SQLAlchemy - Eager loading with primary join relationship

Question

I'm trying to eager load a relationship that has a primaryjoin condition and it keeps telling me Unknown column 'tags.type' in 'on clause' . I believe it is because the table name is hard coded into the join, but the query is trying to use an AS name for the table. Full example below:

#!/usr/bin/env python
# -*- coding: utf-8 -*-

from flask import Flask
from flask_sqlalchemy import SQLAlchemy

db = SQLAlchemy()

def create_app():
    app = Flask(__name__)
    app.config['SQLALCHEMY_DATABASE_URI'] = 'mysql+pymysql://test:test@localhost/JoinTest'
    db.init_app(app)
    return app

class PostToTags(db.Model):
    __tablename__ ='post_to_tags'
    post_id = db.Column(db.Integer, db.ForeignKey('posts.id'), primary_key=True)
    tag_id = db.Column(db.Integer, db.ForeignKey('tags.id'), primary_key=True)

class Tag(db.Model):
    __tablename__ = 'tags'
    id = db.Column(db.Integer, primary_key=True)
    key = db.Column(db.Unicode(1024), nullable=False)
    value = db.Column(db.Unicode(1024))
    type = db.Column(db.Enum('post'), default=None)

class Post(db.Model):
    __tablename__ = 'posts'

    id = db.Column(db.Integer, primary_key=True)

    tags = db.relationship(
        "Tag",
        primaryjoin="and_(Post.id == PostToTags.post_id, Tag.type=='post')",
        secondary='post_to_tags',
    )

if __name__ == "__main__":
    app = create_app()
    with app.app_context():
        db.drop_all()
        db.create_all()
        tag = Tag(key='foo',value='bar', type='post')
        post = Post()
        post.tags.append(tag)
        db.session.add_all((post, tag))
        db.session.commit()

        q = db.session.query(Post).options(db.joinedload('tags'))

        print q
        print q.all()

The actual query it emits looks like this:

SELECT posts.id AS posts_id, tags_1.id AS tags_1_id, tags_1.`key` AS tags_1_key, tags_1.value AS tags_1_value, tags_1.type AS tags_1_type 
FROM posts LEFT OUTER JOIN (post_to_tags AS post_to_tags_1 INNER JOIN tags AS tags_1 ON tags_1.id = post_to_tags_1.tag_id) ON posts.id = post_to_tags_1.post_id AND tags.type = %(type_1)s

Notice that before the last ON statement it names the tag table to tags_1 , but then the last reference to the table tries to access the attribute via tags.type

Whats the best way to rethink this?

Thanks in advance!

Answer 1

You have almost the same problem as in the post you linked , but instead of simply removing the secondary join from the primary and letting SQLAlchemy figure it out you need to define the secondary instead of the primary – and let SQLA figure the primary out for you:

tags = db.relationship(
    "Tag",
    secondaryjoin="and_(Tag.id == PostToTags.tag_id, Tag.type == 'post')",
    secondary='post_to_tags',
)