mongodb-language-model

Parses a MongoDB query and creates an abstract syntax tree (AST) with part of speech
tagging. Currently, only [strict extended json][docs-extended-json] syntax is
supported (which means keys have to be surrounded by double quotes and values
have to match the strict syntax of the extended-json format.);

Usage

The main module exposes two functions: accepts(queryStr) and parse(queryStr).

#### accepts(queryStr)

The accepts(queryStr) function takes a query string and returns true if the
string is a valid MongoDB query, false otherwise.

Example:

``javascript
var accepts = require('mongodb-language-model').accepts;
var assert = require('assert');

assert.ok(accepts('{"foo": 1}'));
assert.ok(accepts('{"age": {"$gt": 35}}'));
assert.ok(accepts('{"$or": [{"email": {"$exists": true}}, {"phone": {"$exists": true}}]}'));

assert.equal(accepts('{"$invalid": "key"}'), false);
`

#### parse(queryStr)

The parse(queryStr) function takes a query string and returns an abstract
syntax tree (AST) as a javascript object, if the query is valid. If the
query is not valid, the function throws a pegjs.SyntaxError with a message
explaining the failure.

Example:

`javascript
var parse = require('mongodb-language-model').parse;
var assert = require('assert');
var pegjs = require('pegjs');

var ast = parse('{"foo": "bar"}');
assert.deepEqual(ast, {
'pos': 'expression',
'clauses': [
{
'pos': 'leaf-clause',
'key': 'foo',
'value': {
'pos': 'leaf-value',
'value': 'bar'
}
}
]
});
`

UML diagram

This is the hierarchical model that is created when a query is parsed:

![](./docs/query_language_uml.png)

Installation

`
npm install --save mongodb-language-model
`

Testing

`
npm test
``

[docs-extended-json]: https://docs.mongodb.com/manual/reference/mongodb-extended-json/