I have read the SQLAlchemy documentation and tutorial about building many-to-many relation but I could not figure out how to do it properly when the association table contains more than the 2 foreign keys.
I have a table of items and every item has many details. Details can be the same on many items, so there is a many-to-many relation between items and details
I have the following:
class Item(Base):
__tablename__ = 'Item'
id = Column(Integer, primary_key=True)
name = Column(String(255))
description = Column(Text)
class Detail(Base):
__tablename__ = 'Detail'
id = Column(Integer, primary_key=True)
name = Column(String)
value = Column(String)
My association table is (It's defined before the other 2 in the code):
class ItemDetail(Base):
__tablename__ = 'ItemDetail'
id = Column(Integer, primary_key=True)
itemId = Column(Integer, ForeignKey('Item.id'))
detailId = Column(Integer, ForeignKey('Detail.id'))
endDate = Column(Date)
In the documentation, it's said that I need to use the "association object". I could not figure out how to use it properly, since it's mixed declarative with mapper forms and the examples seem not to be complete. I added the line:
details = relation(ItemDetail)
as a member of Item class and the line:
itemDetail = relation('Detail')
as a member of the association table, as described in the documentation.
when I do item = session.query(Item).first(), the item.details is not a list of Detail objects, but a list of ItemDetail objects.
How can I get details properly in Item objects, i.e., item.details should be a list of Detail objects?
From the comments I see you've found the answer. But the SQLAlchemy documentation is quite overwhelming for a 'new user' and I was struggling with the same question. So for future reference:
ItemDetail = Table('ItemDetail',
Column('id', Integer, primary_key=True),
Column('itemId', Integer, ForeignKey('Item.id')),
Column('detailId', Integer, ForeignKey('Detail.id')),
Column('endDate', Date))
class Item(Base):
__tablename__ = 'Item'
id = Column(Integer, primary_key=True)
name = Column(String(255))
description = Column(Text)
details = relationship('Detail', secondary=ItemDetail, backref='Item')
class Detail(Base):
__tablename__ = 'Detail'
id = Column(Integer, primary_key=True)
name = Column(String)
value = Column(String)
items = relationship('Item', secondary=ItemDetail, backref='Detail')
Like Miguel, I'm also using a Declarative approach for my junction table. However, I kept running into errors like
sqlalchemy.exc.ArgumentError: secondary argument <class 'main.ProjectUser'> passed to to relationship() User.projects must be a Table object or other FROM clause; can't send a mapped class directly as rows in 'secondary' are persisted independently of a class that is mapped to that same table.
With some fiddling, I was able to come up with the following. (Note my classes are different than OP's but the concept is the same.)
Example
Here's a full working example
from sqlalchemy import create_engine, Column, Integer, String, ForeignKey
from sqlalchemy.orm import declarative_base, relationship, Session
# Make the engine
engine = create_engine("sqlite+pysqlite:///:memory:", future=True, echo=False)
# Make the DeclarativeMeta
Base = declarative_base()
class User(Base):
__tablename__ = "users"
id = Column(Integer, primary_key=True)
name = Column(String)
projects = relationship('Project', secondary='project_users', back_populates='users')
class Project(Base):
__tablename__ = "projects"
id = Column(Integer, primary_key=True)
name = Column(String)
users = relationship('User', secondary='project_users', back_populates='projects')
class ProjectUser(Base):
__tablename__ = "project_users"
id = Column(Integer, primary_key=True)
notes = Column(String, nullable=True)
user_id = Column(Integer, ForeignKey('users.id'))
project_id = Column(Integer, ForeignKey('projects.id'))
# Create the tables in the database
Base.metadata.create_all(engine)
# Test it
with Session(bind=engine) as session:
# add users
usr1 = User(name="bob")
session.add(usr1)
usr2 = User(name="alice")
session.add(usr2)
session.commit()
# add projects
prj1 = Project(name="Project 1")
session.add(prj1)
prj2 = Project(name="Project 2")
session.add(prj2)
session.commit()
# map users to projects
prj1.users = [usr1, usr2]
prj2.users = [usr2]
session.commit()
with Session(bind=engine) as session:
print(session.query(User).where(User.id == 1).one().projects)
print(session.query(Project).where(Project.id == 1).one().users)
Notes
reference the table name in the secondary argument like secondary='project_users' as opposed to secondary=ProjectUser
use back_populates instead of backref
I made a detailed writeup about this here.
Previous Answer worked for me, but I used a Class base approach for the table ItemDetail. This is the Sample code:
class ItemDetail(Base):
__tablename__ = 'ItemDetail'
id = Column(Integer, primary_key=True, index=True)
itemId = Column(Integer, ForeignKey('Item.id'))
detailId = Column(Integer, ForeignKey('Detail.id'))
endDate = Column(Date)
class Item(Base):
__tablename__ = 'Item'
id = Column(Integer, primary_key=True)
name = Column(String(255))
description = Column(Text)
details = relationship('Detail', secondary=ItemDetail.__table__, backref='Item')
class Detail(Base):
__tablename__ = 'Detail'
id = Column(Integer, primary_key=True)
name = Column(String)
value = Column(String)
items = relationship('Item', secondary=ItemDetail.__table__, backref='Detail')
Related
I try to set fk which parent_id contains id of a person in People table in orm manner and backpopulate between them but it does not work.
class People(Base):
__tablename__ = "people"
id = Column(Integer, primary_key=True)
name = Column(String(20), nullable=False, unique=True)
parent_id = Column(Integer, ForeignKey('people.id'))
parent = relationship("People", back_populates="parent", uselist=False)
engine = create_engine(
f'mssql://{username}:{password}#{server_name}/{db_name}?driver=SQL+Server&trusted_connection=yes')
Session = sessionmaker(bind=engine)
session = Session()
session.add(People(name='me'))
raise sa_exc.ArgumentError(
sqlalchemy.exc.ArgumentError: People.parent and back-reference People.parent are both of the same direction symbol('ONETOMANY'). Did you mean to set remote_side on the many-to-one side ?
You can use the remote_side argument.
Here's code I'm using adapted to your example:
class People(Base):
__tablename__ = "people"
id = Column(Integer, primary_key=True)
parent_id = Column(Integer, ForeignKey('people.id'))
parent = relationship('People', foreign_keys=parent_id, remote_side=id)
children = relationship('People', back_populates='parent')
Situation
I have the Self-Referential Many-to-Many Relationship (almost identical to the sqlalchemy manual entry of the same heading). This relationship is governed by the table entity_weights. This code works!
Question
How do I include an attribute in class Entity representing the table column entity_weights.weight. Lets say the attribute would be called Entity.child_weights. It is important that the rank of Entity.child_entities and Entity.child_weights are identical.
entity_weights = Table('entity_weights', Base.metadata,
Column('id',Integer, primary_key=True),
Column('parent_entity_id',Integer, ForeignKey('entity.id')),
Column('entity_id',Integer, ForeignKey('entity.id')),
Column('weight',Float))
class Entity(Base):
__tablename__ = 'entity'
id = Column(Integer, primary_key=True)
name = Column(String)
domicile_id = Column(Integer, ForeignKey('domicile.id'))
entity_type = Column('type',Enum('asset','institution','model'))
source_table_id = Column(Integer)
child_entities = relationship('Entity',
secondary=entity_weights,
primaryjoin=id==entity_weights.c.parent_entity_id,
secondaryjoin=id==entity_weights.c.entity_id,
backref='parent_entity_id'
)
The cleanest solution I've found for this scenario is to break up the child_entities relationship by adding entity_weights as a one-to-many relationship on Entity and use an association proxy to proxy the weight value as well as the remote side of the many-to-many relationship:
class EntityWeight(Base):
__tablename__ = "entity_weights"
id = Column(Integer, primary_key=True)
parent_entity_id = Column(Integer, ForeignKey('entity.id'))
entity_id = Column(Integer, ForeignKey('entity.id'))
weight = Column(Float)
entity = relationship("Entity", primaryjoin=lambda: EntityWeight.entity_id == Entity.id)
class Entity(Base):
...
_child_weights = relationship(EntityWeight, primaryjoin=id == EntityWeight.parent_entity_id)
child_weights = association_proxy("_child_weights", "weight")
child_entities = association_proxy("_child_weights", "entity")
This question already has an answer here:
The foreign key associated with column 'x.y' could not ... generate a foreign key to target column 'None'
(1 answer)
Closed 4 years ago.
I've got three tables: users, funds, and fund types. Each fund has a fund type, each user has a list of funds, and each user can also have a list of fund types that they have created.
Schema:
class Fund(Base):
__tablename__ = 'fds_funds'
fds_fund_id = Column(Integer, primary_key=True)
fds_name = Column(String(128))
fds_symbol = Column(String(5))
fds_fdt_fund_type_id = Column(Integer, ForeignKey('fdt_fund_type_id'))
fund_type = relationship('FundType', backref=backref('fds_funds', uselist=False))
class FundType(Base):
__tablename__ = 'fdt_fund_types'
fdt_fund_type_id = Column(Integer, primary_key=True)
fdt_type_name = Column(String(128))
fdt_usr_user_id = Column(Integer, ForeignKey('usr_user_id'), nullable=True)
user_funds = Table('usf_user_funds', Base.metadata,
Column('usf_usr_user_id', Integer, ForeignKey('usr_users.usr_user_id')),
Column('usf_fds_fund_id', Integer, ForeignKey('fds_funds.fds_fund_id'))
)
class User(Base):
"""
Application's user model.
"""
__tablename__ = 'usr_users'
usr_user_id = Column(Integer, primary_key=True)
usr_email = Column(Unicode(50))
_usr_password = Column('password', Unicode(64))
fund_types = relationship('FundType', foreign_keys='FundType.fdt_usr_user_id')
funds = relationship('Fund', secondary=user_funds)
I'm following the documentation here, and it appears to very clearly say that the first argument to the foreign key designation should be the column name, not the table name, but I'm getting this error when I run the initialize_DB script:
sqlalchemy.exc.NoReferencedTableError: Foreign key associated with column 'fdt_fund_types.fdt_usr_user_id' could not find table
'usr_user_id' with which to generate a foreign key to target column
'None'
Have I misread the documentation?
I'm not really sure what fixed it, but I did some rearranging and messing with the references. For posterity, this works:
user_funds = Table('usf_user_funds', Base.metadata,
Column('usf_usr_user_id', Integer, ForeignKey('usr_users.usr_user_id')),
Column('usf_fds_fund_id', Integer, ForeignKey('fds_funds.fds_fund_id'))
)
class User(Base):
"""
Application's user model.
"""
__tablename__ = 'usr_users'
usr_user_id = Column(Integer, primary_key=True)
usr_email = Column(Unicode(50))
_usr_password = Column('password', Unicode(64))
fund_types = relationship('FundType', foreign_keys='FundType.fdt_usr_user_id')
funds = relationship('Fund', secondary=user_funds)
class FundType(Base):
__tablename__ = 'fdt_fund_types'
fdt_fund_type_id = Column(Integer, primary_key=True)
fdt_type_name = Column(String(128))
fdt_usr_user_id = Column(Integer, ForeignKey('usr_users.usr_user_id'), nullable=True)
class Fund(Base):
__tablename__ = 'fds_funds'
fds_fund_id = Column(Integer, primary_key=True)
fds_name = Column(String(128))
fds_symbol = Column(String(5))
fds_fdt_fund_type_id = Column(Integer, ForeignKey('fdt_fund_types.fdt_fund_type_id'))
fund_type = relationship('FundType', backref=backref('funds', uselist=False))
I have two models:
class Report(Base):
__tablename__ = 'report'
id = Column(Integer, primary_key=True)
class ReportPhoto(Base):
__tablename__ = 'report_photo'
id = Column(Integer, primary_key=True)
report_id = Column(Integer, ForeignKey(Report.id), nullable=False)
report = relationship(Report, uselist=False, backref=backref('report_photo', uselist=True))
And I would like to add column to Report model which indicates is there any records within ReportPhoto. I try to use column_property this way:
class Report(Base):
__tablename__ = 'report'
id = Column(Integer, primary_key=True)
has_photo = column_property(
select(ReportPhoto.any())
)
but get an error NameError: name 'ReportPhoto' is not defined. How I can fix this issue?
something like that should work:
class ReportPhoto(Base):
__tablename__ = 'report_photo'
id = Column(Integer, primary_key=True)
report_id = Column(Integer, ForeignKey('report.id'), nullable=False)
class Report(Base):
__tablename__ = 'report'
id = Column(Integer, primary_key=True)
report_photos = relationship(ReportPhoto, backref='report')
has_photo = column_property(
exists().where(ReportPhoto.report_id==id)
)
I will add to #Vladimir lliev's response with some clarification for anyone else that might not see how to do this.
Place the table that will have the 'foreign table referencing' column_property after that which it references. In this case, it means placing Report after ReportPhoto. This will solve your NameError, however, you would be left with a new error on your ReportPhoto foreign key reference. To solve this, place your foreign key table reference in quotes. You can read more by referencing the declarative documentation (e.g., declarative.py) and looking under "Configuring Relationships" --- specifically, read the portion on quoting your foreign references.
With your code, this would look like:
class ReportPhoto(Base):
# This now goes first
__tablename__ = 'report_photo'
id = Column(Integer, primary_key=True)
# Notice the quotations around Report references here
report_id = Column(Integer, ForeignKey("Report.id"), nullable=False)
# Notice the quotations around Report references here
report = relationship("Report",
uselist=False,
backref=backref("report_photo", uselist=True))
class Report(Base):
# This is now _after_ ReportPhoto
__tablename__ = 'report'
id = Column(Integer, primary_key=True)
# ReportPhoto now exists and we will not trip a NameError exception
has_photo = column_property(
select(ReportPhoto.any())
)
I created a many to many relationship with sqlalchemy like this:
subject_books = Table('subject_books', Base.metadata,
Column('subject_id', Integer, ForeignKey('subjects.id')),
Column('book_id', Integer, ForeignKey('books.id')),
Column('group', Integer)
)
class Subject(Base):
__tablename__ = 'subjects'
id = Column(Integer, primary_key=True)
value = Column(Unicode(255), unique=True)
class Book(Base):
__tablename__ = 'books'
id = Column(Integer, primary_key=True)
title = Column(Unicode(255))
isbn = Column(Unicode(24))
subjects = relationship('Subject', secondary=subject_books, collection_class=attribute_mapped_collection('group'), backref='books')
after that I created a test like following:
book = Book(title='first book',isbn='test')
book.subjects[0] = Subject(value='first subject')
book.subjects[1] = Subject(value='second subject')
session.add(book)
transaction.commit()
and it works fine. But what I really want is to store more than one subject with the same group value, so I tried the following test:
book = Book(title='first book',isbn='test')
book.subjects[0] = [Subject(value='first subject'),Subject(value='second subject')]
book.subjects[1] = [Subject(value='third subject'),Subject(value='forth subject')]
session.add(book)
transaction.commit()
but it does not work.
Can this be done using sqlalchemy?
Thanks in Advance
Razi
I think you are constructing wrong relation ship.
Your relation ship must be
book M2M subject
subject M2M group
So you have to create one more model for group and that must be assign as m2m in Subject
Your models will be like.
subject_books = Table('subject_books', Base.metadata,
Column('subject_id', Integer, ForeignKey('subjects.id')),
Column('book_id', Integer, ForeignKey('books.id')),
)
subject_group = Table('subject_groups', Base.metadata,
Column('group_id', Integer, ForeignKey('groups.id')),
Column('subject_id', Integer, ForeignKey('subjects.id')),
)
class Subject(Base):
__tablename__ = 'subjects'
id = Column(Integer, primary_key=True)
value = Column(Unicode(255), unique=True)
groups = relationship('Groups', secondary=subject_groups, backref='subjects')
class Groups(Base):
__tablename__ = 'groups'
id = Column(Integer, primary_key=True)
name = Column(Unicode(255), unique=True)
class Book(Base):
__tablename__ = 'books'
id = Column(Integer, primary_key=True)
title = Column(Unicode(255))
isbn = Column(Unicode(24))
subjects = relationship('Subject', secondary=subject_books, backref='books')
I also check the docs for attribute_mapped_collection. But each time I found that each key is associated with only one object not more then one. If you read anywhere then please provide the link so I can check that how it will be fit in your code.
I think this will be help you.