nomilkinmyhome/elasticmapper

Easy generating ElasticSearch mappings based on ORM's models.

ElasticMapper

ElasticSearch mapper for three ORMs - SQLAlchemy, Peewee, DjangoORM.
Easy generating ElasticSearch mappings based on models.

Installation

pip install elasticmapper

Basic usage

Peewee example

from peewee import *
from elasticmapper import PeeweeMapper

db = SqliteDatabase('my_app.db')

class BaseModel(Model):
    class Meta:
        database = db

class User(BaseModel):
    username = CharField(unique=True)
    is_active = BooleanField(default=True)
    age = IntegerField()

user_elastic_mapping = PeeweeMapper(model=User).load()

SQLAlchemy example

from sqlalchemy.orm import declarative_base
from sqlalchemy import Column, Integer, String, Boolean
from elasticmapper import SQLAlchemyMapper

Base = declarative_base()

class User(Base):
    __tablename__ = 'users'
    id = Column(Integer, primary_key=True)
    username = Column(String)
    is_active = Column(Boolean)
    age = Column(Integer)

user_elastic_mapping = SQLAlchemyMapper(model=User).load()

DjangoORM example

from django.db import models
from elasticmapper import DjangoMapper

class User(models.Model):
    username = models.CharField(max_length=30)
    is_active = models.BooleanField(default=True)
    age = models.IntegerField()

user_elastic_mapping = DjangoMapper(model=User).load()

Output for all examples:

{
    'id': {'type': 'integer'},
    'username': {'type': 'text'},
    'age': {'type': 'integer'},
    'is_active': {'type': 'boolean'}
}

Documentation

Documentation lives here.

Contributing

PR are welcome! If you want to help, please visit the contribution guide, then take one of the issues. Thank you for your contribution.

License

Copyright © 2022 Polina Beskorovaynaya ihatemilk

This project has MIT License.

elasticmapper's Issues

improve tests

add pytest.mark.parametrize for each test with different cases

m2m support

uuid pk support

add poetry

move docs from readme

self joins support

add custom values

new parameter for load() - custom_values.
it will be a dict that contains custom values per field.
for example:

from django.db import models

from elasticmapper import load, SupportedORMs

class User(models.Model):
    username = models.CharField(max_length=30)
    timestamp = models.DateTimeField()

custom_values = {'timestamp': {'type': 'date', 'format': `'yyyy-MM-dd'}}
user_elastic_mapping = load(
    model=User, 
    orm=SupportedORMs.DjangoORM,
    custom_values=custom_values,
)

expected output:

{
    'id': 'int',
    'username': 'text',
    'timestamp': {
        'type': 'date',
        'format': 'yyyy-MM-dd'
    }
}

configure CI (travis.ci)

add support for nested types and FK for each ORM

nested types like arrays or others

replace sphinx with mddocs

more orms support

TortoiseORM, PonyORM, etc

support more python versions

many models support

add tests and linters

install pytest and flake8.
tests should cover common use cases: one test for each load() parameter and tests for each supported orm.