mongocapsule

Encapsulated MongoEngine.


Keywords
mongodb, mongoengine, orm, python
License
MIT
Install
pip install mongocapsule==0.1.2

Documentation

MongoCapsule

Build Status

Code Climate

Test Coverage

PyPI

PyPI

Overview

MongoCapsule is a very thin wrapper around MongoEngine built for your happiness. It encapsulates MongoEngine attributes under a single namespace and hence allows explicit declaration without context switches.

In addition to that, MongoCapsule adds pagination support to all the query results.

MongoEngine is a great ORM for using MongoDB in any Python project. However, since MongoEngine works in "contexts", using multiple databases requires trickery such as db_alias and switch_db. MongoCapsule solves this by attaching references to MongoEngine attributes to itself.

Quickstart

If you are familiar with MongoEngine, you can use MongoCapsule already! Create the database object and use it to define your document and fields.

from mongocapsule import MongoCapsule

db = MongoCapsule('test_db')

class Fruits(db.Document):
    name = db.StringField()

Refer to MongoEngine Docs for details.

Installation

To install use pip:

    pip install mongocapsule

Or clone the repo:

    git clone https://github.com/prashnts/mongocapsule.git
    python setup.py install

Additional API

MongoCapsule adds Pagination support to the MongoEngine QuerySet object. It returns 10 objects per page, however, this can be changed.

# Obtain nth Page of any arbitrary query:

query_results = Document.objects(...).sort(...)

result_page = query_results.page(2)     # Obtain second page

total_pages = query_results.page_count

# Update number of items returned per page:

db.QuerySet.set_page_limit(20)

Contributing

Code Patches, suggestions and bug reports welcome! Please use GitHub issues for the same.

Rant

I wrote this module because the examples in official MongoEngine documentation encourages using from mongoengine import * which not only pollutes the local namespace, but makes class definitions implicit. Of course, cherrypicked imports are possible, however that requires a lot of extra imports in each files.

The biggest problem, however, comes when you're using multiple databases or hosts -- in those cases, you need to use context switches or ugly meta attributes in the declaration.