Setup

1. Add {:mongodb, "0.1.1"}, {:poolboy, ">= 0.0.0"}, {:mongo_ecto, git: "https://github.com/eyrmedical/mongo_ecto.git", tag: "v0.1.20-beta"} to mix.exs under deps function, add :poolboy and :mongodb in your application list.
2. Run mix deps.get && mix deps.compile.
3. Add MongoEcto.Repo into your supervision tree, on Phoenix it should look like: worker(MongoEcto.Repo, [[database: "eyr", hostname: "localhost"]]).
4. In web.exs either remove import Ecto.Changeset completly if you plan to use different database endpoints, or replace it with import Ecto.Changeset, except: [unique_constraint: 2] if you plan to use MongoEcto as single repo for all your schemas.

Usage

MongoEcto.Repo implements most of the Ecto.Repo functions with exception to transactions, mass operations (insert_all, update_all etc) and system calls defined for the Ecto.Repo Behaviour.

You can use MongoEcto.Repo in the same way you are using Ecto.Repo

The main difference is that you should use Mongo queries in form of map() instead of Ecto.Query keyword clause.

• [] = MongoEcto.Repo.all(schema)
• [] = MongoEcto.Repo.all(schema, %{ "$query": %{ "$or": [ %{account_id: account1_id}, %{account_id: account2_id} ] } })
• schema_record = MongoEcto.Repo.get(schema, id)
• schema_record = MongoEcto.Repo.one(schema, %{field: value})
• {:ok, schema_record} = MongoEcto.Repo.insert(changeset)
• schema_record = MongoEcto.Repo.insert!(changeset)
• schema_record = MongoEcto.Repo.delete!(changeset)

Model

• Add use MongoEcto.Model, :model to your Mongo models.
• Define @collection_name with the name of the related mongo collection.
• Define @primary_key {:id, :binary_id, autogenerate: true} to comply Mongo ObjectId UID.
• Use embedded_schema instead of schema to define fields.

Helper functions field_timestamp/2 and now_timestamp/0 can be used to set current timestamp, like: Schema.field_timestamp(record, :inserted_at)

Diclaimer

This module is in it's early beta, use on your own risk.