How to define metaclass for a class that extends from sqlalchemy declarative base ?

[Ven Karri]

I use: Python 2.6 and sqlalchemy 0.6.1

This is what I am trying to do:

from sqlalchemy.types import (

   Integer,

   String,

   Boolean

)

from sqlalchemy.ext.declarative import declarative_base

    Base = declarative_base()

class SampleMeta(type):

        def __new__(cls, name, bases, attrs):

            attrs.update({   'id': Column('Id', Integer, primary_key=True),

                        'name': Column('Name', String),

                        'description': Column('Description', String),

                        'is_active': Column('IsActive', Boolean)

                    })

            return super(SampleMeta, cls).__new__(cls, name, bases, attrs)

    class Sample(Base):

        __tablename__ = 'Sample'

        __table_args__ = {'useexisting': True}

        __metaclass__ = SampleMeta

        def __init__(self, id, name, description, is_active):

            self.id = id

            self.name = name

            self.description = description

            self.is_active = is_active

        def __repr__(self):

            return "<(%d, '%s', '%s', %r)>" % (self.id, self.name, self.description, self.isactive)

And the error I am getting is this:

    TypeError: Error when calling the metaclass bases

        metaclass conflict: the metaclass of a derived class must be a (non-strict) subclass of the metaclasses of all its bases

Now, if I do the same thing above by using

    class Sample(object)

instead of

    class Sample(Base)

it works absolutely fine.

I need to update the attributes of the class dynamically. So, I will be using dynamic attribute and column names. And I need the above piece code to work in order to be able to get there.

**Please help** 

[Michael Bayer]

your metaclass must derive from the DeclarativeMeta class.

Also, I disagree that you need this metaclass,  what you're trying to do is very easy using mixins, which are supported in version 0.6: http://docs.sqlalchemy.org/en/rel_0_6/orm/extensions/declarative.html#mixing-in-columns

--
You received this message because you are subscribed to the Google Groups "sqlalchemy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sqlalchemy+...@googlegroups.com.
To post to this group, send email to sqlal...@googlegroups.com.
Visit this group at http://groups.google.com/group/sqlalchemy.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Praveen]

The problem with using Mixins is that you need to know the definition of columns already for creating the mixin class. What I am trying to do is more like get the definition dynamically on the fly.Take a look at this:

def get_properties(tablename, map):

    table_inspector = reflection.Inspector.from_engine(DB_ENGINE.connect())

    table = Table(tablename, metadata)

    table_inspector.reflecttable(table, None)

    columns = []

    for child in table.get_children():

        if isinstance(child, Column):

            column = list(child.base_columns)[0]

            column.table = None

            columns.append(column)

    return dict([(map[column.key],  column) for column in columns])

class CustomDeclarativeMeta(DeclarativeMeta):

    def __new__(cls, name, bases, attrs):

        attrs.update(get_properties(attrs.get('__tablename__'), attrs.get('__map__')))

        return super(CustomDeclarativeMeta, cls).__new__(cls, name, bases, attrs)

# Base = declarative_base(metaclass=CustomDeclarativeMeta)

Base = declarative_base()

class Enum_SampleBase):

    __tablename__ = 'Enum_Sample'

    __table_args__ = {'useexisting': True}

    __metaclass__ = CustomDeclarativeMeta

    __map__ = {'Id': 'id', 'Name': 'name', 'Description': 'description', 'IsActive': 'is_active'}

    def __init__(self, id, name, description, is_active):

        self.id = id

        self.name = name

        self.description = description

        self.is_active = is_active

    def __repr__(self):

        return "<(%d, '%s', '%s', %r)>" % (self.id, self.name, self.description, self.isactive)

Unfortunately, this isn't working. I want to declare column types by getting them from a table that's already created in the database.

--
You received this message because you are subscribed to a topic in the Google Groups "sqlalchemy" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/sqlalchemy/37M-1Qf8HO8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to sqlalchemy+...@googlegroups.com.

To post to this group, send email to sqlal...@googlegroups.com.
Visit this group at http://groups.google.com/group/sqlalchemy.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Have a nice day !!!

[Praveen]

Sorry, it should've been:

class Enum_Sample(Base):

Typo.

[Michael Bayer]

On Aug 26, 2013, at 4:35 PM, Praveen <praveen...@gmail.com> wrote:

The problem with using Mixins is that you need to know the definition of columns already for creating the mixin class. What I am trying to do is more like get the definition dynamically on the fly.Take a look at this:

here's a simple way to add reflection events which intercept the Column and add the .key of your choice, and saves on code by making use of the existing DeferredReflection mixin instead of hand-coding it.  There's many more variants on this kind of thing as well.   DeferredReflection is using some not-yet-public APIs to do its work, but there are other ways of deferring the mapping as well (typically by overriding the __mapper__ attribute).   Basically your class creates itself an empty Table object, to which we apply events.  then DeferredReflection.prepare() fills in those Table objects with their columns using reflection.

from sqlalchemy import *

from sqlalchemy.orm import *

from sqlalchemy.ext.declarative import declarative_base

from sqlalchemy.ext.declarative import DeferredReflection

from sqlalchemy import event

Base = declarative_base()

e = create_engine("sqlite://", echo=True)

e.execute("""

    create table sample (

        Id integer primary key,

        Name varchar,

        Description varchar,

        IsActive varchar

    )

""")

class MagicMappyThing(DeferredReflection):

    @classmethod

    def __declare_last__(cls):

        @event.listens_for(cls.__table__, "column_reflect")

        def new_col(inspector, table, column_info):

            column_info['key'] = cls.__map__.get(column_info['name'], column_info['name'])

class Sample(MagicMappyThing, Base):

    __tablename__ = 'sample'

    __map__ = {'Id': 'id', 'Name': 'name', 'Description': 'description', 'IsActive': 'is_active'}

    def __init__(self, id, name, description, is_active):

        self.id = id

        self.name = name

        self.description = description

        self.is_active = is_active

MagicMappyThing.prepare(e)

s = Session(e)

s.add(Sample(id=1, name='some name', description='foo', is_active='foo'))

s.commit()

[Michael Bayer]

On Aug 26, 2013, at 5:16 PM, Michael Bayer <mik...@zzzcomputing.com> wrote:

On Aug 26, 2013, at 4:35 PM, Praveen <praveen...@gmail.com> wrote:

The problem with using Mixins is that you need to know the definition of columns already for creating the mixin class. What I am trying to do is more like get the definition dynamically on the fly.Take a look at this:

here's a simple way to add reflection events which intercept the Column and add the .key of your choice, and saves on code by making use of the existing DeferredReflection mixin instead of hand-coding it.

except it doesnt work yet, give me 10 minutes.

[Michael Bayer]

OK here we are, had to switch approaches due to a bug with the column reflect event, to use the aforementioned __mapper_cls__ (had the name wrong), so I think you'll see this is a pretty open-ended way to control how something maps as you're given total access to mapper() here:

from sqlalchemy import *

from sqlalchemy.orm import *

from sqlalchemy.ext.declarative import declarative_base, declared_attr

from sqlalchemy.ext.declarative import DeferredReflection

from sqlalchemy import event

Base = declarative_base()

e = create_engine("sqlite://", echo=True)

e.execute("""

    create table sample (

        Id integer primary key,

        Name varchar,

        Description varchar,

        IsActive varchar

    )

""")

class MagicMappyThing(DeferredReflection):

    @declared_attr

    def __mapper_cls__(cls):

        def map_(cls, *arg, **kw):

            props = kw.setdefault("properties", {})

            for k, v in cls.__map__.items():

                props[v] = cls.__table__.c[k]

            return mapper(cls, *arg, **kw)

        return map_

class Sample(MagicMappyThing, Base):

    __tablename__ = 'sample'

    __map__ = {'Id': 'id', 'Name': 'name', 'Description': 'description', 'IsActive': 'is_active'}

    def __init__(self, id, name, description, is_active):

        self.id = id

        self.name = name

        self.description = description

        self.is_active = is_active

MagicMappyThing.prepare(e)

s = Session(e)

s.add(Sample(id=1, name='some name', description='foo', is_active='foo'))

s.commit()

[Praveen]

Does this work in sqlalchemy 0.6.1 ?

[Praveen]

I am getting ImportError for the following:

from sqlalchemy.ext.declarative import DeferredReflection

from sqlalchemy import event

I use sqlalchemy 0.6.1. Is there any way I can make it work in 0.6.1 ?

[Michael Bayer]

you'd need to hand-roll the deferred reflection part, there's an example in 0.7 called "declarative_reflection" but it might require features that aren't in 0.6.

I'd not be looking to add any kind of slick/magic systems to an 0.6 app, 0.6 is very early in the curve for declarative techniques.  upgrade it first, otherwise stick with the hacky approaches you have.

[Praveen]

Could you please point me to the link where I can find the example ?

[Praveen]

nvm... i found it.

[Praveen]

It works only in 0.7 like you said. I can't find any way to crack this situation in 0.6. :-(

[Praveen]

I tried your example in sqlalchemy 0.6 by manually plugging in api.py library (attached) that I got from here. 

I get this error:

File "path\to\sample_orm.py", line 33, in map_

    props[v] = cls.__table__.c[k]

  File "path\to\lib\python2.6\sqlalchemy\util.py", line 731, in __getitem__

KeyError: 'Description'

Here's my code:

from sqlalchemy.orm import mapper

from sqlalchemy import create_engine

from sqlalchemy.ext.declarative import declarative_base

from .api import DeferredReflection, declared_attr --> this is coming from attached api.py file

Base = declarative_base()

e = create_engine("sqlite://", echo=True)

e.execute("""

    create table Sample (

        Id integer primary key,

        Name varchar,

        Description varchar,

        IsActive varchar

    )

""")

class MagicMappyThing(DeferredReflection):

    @declared_attr

    def __mapper_cls__(cls):

        def map_(cls, *arg, **kw):

            props = kw.setdefault("properties", {})

            for k, v in cls.__map__.items():

                props[v] = cls.__table__.c[k]

            return mapper(cls, *arg, **kw)

        return map_

class Sample(MagicMappyThing, Base):

    __tablename__ = 'Sample'

    __map__ = {'Id': 'id', 'Name': 'name', 'Description': 'description', 'IsActive': 'is_active'}

    def __init__(self, id, name, description, is_active):

        self.id = id

        self.name = name

        self.description = description

        self.is_active = is_active

MagicMappyThing.prepare(e)

# s = Session(e)

# s.add(Sample(id=1, name='some name', description='foo', is_active='foo'))

# s.commit()

[Michael Bayer]

On Aug 26, 2013, at 7:32 PM, Praveen <praveen...@gmail.com> wrote:

I tried your example in sqlalchemy 0.6 by manually plugging in api.py library (attached) that I got from here. 

its not going to work.  you need to upgrade.