feifangit / mongo_reader

MongoDB HTTP interface

Mongo Reader

A standalone flask application.

MongoDB HTTP interface. Inspired by sleepy.mongoose

Based on Flask and pymongo

Compare with sleepy.mongoose:

• Follows interfaces in sleepy.mongoose

Add

• Replica set support

Changed

• Replace pymongo.connection.Connection with pymongo.MongoClient
• Use WSGI interface provided by Flask instead of customized BaseHTTPServer.HTTPServer

Interfaces

Interface format: /<database>/<collection>/<command>

Current Supporting Commands:

_status:

return collection status read from MongoDB

request parameters

sample response

/_status

{
  "ok": 1, 
  "status": {
    "avgObjSize": 230.44444444444446, 
    "count": 504, 
    "indexSizes": {
      "_id_": 40880, 
      "serverTime_1_errorCode_1_company_1_site_1": 40880, 
      "timeCreated_1_errorCode_1_company_1_site_1": 40880
    }, 
    "lastExtentSize": 16990208, 
    "nindexes": 3, 
    "ns": "stage5.transaction_p1", 
    "numExtents": 7, 
    "ok": 1.0, 
    "paddingFactor": 1.0, 
    "size": 116144, 
    "storageSize": 33763328, 
    "systemFlags": 1, 
    "totalIndexSize": 122640, 
    "userFlags": 0
  }
}

_count:

return size of query result

request parameters

• criteria, JSON dict, e.g: {'code':1}

sample response

/count?criteria:{"errorcode":1}

{
  "count": 504, 
  "ok": 1
}

• count: total size of the result data set based on criteria
• ok: 1 for successfully quried, 0 for error

_find:

return query result

request parameters

• criteria, JSON dict, e.g: {'code':1}
• fields, JSON list, e.g: ['field1', 'field2']
• limit, a number, e.g: 50
• skip, a number, e.g: 20

sample response

/_find?criteria:{"errorcode":1}&limit=2

{
  "ok": 1, 
  "results": [
    {
      "_id": "da1fe7d3-2335-49cc-bdfb-dc92d248e1ec", 
      "company": 7, 
      "errorcode":1,
      "model": "xperia B", 
      "timecreated": {
        "$date": 1386103405000
      }
    }, 
    {
      "_id": "1a607f85-4f1f-4b54-915f-428e52244f9b", 
      "company": "5", 
      "errorcode":1,
      "model": "xperia C", 
      "timecreated": {
        "$date": 1386100597000
      }
    }
  ]
}

• count: total size of the result data set based on criteria
• ok: 1 for successfully quried, 0 for error
• results: list of documents
• size: size of result data set based on criteria, limit

About date/time in criteria:

.../mydb/mycol/_find?criteria={"serverTime":{"$gte":{"$date":1393459200000}}}

Run Application

Start from command line

Available command line parameters:

• -p, --port application listen port, 5000 by default
• -i, --ip application listen ip, 0.0.0.0 by default
• -d debug flag, will pass to app.run(debug=?), False by default
• --mongoconn mongoDB connection string, mongodb://127.0.0.1:27017/1 by default, MongoDB connection string format http://docs.mongodb.org/manual/reference/connection-string/

python app.py -p 8080

connect to a standalone MongoDB instance on 127.0.0.1:27017 will start web application and listen on port 8080

python app.py -p 8080 --mongoconn "mongodb://db1.example.net,db2.example.net:2500/?replicaSet=test&connectTimeoutMS=300000"

connect to a MongoDB replicaset test on db1.example.net:27017 and db2.example.net:2500 the web application itself is listening on port 8080

Run with Gunicorn

If you're planning to run the app in production environment, you probably need to run it with Gunicorn which provides you a better performance.

parameters -p, --port, -i, --ip, -d is omitted while running with Gunicorn. You can run

gunicorn -k gevent "app:build_app('--mongoport=27018 --mongoip=10.1.1.47')" -b 0.0.0.0:80 to start the application listen on 0.0.0.0:80, run with gevent and connect to a MongoDB instance on 10.1.1.47:27018