Difference and usage of UniqueConstraint in python class

I am using python 2.7 and sqlalchemy 0.9.9. In our code, I found various ways to define UniqueConstraint and I don't know which one is correct or if we need it at all (we don't want to create a table with our code).

Example 1:

class ExampleOne(OrmModelBase, TableModelBase):

__tablename__ = 'example_one'

example_one_id = Column(Integer, Sequence('seq_example_one_example_one_id'), 
                        primary_key=True, nullable=False)

example_one_name = Column(String, unique=True, nullable=False)

def __init__(self):
    self.example_one_id = self.getNextPkValue()
    self.example_one_name = ''

      

        
        
        
      

    

Example 2 (Version 1):

class ExampleTwo(OrmModelBase, TableModelBase):

__tablename__ = 'example_two'
example_two_id = Column(Integer, Sequence(
    'seq_example_two_example_two_id'), primary_key=True,
    nullable=False)

example_two_name = Column(String, nullable=False)

UniqueConstraint(
    "example_two_id", "example_two_name",
    name="uk_example_two_example_two_id_example_two_name")

def __init__(self):
    self.example_two_id = self.getNextPkValue()
    self.example_two_name = example_two_name

      

        
        
        
      

    

Example 2 (version 2):

class ExampleTwo(OrmModelBase, TableModelBase):

__tablename__ = 'example_two'
example_two_id = Column(Integer, Sequence(
    'seq_example_two_example_two_id'), primary_key=True,
    nullable=False)

example_two_name = Column(String, nullable=False)

uk_example_two_example_two_id_example_two_name = UniqueConstraint(
"example_two_id", "example_two_name")

def __init__(self):
    self.example_two_id = self.getNextPkValue()
    self.example_two_name = example_two_name

      

        
        
        
      

    

Example 3:

class ExampleThree(OrmModelBase, TableModelBase):
__tablename__ = 'example_three'

example_three_id = Column(Integer,
                          Sequence('example_three_example_three_id'),
                          primary_key=True)

example_three_name = Column(String, nullable=False)

__table_args__ = (
    UniqueConstraint("example_three_name", name="uk_example_three_name"),
)

def __init__(self):
    self.example_three_id = self.getNextPkValue()
    self.example_three_name = ""

      

        
        
        
      

    

I don't want to create a table with this code, I create the required tables manually with the necessary constraints.

• When I am not creating tables, do I need unique constraint definitions (or primary_key, nullable, etc.)? Is there an effect when I write one of the three versions in a class?

I read that the first example is for unique constraints for one column and the second example is for unique constraints for more than one. But the examples I found always used a table and only one example was used in my case (classes).

2. Is there any difference between using unique constraints and what is the difference?

I hope there is someone who can help me. Thank you for your time!