Re: [sqlalchemy] declarative and late reflection?

2011-12-27 Thread peter sabaini
Cool -- works nicely, thanks again!

On Fri, Dec 23, 2011 at 2:56 AM, Michael Bayer mike...@zzzcomputing.comwrote:


 On Dec 22, 2011, at 7:28 PM, Michael Bayer wrote:

  this could work really nicely with extend_existing, which has been
 enhanced in 0.7.4, but there seem to be some glitches preventing it from
 being super nice, so I can't get that to work right now.  Just send in
 those columns via your own means:

 Those glitches have a pending patch in ticket 2356:
 http://www.sqlalchemy.org/trac/ticket/2356  See the modified example case
 there.  In 0.7.5 your use case will work exactly as you intend, where
 you'll be able to add Column objects to your declared class that will take
 precedence over what's autoloaded.


 --
 You received this message because you are subscribed to the Google Groups
 sqlalchemy group.
 To post to this group, send email to sqlalchemy@googlegroups.com.
 To unsubscribe from this group, send email to
 sqlalchemy+unsubscr...@googlegroups.com.
 For more options, visit this group at
 http://groups.google.com/group/sqlalchemy?hl=en.




-- 
http://sabaini.at

-- 
You received this message because you are subscribed to the Google Groups 
sqlalchemy group.
To post to this group, send email to sqlalchemy@googlegroups.com.
To unsubscribe from this group, send email to 
sqlalchemy+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/sqlalchemy?hl=en.



Re: [sqlalchemy] declarative and late reflection?

2011-12-22 Thread Michael Bayer

On Dec 22, 2011, at 9:37 AM, peter sabaini wrote:

 Hey list,
 
 this sounds like it should be a FAQ, didn't find anything though:
 
 I want to use the ORM in a declarative style and have the table
 definition reflected, eg sth like:
 
 class A(declarative_base()):
__tablename__ = 'A'
__table_args__ = {'autoload' : True}
 
 However to do this SA (quite reasonably, really) already needs an
 engine. For various reasons I can only construct one after import time
 however. Is there a way to do a kind of late reflection, ie. have
 the above class definition but trigger the reflection part later? I
 could of course generate the class object later in a function when the
 engine is already available but maybe there's something more
 elegant...

I was about to type up this recipe on the wiki and then most awesomely I 
already did it for someone !   hooray.  The current technique for this is at 
http://www.sqlalchemy.org/trac/wiki/UsageRecipes/DeclarativeReflectedBase .


-- 
You received this message because you are subscribed to the Google Groups 
sqlalchemy group.
To post to this group, send email to sqlalchemy@googlegroups.com.
To unsubscribe from this group, send email to 
sqlalchemy+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/sqlalchemy?hl=en.



Re: [sqlalchemy] declarative and late reflection?

2011-12-22 Thread peter sabaini
Hey!

This works for me -- almost :-)

In my use case I need to override a column (to provide an artificial FK --
some *erm old school mysql db) which seems to trigger SA into trying to
reflect early

Observe:

from sqlalchemy import *
from sqlalchemy.orm import *
from sqlalchemy.ext.declarative import declarative_base, declared_attr

class DeclarativeReflectedBase(object):
_mapper_args = []

@classmethod
def __mapper_cls__(cls, *args, **kw):
Declarative will use this function in lieu of
calling mapper() directly.

Collect each series of arguments and invoke
them when prepare() is called.


cls._mapper_args.append((args, kw))

@classmethod
def prepare(cls, engine):
Reflect all the tables and map !
for args, kw in cls._mapper_args:
klass = args[0]
klass.__table__ = table = Table(
klass.__tablename__,
cls.metadata,
autoload=True,
autoload_with=engine)
klass.__mapper__ = mapper(klass, table, **kw)

@declared_attr
def __table__(cls):
Return a placeholder to lull declarative into complacency
return object()

Base = declarative_base(cls=DeclarativeReflectedBase)

class Foo(Base):
__tablename__ = 'foo'
quux = Column(String)
bars = relationship(Bar)

class Bar(Base):
__tablename__ = 'bar'


Gives:

Traceback (most recent call last):
  File stdin, line 1, in module
  File /usr/tmp/Python6244Mux.py, line 40, in module
class Foo(Base):
  File
/usr/lib/python2.6/site-packages/SQLAlchemy-0.7.4-py2.6-linux-x86_64.egg/sqlalchemy/ext/declarative.py,
line 1273, in __init__
_as_declarative(cls, classname, cls.__dict__)
  File
/usr/lib/python2.6/site-packages/SQLAlchemy-0.7.4-py2.6-linux-x86_64.egg/sqlalchemy/ext/declarative.py,
line 1177, in _as_declarative
if not table.c.contains_column(c):
AttributeError: 'object' object has no attribute 'c'

Hm, maybe I can try to add the column override later?

Thanks again

peter.

On Thu, Dec 22, 2011 at 4:41 PM, Michael Bayer mike...@zzzcomputing.comwrote:


 On Dec 22, 2011, at 9:37 AM, peter sabaini wrote:

  Hey list,
 
  this sounds like it should be a FAQ, didn't find anything though:
 
  I want to use the ORM in a declarative style and have the table
  definition reflected, eg sth like:
 
  class A(declarative_base()):
 __tablename__ = 'A'
 __table_args__ = {'autoload' : True}
 
  However to do this SA (quite reasonably, really) already needs an
  engine. For various reasons I can only construct one after import time
  however. Is there a way to do a kind of late reflection, ie. have
  the above class definition but trigger the reflection part later? I
  could of course generate the class object later in a function when the
  engine is already available but maybe there's something more
  elegant...

 I was about to type up this recipe on the wiki and then most awesomely I
 already did it for someone !   hooray.  The current technique for this is
 at
 http://www.sqlalchemy.org/trac/wiki/UsageRecipes/DeclarativeReflectedBase.


 --
 You received this message because you are subscribed to the Google Groups
 sqlalchemy group.
 To post to this group, send email to sqlalchemy@googlegroups.com.
 To unsubscribe from this group, send email to
 sqlalchemy+unsubscr...@googlegroups.com.
 For more options, visit this group at
 http://groups.google.com/group/sqlalchemy?hl=en.




-- 
http://sabaini.at

-- 
You received this message because you are subscribed to the Google Groups 
sqlalchemy group.
To post to this group, send email to sqlalchemy@googlegroups.com.
To unsubscribe from this group, send email to 
sqlalchemy+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/sqlalchemy?hl=en.



Re: [sqlalchemy] declarative and late reflection?

2011-12-22 Thread Michael Bayer

On Dec 22, 2011, at 11:48 AM, peter sabaini wrote:

 Hey!
 
 This works for me -- almost :-)
 
 In my use case I need to override a column (to provide an artificial FK -- 
 some *erm old school mysql db) which seems to trigger SA into trying to 
 reflect early
 
 Observe:
 
 from sqlalchemy import *
 from sqlalchemy.orm import *
 from sqlalchemy.ext.declarative import declarative_base, declared_attr
 
 class DeclarativeReflectedBase(object):
 _mapper_args = []
 
 @classmethod
 def __mapper_cls__(cls, *args, **kw):
 Declarative will use this function in lieu of 
 calling mapper() directly.
 
 Collect each series of arguments and invoke
 them when prepare() is called.
 
 
 cls._mapper_args.append((args, kw))
 
 @classmethod
 def prepare(cls, engine):
 Reflect all the tables and map !
 for args, kw in cls._mapper_args:
 klass = args[0]
 klass.__table__ = table = Table(
 klass.__tablename__, 
 cls.metadata, 
 autoload=True, 
 autoload_with=engine)
 klass.__mapper__ = mapper(klass, table, **kw)
 
 @declared_attr
 def __table__(cls):
 Return a placeholder to lull declarative into complacency
 return object()
 
 Base = declarative_base(cls=DeclarativeReflectedBase)
 
 class Foo(Base):
 __tablename__ = 'foo'
 quux = Column(String)
 bars = relationship(Bar)
 
 class Bar(Base):
 __tablename__ = 'bar'
 
 

yah well you want to not declare Column objects on the declared class directly 
in this case because that trips off declarative into building up a Table, which 
you don't have here, it only returns object.

this could work really nicely with extend_existing, which has been enhanced in 
0.7.4, but there seem to be some glitches preventing it from being super nice, 
so I can't get that to work right now.  Just send in those columns via your own 
means:

class Base(object):
# ...

extra_cols = []

@classmethod
def prepare(cls, engine):
Reflect all the tables and map !
for args, kw in cls._mapper_args:
klass = args[0]
klass.__table__ = table = Table(
klass.__tablename__, 
cls.metadata, 
autoload=True, 
autoload_with=engine,
*klass.extra_cols
)
klass.__mapper__ = mapper(klass, table, **kw)


class Bar(Base):
__tablename__ = 'bar'

extra_cols = [
Column('foo_id', Integer, ForeignKey('foo.id')),
]

-- 
You received this message because you are subscribed to the Google Groups 
sqlalchemy group.
To post to this group, send email to sqlalchemy@googlegroups.com.
To unsubscribe from this group, send email to 
sqlalchemy+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/sqlalchemy?hl=en.



Re: [sqlalchemy] declarative and late reflection?

2011-12-22 Thread Michael Bayer

On Dec 22, 2011, at 7:28 PM, Michael Bayer wrote:

 this could work really nicely with extend_existing, which has been enhanced 
 in 0.7.4, but there seem to be some glitches preventing it from being super 
 nice, so I can't get that to work right now.  Just send in those columns via 
 your own means:

Those glitches have a pending patch in ticket 2356: 
http://www.sqlalchemy.org/trac/ticket/2356  See the modified example case 
there.  In 0.7.5 your use case will work exactly as you intend, where you'll be 
able to add Column objects to your declared class that will take precedence 
over what's autoloaded.


-- 
You received this message because you are subscribed to the Google Groups 
sqlalchemy group.
To post to this group, send email to sqlalchemy@googlegroups.com.
To unsubscribe from this group, send email to 
sqlalchemy+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/sqlalchemy?hl=en.