credativ team mailing list archive

[Pieter van der Merwe <911133@xxxxxxxxxxxxxxxxxx>]

Public bug reported:

If you are bulk importing historical data or master files and a records
is linked to an inactive master file (res.partner, account.account) etc
the row is rejected.

You can set the property 'active_test' in the context object to false
but the context object isn't passed through by the import function when
a lookup is being done.

Here is the fix to the problem (against trunk in the import_data
function):

bzr diff openerp/osv/orm.py:
=== modified file 'openerp/osv/orm.py'
--- openerp/osv/orm.py  2011-12-01 12:15:03 +0000
+++ openerp/osv/orm.py  2012-12-02 08:23:36 +0000
@@ -1219,7 +1219,7 @@
             if mode=='.id':
                 id = int(id)
                 obj_model = self.pool.get(model_name)
-                ids = obj_model.search(cr, uid, [('id', '=', int(id))])
+                ids = obj_model.search(cr, uid, [('id', '=', int(id))], context=context)
                 if not len(ids):
                     raise Exception(_("Database ID doesn't exist: %s : %s") %(model_name, id))
             elif mode=='id':

For example I am using the new OpenERP plugin in Kettle to load
account.account records from an existing ERP.  I can only load active
accounts since inactive accounts can't be referenced.  I will lose all
historical data for those accounts if I can't link to them.

Kind Regards,
Pieter

Edit: I see the patch isn't formed correctly, but i'm sure you can see
that context=context needs to be added to the search function.

** Affects: openobject-server
     Importance: Undecided
         Status: New

** Description changed:

  If you are bulk importing historical data or master files and a records
  is linked to an inactive master file (res.partner, account.account) etc
  the row is rejected.
  
  You can set the property 'active_test' in the context object to false
  but the context object isn't passed through by the import function when
  a lookup is being done.
  
  Here is the fix to the problem (against trunk in the import_data
  function):
  
  bzr diff openerp/osv/orm.py:
  === modified file 'openerp/osv/orm.py'
  --- openerp/osv/orm.py  2011-12-01 12:15:03 +0000
  +++ openerp/osv/orm.py  2012-12-02 08:23:36 +0000
  @@ -1219,7 +1219,7 @@
-              if mode=='.id':
-                  id = int(id)
-                  obj_model = self.pool.get(model_name)
+              if mode=='.id':
+                  id = int(id)
+                  obj_model = self.pool.get(model_name)
  -                ids = obj_model.search(cr, uid, [('id', '=', int(id))])
  +                ids = obj_model.search(cr, uid, [('id', '=', int(id))], context=context)
-                  if not len(ids):
-                      raise Exception(_("Database ID doesn't exist: %s : %s") %(model_name, id))
-              elif mode=='id':
+                  if not len(ids):
+                      raise Exception(_("Database ID doesn't exist: %s : %s") %(model_name, id))
+              elif mode=='id':
  
  For example I am using the new OpenERP plugin in Kettle to load
  account.account records from an existing ERP.  I can only load active
  accounts since inactive accounts can't be referenced.  I will lose all
  historical data for those accounts if I can't link to them.
  
  Kind Regards,
  Pieter
+ 
+ Edit: I see the patch isn't formed correctly, but i'm sure yuo can see
+ that context=context needs to be added to the search function.

** Description changed:

  If you are bulk importing historical data or master files and a records
  is linked to an inactive master file (res.partner, account.account) etc
  the row is rejected.
  
  You can set the property 'active_test' in the context object to false
  but the context object isn't passed through by the import function when
  a lookup is being done.
  
  Here is the fix to the problem (against trunk in the import_data
  function):
  
  bzr diff openerp/osv/orm.py:
  === modified file 'openerp/osv/orm.py'
  --- openerp/osv/orm.py  2011-12-01 12:15:03 +0000
  +++ openerp/osv/orm.py  2012-12-02 08:23:36 +0000
  @@ -1219,7 +1219,7 @@
               if mode=='.id':
                   id = int(id)
                   obj_model = self.pool.get(model_name)
  -                ids = obj_model.search(cr, uid, [('id', '=', int(id))])
  +                ids = obj_model.search(cr, uid, [('id', '=', int(id))], context=context)
                   if not len(ids):
                       raise Exception(_("Database ID doesn't exist: %s : %s") %(model_name, id))
               elif mode=='id':
  
  For example I am using the new OpenERP plugin in Kettle to load
  account.account records from an existing ERP.  I can only load active
  accounts since inactive accounts can't be referenced.  I will lose all
  historical data for those accounts if I can't link to them.
  
  Kind Regards,
  Pieter
  
- Edit: I see the patch isn't formed correctly, but i'm sure yuo can see
+ Edit: I see the patch isn't formed correctly, but i'm sure you can see
  that context=context needs to be added to the search function.

-- 
You received this bug notification because you are a member of OpenERP
Framework Experts, which is subscribed to OpenERP Server.
https://bugs.launchpad.net/bugs/911133

Title:
  import_data - osv.  Context object isn't passed when obj_model.search
  is called

Status in OpenERP Server:
  New

Bug description:
  If you are bulk importing historical data or master files and a
  records is linked to an inactive master file (res.partner,
  account.account) etc the row is rejected.

  You can set the property 'active_test' in the context object to false
  but the context object isn't passed through by the import function
  when a lookup is being done.

  Here is the fix to the problem (against trunk in the import_data
  function):

  bzr diff openerp/osv/orm.py:
  === modified file 'openerp/osv/orm.py'
  --- openerp/osv/orm.py  2011-12-01 12:15:03 +0000
  +++ openerp/osv/orm.py  2012-12-02 08:23:36 +0000
  @@ -1219,7 +1219,7 @@
               if mode=='.id':
                   id = int(id)
                   obj_model = self.pool.get(model_name)
  -                ids = obj_model.search(cr, uid, [('id', '=', int(id))])
  +                ids = obj_model.search(cr, uid, [('id', '=', int(id))], context=context)
                   if not len(ids):
                       raise Exception(_("Database ID doesn't exist: %s : %s") %(model_name, id))
               elif mode=='id':

  For example I am using the new OpenERP plugin in Kettle to load
  account.account records from an existing ERP.  I can only load active
  accounts since inactive accounts can't be referenced.  I will lose all
  historical data for those accounts if I can't link to them.

  Kind Regards,
  Pieter

  Edit: I see the patch isn't formed correctly, but i'm sure you can see
  that context=context needs to be added to the search function.

To manage notifications about this bug go to:
https://bugs.launchpad.net/openobject-server/+bug/911133/+subscriptions