Get a full REST API with zero coding in less than 30 seconds (seriously) Hyper-Heavily inspired on json-server
This is a rewrite of the great mongodb-rest (https://github.com/ltonetwork/mongodb-rest) package to be compatible to json-server query definitions (_start, _limit, ...).
Changes include:
- Similar API as https://github.com/typicode/json-server
- Cleaner code
A simple but incredibly useful REST API server for MongoDB using Node, Express and the native node.js MongoDB driver.
- Endpoint based policies
- User roles
- Be able to switch to AQP query style
https://www.npmjs.com/package/api-query-params
https://nomadas.gitbook.io/mongo-server https://github.com/ltonetwork/mongodb-rest
Installation is via npm:
npm install mongoapi
You can install globally using -g:
npm install -g mongoapi
Now issue mongoapi on the command line and the server should start.
NOTE: Make sure you are running a MongoDB database in addition to the mongoapi server.
After installation you can quickly test it by issuing the following from the command line:
curl -d '{ "A1" : 201 }' -H "Content-Type: application/json" http://localhost:3000/test/example1
This should add a document to the "test" db.example1 collection:
{
"A1": 201,
"_id": ObjectId("4e90e196b0c7f4687000000e")
}
mongoapi can easily be started programmatically by 'requiring' the module and calling startServer.
const mongodbRest = require('mongoapi/server.js');
mongodbRest.startServer();You can optionally pass in a configuration object:
mongodbRest.startServer(config);When starting from the command line you should have config.json in the current working directory. The project includes an example configuration file.
When starting the server programmatically you can pass in a Javascript object for mongoapi configuration.
Here is an example JSON configuration object:
{
"db": "mongodb://localhost:27017",
"endpoint_root": "server",
"server": {
"port": 3000,
"address": "0.0.0.0"
},
"accessControl": {
"allowOrigin": "*",
"allowMethods": "GET,POST,PUT,DELETE,HEAD,OPTIONS",
"allowCredentials": false
},
"dbAccessControl": {
"foo_database": ["collection1", "collection2"],
"bar_database": ["collection2", "collection3"],
"zoo_database": [],
},
"mongoOptions": {
"serverOptions": {
},
"dbOptions": {
"w": 1
}
},
"humanReadableOutput": true,
"urlPrefix": "",
"schema": {
"foo_database": {
"collection1": {
"definitions": {},
"$schema": "http://json-schema.org/draft-06/schema#",
"$id": "http://json-schema.org/draft-06/schema#",
"type": "object",
"properties": {
"value": {
"$id": "/properties/value",
"type": "boolean",
"title": "Foo boolean value",
"description": "An explanation about the purpose of this instance.",
"default": false,
"examples": [
false
]
}
}
}
}
}
}
db specifies the mongodb connection string for connection to the database. It defaults when not specified.
For documentation on the mongodb connection string: http://docs.mongodb.org/manual/reference/connection-string/
For backward compatibility db can also be set to an object that specified host and port as follows:
"db": {
"port": 27017,
"host": "localhost"
}
endpoint_root can have one of two values: server, database. If it is ommited, the server value is presumed. server means that we can select a database for each query, setting its name in an url, like GET /test_db/test_collection/foo_id. If instead database value is set, than connection is restricted to a single database, given in config connection options: "db": "mongodb://localhost:27017/test_db". Then all the urls should ommit db parameter. So the previous query will look like GET /test_collection/foo_id.
server specifies the configuration for the REST API server, it also defaults if not specified.
mongoOptions specifies MongoDB server and database connection parameters. These are passed directly to the MongoDB API.
Valid options under serverOptions are documented here: http://mongodb.github.io/node-mongodb-native/api-generated/server.html.
auto_reconnect is automatically enabled, don't override this or mongoapi may not work as expected.
Valid options under dbOptions are documented here: http://mongodb.github.io/node-mongodb-native/api-generated/db.html.
w (write concern) is set to 1 so that acknowledgement of the write is recieved by mongoapi, currently this must be enabled for error checking.
Set collectionOutputType to csv to returns collections as csv data rather than json.
If you are configuring the server procedurally you can assign a Javascript function to transformCollection which will transform each collection before returning it via HTTP.
The accessControl options allow you to set the following headers on the HTTP response:
- Access-Control-Allow-Origin
- Access-Control-Allow-Methods
- Access-Control-Allow-Credentials
Help for these headers can be found here: https://developer.mozilla.org/en-US/docs/Web/HTTP/Access_control_CORS
dbAccessControl can be used for limiting access only to certain databases or collections. If ommited, user can reach to any database and collection.
If endpoint_root is set to server, than the syntax for this option is as follows:
{
"database_name": ["collection_name1", "collection_name2"],
"database_name2": [],
}
This example allows access only to two databases. For database_name only two collections are accesible. For database_name2 all collections are accesible.
If endpoint_root is set to database, than the syntax is as follows:
[
"collection_name1", "collection_name2"
]
So it's just a list of accesible collections. If array is empty, all collections are accesible.
The urlPrefix option allows specification of a prefix for the REST API URLs. This defaults to an empty string, meaning no prefix which was the original behavior. For example, given the following REST API URL:
/database/collection
Setting a URL prefix of /blah will change the example REST API URL to:
/blah/database/collection
The URL prefix should allow the REST API to co-exist with another REST API and can also be used a very primitive form of security (by setting the prefix to a secret key).
schema option defines json schemas for collections. So all the documents in given collections should match defined schemas. Schema validation is performed on insert, replace and update operations. If new document does not passes schema validation, response code 400 is returned.
Winston logging is supported if you configure the REST API programmatically. When you call startServer and pass in configuration options set the logger option to your Winston logger. Mongoapi uses the following functions: verbose, info, warn and error.
Please see the Winston documentation for more setup details: https://github.com/flatiron/winston
Listing Databases:
Format: GET /dbs
$ curl 'http://127.0.0.1:3000/dbs/' \
> -D - \
> -H 'Accept: application/json'
HTTP/1.1 200 OK
X-Powered-By: Express
Content-Type: application/json; charset=utf-8
Content-Length: 27
ETag: W/"1b-134804454"
Date: Thu, 02 Jul 2015 08:02:26 GMT
Connection: keep-alive
[
"local",
"test"
]
Listing Collections:
Format:GET /<db>/
$ curl 'http://127.0.0.1:3000/test/' \
> -D - \
> -H 'Accept: application/json'
HTTP/1.1 200 OK
X-Powered-By: Express
Content-Type: application/json; charset=utf-8
Content-Length: 27
ETag: W/"1b-134804454"
Date: Thu, 02 Jul 2015 08:02:26 GMT
Connection: keep-alive
[
"new-collection",
"startup_log",
"system.indexes"
]
List Documents in a Collection:
Format: GET /<db>/<collection>
$ curl 'http://127.0.0.1:3000/test/new-collection' \
> -D - \
> -H 'Accept: application/json'
HTTP/1.1 200 OK
X-Powered-By: Express
Content-Type: application/json; charset=utf-8
Content-Length: 27
ETag: W/"1b-134804454"
Date: Thu, 02 Jul 2015 08:02:26 GMT
Connection: keep-alive
[
{
"_id": "5594bf2b019d364a083f2e03",
"attribute": "hello"
}
]
Output a CSV collection:
Format:GET /<db>/<collection>?output=csv
$ curl http://127.0.0.1:3000/test/newcollection?output=csv > Sample.csv
List documents satisfying a query:
Format:GET /<db>/<collection>?query={"key":"value"}
$ curl -X "GET" http://localhost:3000/test/newcollection \
-d 'query={"attribute":"value"}
[
{
"_id": "5594bf2b019d364a083f2e03",
"attribute": "value"
}
]
List documents with nested queries:
Format:GET /<db>/<collection>?query={"key":{"second_key":{"_id":"value"}}}
$ curl -X "GET" http://localhost:3000/test/newcollection \
-d 'query={"attribute":{"other_attribute:{"_id":"5063114bd386d8fadbd6b004"}}}
[
{
"_id": "5594bf2b019d364a083f2e03",
"attribute": {
other_attribute: "5063114bd386d8fadbd6b004"
}
}
]
Return document by id:
Format GET /<db>/<collection>/id
$ curl -X "GET" http://localhost:3000/test/nested/5594bf2b019d364a083f2e03
{
"_id": "5594bf2b019d364a083f2e03",
"attribute": "hello"
}
Inserting documents:
Format: POST /<db>/<collection>
$ curl 'http://localhost:3000/test/newcollection' \
> -D - \
> -X POST \
> -H 'Content-Type: application/json' \
> -H 'Accept: application/json' \
> --data '{"title": "Some title", "content": "document content"}'
HTTP/1.1 201 CREATED
Date: Thu, 02 Jul 2015 12:50:34 GMT
Connection: keep-alive
Content-Type: application/json; charset=utf-8
X-Powered-By: Express
Content-Length: 15
{
"_id": "5595339aa73107ad070e891a",
"title": "Some title",
"content": "document content"
}
Replacing a document:
Format: PUT /<db>/<collection>/id
$ curl -X "PUT" "http://localhost:3000/test/nested/5595339aa73107ad070e891a \
> --data {"title": "New title", "content": "New document content"}'
HTTP/1.1 200 OK
Connection: keep-alive
Content-Type: application/json; charset=utf-8
X-Powered-By: Express
Content-Length: 15
Date: Thu, 02 Jul 2015 12:53:00 GMT
{
"_id": "5595339aa73107ad070e891a",
"title": "New title",
"content": "New document content"
}
Updating a document:
Format: PATCH /<db>/<collection>/id
$ curl -X "PUT" "http://localhost:3000/test/nested/5595339aa73107ad070e891a \
> --data {"title": "New title", "content": "New document content", "field_to_delete": null}'
HTTP/1.1 200 OK
Connection: keep-alive
Content-Type: application/json; charset=utf-8
X-Powered-By: Express
Content-Length: 15
Date: Thu, 02 Jul 2015 12:53:00 GMT
{
"_id": "5595339aa73107ad070e891a",
"title": "New title",
"content": "New document content"
}
Deleting a document by id:
Format: DELETE /<db>/<collection>/id
$ curl -X "DELETE" "http://localhost:3000/test/nested/5595339aa73107ad070e891a
HTTP/1.1 200 OK
Connection: keep-alive
Content-Type: application/json; charset=utf-8
X-Powered-By: Express
Content-Length: 15
Date: Thu, 02 Jul 2015 12:53:00 GMT
{
"ok": 1
}
Bulk write (insert, update and delete)
Format: POST /<db>/bulk
$ curl 'http://localhost:3000/test/bulk' \
> -D - \
> -X POST \
> -H 'Content-Type: application/json' \
> -H 'Accept: application/json' \
> --data '{"data": {"collection1": {"insert": [{"Title": "Some title"}, {"_id": "5595339aa73107ad070e891a", "Key": "Value"}], "update": [{"_id": 123, "New field": "new value"}]}, "collection2": {"delete": [{"name": "John"}, {"_id": "5595339aa73107ad070e891b"}]}}}'
HTTP/1.1 200 OK
Date: Thu, 02 Jul 2015 12:50:34 GMT
Connection: keep-alive
Content-Type: application/json; charset=utf-8
X-Powered-By: Express
Content-Length: 15
{
"ok": 1
}
For bulk write operation the following syntax of POST body should be used:
{
"data": {
"collection1": {
"insert": [<doc1>, <doc2>, ...],
"update": [<doc3>, <doc4>, ...],
"delete": [<doc5>, <doc6>, ...],
},
"collection2": {
"insert": [<doc1>, <doc2>, ...],
"update": [<doc3>, <doc4>, ...],
"delete": [<doc5>, <doc6>, ...],
},
...
}
}
So insert, update and delete operations can be performed in a single request for multiple collections.
Documents in update section should contain an _id field, that acts as a filter. The rest fields are used in mongo $set operator to update existing document.
Documents in insert and delete section are not obligated to contain _id field.
Content Type:
Please make sure application/json is used as Content-Type when using POST/PUT with request bodies.
When performing a query GET /<db>/<collection>, some options can be applyed together with filter. The following options are supported:
- skip (int)
- limit (int)
- sort (object)
- hint (object)
- fields (object)
- snapshot (boolean)
- count (boolean)
- explain (boolean)
For explain option, the explain is performed and returned for given query, no documents are returned.
For count option the response looks like {count: 24}, no documents are returned. Limit and skip options do influence the count.
An example of query with options:
GET /<db>/<collection>?query={"key":"value"}&fields={"name":1,"surname":1}&limit=10&skip=2&snapshot=1&sort={"name":-1}&hint=index_name
WARNING: This is a prototype feature and may change in the future.
mongoapi supports a simple token-based auth system. Login is accomplilshed by a HTTP POST to /login with username and password, the server will verify the user's password against a secret database. Upon authentication an access token is returned that must be attached to each subsequent API requests.
An authorization token is specified via query parameter as follows:
GET /db/collection?token=234d43fdg-34324d-dd-dsdf-f435d
Authentication is enabled by adding auth to config.json as follows:
"auth": {
"usersDBConnection": "mongodb://localhost/auth",
"usersCollection": "users",
"tokenDBConnection": "mongodb://localhost/auth",
"tokensCollectionName": "tokens",
"universalAuthToken": "this-token-grants-universal-access-so-please-change-it",
"tokenExpirationTimeHours": 8
}
auth requires at least:
- usersDBConnection - mongodb connection string for the users database.
- tokenDBConnection - mongodb connection string for the tokens database.
Here are the docs for mongodb connection strings: http://docs.mongodb.org/manual/reference/connection-string/
The following are optional:
- usersCollection - The auth database collection where users are stored.
- tokensCollectionName - The auth database collection where tokens are stored.
- universalAuthToken - Specifies a token that can be used for universal authorization.
- tokenExpirationTimeHours - Specifies the timeout in hours before tokens must be renewed by 'login'.
An example configuration example config with auth.json is included with a working authentication setup.
** Please note that mongodb exposes all databases in the server, including your secret authentication database. Move your auth database to a different server on the same machine or ensure MongoDB authentication is setup correctly. Work will be done in the future that allows particular databases to be whitelisted/blacklisted and not exposed. **