[sqlalchemy] Re: when is object.id initialized
OK, thanks, this was part of the ActiveRecord kind of approach I was playing with, which after reading your article at zzzeek and the alternative described there I will probably shelve. On Apr 6, 9:59 pm, Michael Bayer wrote: > On Apr 6, 2011, at 6:38 AM, farcat wrote: > > > Thank you, > > > I now experiment with putting session.add and session.flush in > > object.__init__ .. > > > Are there any general disadvantages of that approach? > > Using add() inside of __init__ is somewhat common. Using flush() has the > significant downside that flushes occur too often which is wasteful and > performs poorly for larger scale operations (like bulk loads and such). The > ORM is designed such that primary key identifiers are not needed to be > explicitly accessed outside of a flush except for query situations that wish > to avoid the usage of relationships. When you build your application to be > strongly dependent on primary key identifiers being available within > otherwise fully pending object graphs, you begin to work against the usage > goals of the ORM. > > > > > > > > > > > On Apr 3, 7:44 pm, Michael Bayer wrote: > >> Integer primary key identifiers are generated by the database itself using > >> a variety of techniques which are all database-dependent. This process > >> occurs when the session flushes. > > >> If you read the object relational tutorial starting > >> athttp://www.sqlalchemy.org/docs/orm/tutorial.html#setting-up-the-mappi...working > >> down through the end > >> ofhttp://www.sqlalchemy.org/docs/orm/tutorial.html#adding-new-objectsyouwill > >> see that this interaction is described. > > >> You can of course set .id to any value you'd like and that will be the > >> value used when the flush happens. > > >> On Apr 3, 2011, at 1:09 PM, farcat wrote: > > >>> Hi all, > > >>> I use a kind of dynamic reference from parent_table to other tables. > >>> For that parent_table uses columns table_name and a record_id. This > >>> makes it possible to have a reference from parent_table to any record > >>> in any table in the database. However, say that i want to reference a > >>> record of table_name, i need the record.id to initialize > >>> parent_table.record_id. However, when i create a record and > >>> session.add it to the database, record.id == None. > > >>> I was wondering when and how record.id is initialized and how it can > >>> be forced. > > >>> Cheers, Lars > > >>> -- > >>> 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 > >>> athttp://groups.google.com/group/sqlalchemy?hl=en. > > > -- > > 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 > > athttp://groups.google.com/group/sqlalchemy?hl=en. -- 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] Re: when is object.id initialized
On Apr 6, 2011, at 6:38 AM, farcat wrote: > Thank you, > > I now experiment with putting session.add and session.flush in > object.__init__ .. > > Are there any general disadvantages of that approach? Using add() inside of __init__ is somewhat common. Using flush() has the significant downside that flushes occur too often which is wasteful and performs poorly for larger scale operations (like bulk loads and such). The ORM is designed such that primary key identifiers are not needed to be explicitly accessed outside of a flush except for query situations that wish to avoid the usage of relationships. When you build your application to be strongly dependent on primary key identifiers being available within otherwise fully pending object graphs, you begin to work against the usage goals of the ORM. > > > > On Apr 3, 7:44 pm, Michael Bayer wrote: >> Integer primary key identifiers are generated by the database itself using a >> variety of techniques which are all database-dependent. This process occurs >> when the session flushes. >> >> If you read the object relational tutorial starting >> athttp://www.sqlalchemy.org/docs/orm/tutorial.html#setting-up-the-mappingand >> working down through the end >> ofhttp://www.sqlalchemy.org/docs/orm/tutorial.html#adding-new-objectsyou >> will see that this interaction is described. >> >> You can of course set .id to any value you'd like and that will be the value >> used when the flush happens. >> >> On Apr 3, 2011, at 1:09 PM, farcat wrote: >> >> >> >> >> >> >> >>> Hi all, >> >>> I use a kind of dynamic reference from parent_table to other tables. >>> For that parent_table uses columns table_name and a record_id. This >>> makes it possible to have a reference from parent_table to any record >>> in any table in the database. However, say that i want to reference a >>> record of table_name, i need the record.id to initialize >>> parent_table.record_id. However, when i create a record and >>> session.add it to the database, record.id == None. >> >>> I was wondering when and how record.id is initialized and how it can >>> be forced. >> >>> Cheers, Lars >> >>> -- >>> 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 >>> athttp://groups.google.com/group/sqlalchemy?hl=en. > > -- > 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. > -- 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.
[sqlalchemy] Re: when is object.id initialized
Thank you, I now experiment with putting session.add and session.flush in object.__init__ .. Are there any general disadvantages of that approach? On Apr 3, 7:44 pm, Michael Bayer wrote: > Integer primary key identifiers are generated by the database itself using a > variety of techniques which are all database-dependent. This process occurs > when the session flushes. > > If you read the object relational tutorial starting > athttp://www.sqlalchemy.org/docs/orm/tutorial.html#setting-up-the-mappingand > working down through the end > ofhttp://www.sqlalchemy.org/docs/orm/tutorial.html#adding-new-objectsyou will > see that this interaction is described. > > You can of course set .id to any value you'd like and that will be the value > used when the flush happens. > > On Apr 3, 2011, at 1:09 PM, farcat wrote: > > > > > > > > > Hi all, > > > I use a kind of dynamic reference from parent_table to other tables. > > For that parent_table uses columns table_name and a record_id. This > > makes it possible to have a reference from parent_table to any record > > in any table in the database. However, say that i want to reference a > > record of table_name, i need the record.id to initialize > > parent_table.record_id. However, when i create a record and > > session.add it to the database, record.id == None. > > > I was wondering when and how record.id is initialized and how it can > > be forced. > > > Cheers, Lars > > > -- > > 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 > > athttp://groups.google.com/group/sqlalchemy?hl=en. -- 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.
