongooseQueryParser(); const predefined = { vip: { name: { $in: 'Google', 'Microsoft', 'NodeJs' } }, sentStatus: 'sent' }; const parsed = parser.parse('${vip}&status=${sentStatus}×tamp>2017-10-01&author.firstName=/john/i&limit=100&skip=50&sort=-timestamp&select=name&populate=children.firstName,children.lastName', predefined); // { // select: { name : 1 }, // populate: { path: 'children', select: 'firstName lastName' }, // sort: { timestamp: -1 }, // skip: 50, // limit: 100, // filter: { // name: {{ $in: 'Google', 'Microsoft', 'NodeJs' }}, // status: 'sent', // timestamp: { '$gt': 2017-09-30T14:00:00.000Z }, // 'author.firstName': /john/i // } // } ```Supported features
Filtering operators
| MongoDB | URI | Example | Result | | --------- | -------------------- | ----------------------- | -------------------------------------------------------- | |$eq
| key=val
| type=public
| {filter: {type: 'public'}}
|
| $gt
| key>val
| count>5
| {filter: {count: {$gt: 5}}}
|
| $gte
| key>=val
| rating>=9.5
| {filter: {rating: {$gte: 9.5}}}
|
| $lt
| key<val
| createdAt<2017-10-01
| {filter: {createdAt: {$lt: 2017-09-30T14:00:00.000Z}}}
|
| $lte
| key<=val
| score<=-5
| {filter: {score: {$lte: -5}}}
|
| $ne
| key!=val
| status!=success
| {filter: {status: {$ne: 'success'}}}
|
| $in
| key=val1,val2
| country=GB,US
| {filter: {country: {$in: ['GB', 'US']}}}
|
| $nin
| key!=val1,val2
| lang!=fr,en
| {filter: {lang: {$nin: ['fr', 'en']}}}
|
| $exists
| key
| phone
| {filter: {phone: {$exists: true}}}
|
| $exists
| !key
| !email
| {filter: {email: {$exists: false}}}
|
| $regex
| key=/value/<opts>
| email=/@gmail\.com$/i
| {filter: {email: /@gmail.com$/i}}
|
| $regex
| key!=/value/<opts>
| phone!=/^06/
| {filter: {phone: { $not: /^06/}}}
|
For more advanced usage ($or
, $type
, $elemMatch
, etc.), pass any MongoDB query filter object as JSON string in the filter
query parameter, ie:
```js
parser.parse('filter={"$or":{"key1":"value1"},{"key2":"value2"}}&name=Telstra');
// {
// filter: {
// $or:
// { key1: 'value1' },
// { key2: 'value2' }
// ,
// name: 'Telstra'
// },
// }
```
Populate operators
- Useful to populate sub-document(s) in query. Works with
MongooseJS
. Please see Mongoose Populate for more details
- Supports deep populate with delimiter ":" to indicate deep populate
test-populate.spec.ts
for more details
- See Mongoose Docs
- Allows to populate with only selected fields with "." to indicate field selection
- Default operator key is
populate
Skip / Limit operators
- Useful to limit the number of records returned
- Default operator keys are
skip
andlimit
Select operator
- Useful to limit fields to return in each records
- Default operator key is
select
- It accepts a comma-separated list of fields. Default behavior is to specify fields to return. Use
-
prefixes to return all fields except some specific fields
- Due to a MongoDB limitation, you cannot combine inclusion and exclusion semantics in a single projection with the exception of the id field
Sort operator
- Useful to sort returned records
- Default operator key is
sort
- It accepts a comma-separated list of fields. Default behavior is to sort in ascending order. Use
-
prefixes to sort in descending order
Keys with multiple values
Any operators which process a list of fields ($in
, $nin
, sort and projection) can accept a comma-separated string or multiple pairs of key/value:
country=GB,US
is equivalent tocountry=GB&country=US
sort=-createdAt,lastName
is equivalent tosort=-createdAt&sort=lastName
Embedded documents using .
notation
Any operators can be applied on deep properties using .
notation:
```js
parser.parse('followers0.id=123&sort=-metadata.createdat');
// {
// filter: {
// 'followers0.id': 123,
// },
// sort: { 'metadata.createdat': -1 }
// }
```
Automatic type casting
The following types are automatically casted:Number
, RegExp
, Date
, Boolean
and null
string is also casted:
```js
parser.parse('date=2017-10-01&boolean=true&integer=10®exp=/foobar/i&null=null');
// {
// filter: {
// date: 2017-09-30T14:00:00.000Z,
// boolean: true,
// integer: 10,
// regexp: /foobar/i,
// null: null
// }
// }
```
If you need to disable or force type casting, you can wrap the values with string()
, date()
built-in casters or by specifying your own custom functions (See below):
```js
parser.parse('key1=string(10)&key2=date(2017-10-01)&key3=string(null)');
// {
// filter: {
// key1: '10',
// key2: 2017-09-30T14:00:00.000Z,
// key3: 'null'
// }
// }
```
String template for predefined query/variable
Best to reducing the complexity/length of the query string with the ES template literal delimiter as an "interpolate" delimiter (i.e.${something}
)
```js
const parser = new MongooseQueryParser();
const preDefined = {
isActive: { status: { $in: 'In Progress', 'Pending' } },
secret: 'mysecret'
};
parser.parse('${isActive}&secret=${secret}', preDefined);
// {
// filter: {
// status: { $in: 'In Progress', 'Pending' },
// secret: 'mysecret'
// }
// }
```
Available options (opts
)
Customize operator keys
The following options are useful to change the operator default keys:populateKey
: custom populate operator key (default ispopulate
)
skipKey
: custom skip operator key (default isskip
)
limitKey
: custom limit operator key (default islimit
)
selectKey
: custom select operator key (default isselect
)
sortKey
: custom sort operator key (default issort
)
filterKey
: custom filter operator key (default isfilter
)
Date format
dateFormat
: set date format for auto date casting. Default is ISO8601 format if no custom datetime format provided
- Allows multiple formats. Works with luxon
- For the supported format tokens, please refer to luxon document
Blacklist
The following options are useful to specify which keys to use in thefilter
object. (ie, avoid that authentication parameter like apiKey
ends up in a mongoDB query). All operator keys are (sort
, limit
, etc.) already ignored.
blacklist
: filter on all keys except the ones specified
Add custom casting functions
You can specify you own casting functions to apply to query parameter values, either by explicitly wrapping the value in URL with your custom function name (See example below) or by implicitly mapping a key to a functioncasters
: object to specify custom casters, key is the caster name, and value is a function which is passed the query parameter value as parameter.
lowercase: val => val.toLowerCase(),
int: val => parseInt(val, 10),
},
castParams: {
key3: 'lowercase'
}});
parser.parse('key1=lowercase(VALUE)&key2=int(10.5)&key3=ABC');
// {
// filter: {
// key1: 'value',
// key2: 10,
// key3: 'abc'
// }
// }
```