[WIP] all: (work in progress) changing wizards to avoid using read() or at least take proper care of m2o fields

This is in preparation of the fix for m2o.get() on osv_memory objects, to make
it consistent with the regular read on osv objects. Previously on osv_memory
objects it was only returned the id, instead of the pair (id, name) as is appropriate
to avoid an additional name_get RPC call after each read.
Unfortunately many wizards were using direct read() instead of browse() and
relying on the fact that the returned dict contained a bare id value for
many2one fields.

bzr revid: odo@openerp.com-20110201210839-hwmk60ygksm805c4

1 files changed, 2 insertions, 0 deletions

diff --git a/addons/project/wizard/project_task_delegate.py b/addons/project/wizard/project_task_delegate.py
index 72b32f7db68..16c3ec85653 100644
--- a/addons/project/wizard/project_task_delegate.py
+++ b/addons/project/wizard/project_task_delegate.py
@@ -105,6 +105,8 @@ class project_task_delegate(osv.osv_memory):
         task_id = context.get('active_id', False)
         task_pool = self.pool.get('project.task')
         delegate_data = self.read(cr, uid, ids, context=context)[0]
+        if isinstance(delegate_data['user_id'], tuple):
+            delegate_data['user_id'] = delegate_data['user_id'][0]
         delegate_data['name'] = tools.ustr(delegate_data['name'])
         task_pool.do_delegate(cr, uid, task_id, delegate_data, context=context)
         return {}