SQLAlchemy JSON column - how to perform a contains query

Question

I have the following table in mysql(5.7.12):

class Story(db.Model):
    sections_ids = Column(JSON, nullable=False, default=[])

sections_ids is basicly a list of integers [1, 2, ...,n]. I need to get all rows where sections_ids contains X. I tried the following:

stories = session.query(Story).filter(
        X in Story.sections_ids
    ).all()

but it throws:

NotImplementedError: Operator 'contains' is not supported on this expression

Answer 1

For whoever get here, but is using PostgreSQL instead: your fields should be of the type sqlalchemy.dialects.postgresql.JSONB (and not sqlalchemy_utils.JSONType) -

Then you can use the Comparator object that is associated with the field with its contains (and others) operators.

Example:

Query(Mymodel).filter(MyModel.managers.comparator.contains(["[email protected]"]))

(note that the contained part must be a JSON fragment, not just a string)

Answer 2

Use JSON_CONTAINS(json_doc, val[, path]):

from sqlalchemy import func

# JSON_CONTAINS returns 0 or 1, not found or found. Not sure if MySQL
# likes integer values in WHERE, added == 1 just to be safe
session.query(Story).filter(func.json_contains(Story.section_ids, X) == 1).all()

As you're searching an array at the top level, you do not need to give path. Alternatively beginning from 8.0.17 you can use value MEMBER OF(json_array), but using it in SQLAlchemy is a little less ergonomic in my opinion:

from sqlalchemy import literal

# self_group forces generation of parenthesis that the syntax requires
session.query(Story).filter(literal(X).bool_op('MEMBER OF')(Story.section_ids.self_group())).all()