I have a problem with SQL Alchemy, while trying to think about an SQL schema I encountered the following problem.
My schema is based on 2 classes, Flight and Trip.
A Trip includes 2 fields: flights_to and flights_from.
Any of the fields is basically a list of flights, it could be made of one flight, or many flights (Connection flights).
class Trip(Base):
__tablename__ = "Trip"
__table_args__ = {'sqlite_autoincrement': True}
id = Column(Integer, primary_key = True)
flights_to = relationship("Flight", backref="Trip")
flights_from = relationship("Flight", backref="Trip")
class Flight(Base):
__tablename__ = "Flight"
__table_args__ = {'sqlite_autoincrement': True}
id = Column(Integer, primary_key = True)
arrival_airport = Column(String(20))
departure_airport = Column(String(20))
flight_number = Column(Integer)
trip_id = Column(Integer, ForeignKey('Trip.id'))
The problem happens when I create 2 fields in the same type:
sqlalchemy.exc.ArgumentError: Error creating backref 'Trip' on relationship 'Trip.flights_from': property of that name exists on mapper 'Mapper|Flight|Flight'
I have thought about using 2 inheriting classes of types FlightTo and FlightFrom and saving them at two different tables, but what if I want to use a FlightFrom as a FlightTo? will the flight be duplicated in 2 tables?
I would appreciate your help.
backref is used to define a new property on the other class you are using relationship with. So you can't have two property which have the same name
You should rename your backref for the flights_from to any other name than Trip.
It will work then.
For Example:
class Person(Model):
id = Column(Integer, primary_key=True)
name = Column(String)
address = relationship("Address",backref="address")
class Address(Model):
id = Column(Integer, primary_key=True)
house_no = Column(Integer)
person_id = Column(Integer, ForeignKey('person.id'))
So you can access the person name with house_no 100 by:
query_address = Address.query.filter_by(house_no=100).first()
person = query_address.address
This returns you the person object.
Thus if you have multiple such names , it will give you an error
Related
I am struggling with "many to many" relationships in Sqlalchemy.
My data model:
a chain consists of one or multiple jobs
a job may be part of one or multiple chains
class Chain(Base):
__tablename__ = "chain"
__table_args__ = {'extend_existing': True}
__guarded__ = ['id']
id = Column(Integer, primary_key=True)
name = Column(Text, nullable=False)
# .. some other fields
class JobConfig(Base):
__tablename__ = "jobconfiguration"
__table_args__ = {'extend_existing': True}
__guarded__ = ['id']
chain = relationship("Chain", secondary='chain_job')
id = Column(Integer, primary_key=True)
jobname = Column(String(250).evaluates_none(), nullable=False, unique=True)
# some other fields
class ChainJob(Base):
__tablename__ = "chain_job"
__table_args__ = {'extend_existing': True}
__guarded__ = ['id']
id = Column(Integer, primary_key=True)
chain_id = Column(Integer, ForeignKey('chain.id'))
job_id = Column(Integer, ForeignKey('jobconfiguration.id'))
Now when I try to create a new chain and append a new job I get an error message:
chain = Chain()
chain.name='MyFirstChain'
job = JobConfig()
job.jobname="PartOfChain1"
job.chain.append(chain)
session.add(job)
session.add(chain)
session.commit()
sqlalchemy.orm.exc.UnmappedColumnError: Can't execute sync rule for
source column 'jobconfiguration.id'; mapper 'mapped class
JobConfig->jobconfiguration' does not map this column. Try using an
explicit foreign_keys collection which does not include destination
column 'chain_job.job_id' (or use a viewonly=True relation).
How can I fix this?
Let's assume we have the following code in some Models.py file:
class Person(db.Model):
__tablename__ = 'Persons'
ID = db.Column(db.Integer, primary_key=True, nullable=False)
Name = db.Column(db.String(255), nullable=False)
class House(db.Model):
__tablename__ = 'Houses'
ID = db.Column(db.Integer,primary_key=True,nullable=False)
OwnerID = db.Column(db.Integer, nullable=False)
TenantID = db.Column(db.Integer, nullable=False)
__table_args__ = (
db.ForeignKeyConstraint(
['OwnerID'],
['Persons.ID'],
),
db.ForeignKeyConstraint(
['TenantID'],
['Persons.ID'],
),
)
OwnerBackref = db.relationship('Person', backref='OwnerBackref', lazy=True, foreign_keys=[OwnerID])
TenantBackref = db.relationship('Person', backref='TenantBackref', lazy=True, foreign_keys=[TenantID])
And we want to reflect these models using the automap base, so we have this code in another module called Database.py:
Base = automap_base()
engine = create_engine(DB_CONNECTION, pool_size=10, max_overflow=20)
db_session = scoped_session(sessionmaker(autocommit=False, autoflush=False, bind=engine))
Base.prepare(engine, reflect=True)
Person = Base.classes.Persons
House = Base.classes.Houses
Now, when I import House in some other module I want to be able to do this:
h = db_session.query(House).first()
print(h.OwnerBackref.Name)
print(h.TenantBackref.Name)
But instead I get an error saying that those 2 backrefs do not exist and instead a field called 'persons' gets added to my House object but the problem here is that it links only 1 (either the Tenant either the Owner). By this I mean that if I do this:
print(h.persons.Name)
It will only print the Name either for the respective Tenant either for the Owner leaving me with no way of accessing the informations for the other one. (Note here that the names that I set to the backrefs are nowhere to be found)
So, my question is how can I use the backrefs I created to access my desired informations ? Am I doing something wrong here ?
The error in your code is that you are using foreign_keys= to define the relationship between the tables but you are passing the local key name not the foreign key name to the function. For your code you cannot use foreign_keys= to define the relationship within the House model as there is only one possible foreign key Person.ID but two possible local keys House.OwnerID and House.TenantID. The primaryjoin= argument should be used instead to specify this.
class Person(db.Model):
__tablename__ = 'Persons'
ID = db.Column(db.Integer, primary_key=True)
Name = db.Column(db.String(255), nullable=False)
class House(db.Model):
__tablename__ = 'Houses'
ID = db.Column(db.Integer,primary_key=True)
OwnerID = db.Column(db.Integer, db.ForeignKey('Persons.ID'), nullable=False)
TenantID = db.Column(db.Integer, db.ForeignKey('Persons.ID'), nullable=False)
Owner = db.relationship('Person', backref='HousesOwned', primaryjoin='House.OwnerID == Person.ID')
Tenant = db.relationship('Person', backref='HousesOccupied', primaryjoin='House.TenantID == Person.ID')
If you placed the relationship statements in in the Person model rather than the House model then you could use either foreign_keys= or primaryjoin= to define the relationship. The following code will result in exactly the same relationships as in the previous code.
class Person(db.Model):
__tablename__ = 'Persons'
ID = db.Column(db.Integer, primary_key=True)
Name = db.Column(db.String(255), nullable=False)
HousesOwned = db.relationship('House', backref='Owner', foreign_keys='[House.OwnerID]')
HousesOccupied = db.relationship('House', backref='Tenant', foreign_keys='[House.TenantID]')
class House(db.Model):
__tablename__ = 'Houses'
ID = db.Column(db.Integer,primary_key=True)
OwnerID = db.Column(db.Integer, db.ForeignKey('Persons.ID'), nullable=False)
TenantID = db.Column(db.Integer, db.ForeignKey('Persons.ID'), nullable=False)
I'm a little confused over the use of the two modules from SQLAlchemy. This is the code I have:
Base = declarative_base()
class Restaurant(Base):
__tablename__ = 'restaurant'
id = Column(Integer, primary_key=True)
name = Column(String(250), nullable=False)
class MenuItem(Base):
__tablename__ = 'menu_item'
name =Column(String(80), nullable = False)
id = Column(Integer, primary_key = True)
description = Column(String(250))
price = Column(String(8))
course = Column(String(250))
restaurant_id = Column(Integer,ForeignKey('restaurant.id'))
restaurant = relationship(Restaurant)
I understand that ForeignKey is used to define the foreign key relationship between the restaurant_id column of menu_item table and the id column of restaurant table. But why then is restaurant = relationship(Restaurant) used?
restaurant_id will refer to an id (the column value). restaurant will refer to a Restaurant instance that will be lazy loaded from the db on access (or eager loaded if you set up the right stuff earlier). If you set backref on the relationship you can also access a list of MenuItem objects from a Restaurant.
SQLAlchemy newbie here.
I'm trying to define a model subclass that represents a subset of table data. Specifically, I want the subclass to map the most recent row for a given ID.
For example, suppose I have the following model:
class AddressHistory(Base):
__table__ = 'address_table'
date = Column(Date, index=True, nullable=False)
id = Column(BigInteger, primary_key=True)
street = Column(String(2000))
city = Column(String(2000))
state = Column(String(2000))
zip = Column(Integer)
What I want to do is define a subclass of this model which represents the most recent address record for a given id:
class MostRecentAddress(Address):
“””
Represents a row in AddressHistory with the most recent date for a given id.
”””
Is there some sort of subquery I can pass to the mapper_args ? Or is there a way to define the table as a select statement?
You're looking for single table inheritance.
https://docs.sqlalchemy.org/en/13/orm/inheritance.html#single-table-inheritance
Your code sample is very nearly exactly how to go about doing this. You just need to add the mapper.
class Person(Base):
__tablename__ = 'people'
id = Column(Integer, primary_key=True)
discriminator = Column('type', String(50))
__mapper_args__ = {'polymorphic_on': discriminator}
class Engineer(Person):
__mapper_args__ = {'polymorphic_identity': 'engineer'}
primary_language = Column(String(50))
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')