SQLAlchemy-JSONField

SQLALchemy JSONField implementation for storing dicts at SQL independently from JSON type support.

Why?

SqlAlchemy provides JSON field support for several database types (PostgreSQL and MySQL for now) and semi-working dict <-> JSON <-> VARCHAR example, but... In real scenarios we have tests on sqlite, production on MySQL/MariaDB/Percona/PostgreSQL and some of them (modern Oracle MySQL & PostgreSQL) support JSON, some of them (SQLite, Percona & MariaDB) requires data conversion to Text (not VARCHAR).

As addition, we have different levels of Unicode support on database and connector side, so we may be interested to switch JSON encoding between deployments.

Note

SQLite 3.9 supports JSON natively and SQLAlchemy can handle this.

Solution:

SQLALchemy JSONField has API with suport for automatic switch between native JSON and JSON encoded data, and encoding to JSON string can be enforced.

Pros:

• Free software: Apache license
• Open Source: https://github.com/penguinolog/sqlalchemy_jsonfield
• Self-documented code: docstrings with types in comments
• Uses native JSON by default, but allows to specify different library.
• Support multiple Python versions

Usage

Direct usage with MariaDB (example extracted from functional tests):

import sqlalchemy_jsonfield

class ExampleTable(Base):
    __tablename__ = table_name
    id = sqlalchemy.Column(sqlalchemy.Integer, primary_key=True)
    row_name = sqlalchemy.Column(
        sqlalchemy.Unicode(64),
        unique=True,
    )
    json_record = sqlalchemy.Column(
        sqlalchemy_jsonfield.JSONField(
            # MariaDB does not support JSON for now
            enforce_string=True,
            # MariaDB connector requires additional parameters for correct UTF-8
            enforce_unicode=False
        ),
        nullable=False
    )

Usage with alternate JSON library:

import sqlalchemy_jsonfield
import ujson

class ExampleTable(Base):
    __tablename__ = table_name
    id = sqlalchemy.Column(sqlalchemy.Integer, primary_key=True)
    row_name = sqlalchemy.Column(
        sqlalchemy.Unicode(64),
        unique=True,
    )
    json_record = sqlalchemy.Column(
        sqlalchemy_jsonfield.JSONField(
            enforce_string=True,
            enforce_unicode=False,
            json=ujson,  # Use ujson instead of standard json.
        ),
        nullable=False
    )

Usage on PostgreSQL/Oracle MySQL(modern version)/SQLite(testing) environments allows to set enforce_string=False and use native JSON fields.

Testing

The main test mechanism for the package sqlalchemy_jsonfield is using tox. Available environments can be collected via tox -l

CI systems

For code checking several CI systems is used in parallel:

1. GitHub actions: is used for checking: PEP8, pylint, bandit, installation possibility and unit tests.