[IMP] Process jobs 'validate_one_move' in a dedicated job channel

To avoid concurrent update when validating the account_move
on ir_sequence, jobs must be processed one by one. Using a
dedicated channel enable to isolate the processing of
the 'validate_one_move' jobs from the root channel

account_move_batch_validate/__openerp__.py

@@ -32,6 +32,7 @@
     'website': 'http://www.camptocamp.com',
     'data': [
         'account_view.xml',
+        'account_data.xml',
         'wizard/move_marker_view.xml',
     ],
     'test': [

account_move_batch_validate/account.py

@@ -28,6 +28,10 @@ from openerp.tools.translate import _
 from openerp.addons.connector.queue.job import job
 from openerp.addons.connector.session import ConnectorSession
 from openerp.addons.connector.queue.job import OpenERPJobStorage
+from openerp.addons.connector.connector import install_in_connector
+
+# install the module in connector to register the job function
+install_in_connector()
 
 _logger = logging.getLogger(__name__)
 

account_move_batch_validate/account_data.xml

@@ -0,0 +1,19 @@
+<?xml version="1.0" encoding="utf-8"?>
+<openerp>
+    <data noupdate="0">
+
+        <!-- Connector channel: To avoid concurrent update when validating the account_move
+        on ir_sequence, jobs must be processed one by one. By default channels are created
+        with a capacity of 1. Be careful to not increase the capacity of the channel. -->
+        <record model="queue.job.channel" id="channel_account_move_batch_validate">
+            <field name="name">account_move_batch_validate</field>
+            <field name="parent_id" ref="connector.channel_root"/>
+        </record>
+
+        <record model="queue.job.function" id="function_account_validate_one_move">
+            <field name="name">openerp.addons.account_move_batch_validate.account.validate_one_move</field>
+            <field name="channel_id" ref="channel_account_move_batch_validate"/>
+        </record>
+
+    </data>
+</openerp>