[FIX] Wrong domain value was set while converting XML records of action to database...
[odoo/odoo.git] / bin / tools / convert.py
index 24e578a..dc85f19 100644 (file)
@@ -1,36 +1,30 @@
+# -*- encoding: utf-8 -*-
 ##############################################################################
 #
-# Copyright (c) 2004-2008 Tiny SPRL (http://tiny.be) All Rights Reserved.
+#    OpenERP, Open Source Management Solution
+#    Copyright (C) 2004-2009 Tiny SPRL (<http://tiny.be>). All Rights Reserved
+#    $Id$
 #
-# $Id$
+#    This program is free software: you can redistribute it and/or modify
+#    it under the terms of the GNU General Public License as published by
+#    the Free Software Foundation, either version 3 of the License, or
+#    (at your option) any later version.
 #
-# WARNING: This program as such is intended to be used by professional
-# programmers who take the whole responsability of assessing all potential
-# consequences resulting from its eventual inadequacies and bugs
-# End users who are looking for a ready-to-use solution with commercial
-# garantees and support are strongly adviced to contract a Free Software
-# Service Company
+#    This program is distributed in the hope that it will be useful,
+#    but WITHOUT ANY WARRANTY; without even the implied warranty of
+#    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+#    GNU General Public License for more details.
 #
-# This program is Free Software; you can redistribute it and/or
-# modify it under the terms of the GNU General Public License
-# as published by the Free Software Foundation; either version 2
-# of the License, or (at your option) any later version.
+#    You should have received a copy of the GNU General Public License
+#    along with this program.  If not, see <http://www.gnu.org/licenses/>.
 #
-# This program is distributed in the hope that it will be useful,
-# but WITHOUT ANY WARRANTY; without even the implied warranty of
-# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
-# GNU General Public License for more details.
-#
-# You should have received a copy of the GNU General Public License
-# along with this program; if not, write to the Free Software
-# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
-###############################################################################
-#----------------------------------------------------------
-# Convert
-#----------------------------------------------------------
+##############################################################################
 import re
-import StringIO,xml.dom.minidom
-import osv,ir,pooler
+import cStringIO
+from lxml import etree
+import osv
+import ir
+import pooler
 
 import csv
 import os.path
@@ -39,742 +33,877 @@ import netsvc
 
 from config import config
 import logging
+import tools
 
-from lxml import etree
+import sys
+import pickle
+
+unsafe_eval = eval
 
+from tools.safe_eval import safe_eval as eval 
 
 class ConvertError(Exception):
-       def __init__(self, doc, orig_excpt):
-               self.d = doc
-               self.orig = orig_excpt
+    def __init__(self, doc, orig_excpt):
+        self.d = doc
+        self.orig = orig_excpt
 
-       def __str__(self):
-               return 'Exception:\n\t%s\nUsing file:\n%s' % (self.orig, self.d)
+    def __str__(self):
+        return 'Exception:\n\t%s\nUsing file:\n%s' % (self.orig, self.d)
 
 def _ref(self, cr):
-       return lambda x: self.id_get(cr, False, x)
+    return lambda x: self.id_get(cr, False, x)
 
 def _obj(pool, cr, uid, model_str, context=None):
-       model = pool.get(model_str)
-       return lambda x: model.browse(cr, uid, x, context=context)
+    model = pool.get(model_str)
+    return lambda x: model.browse(cr, uid, x, context=context)
+
+def _get_idref(self, cr, uid, model_str, context, idref):
+    import time
+    from mx import DateTime
+    idref2 = idref.copy()
+    idref2['time'] = time
+    idref2['DateTime'] = DateTime
+    import release
+    idref2['version'] = release.major_version
+    idref2['ref'] = _ref(self, cr)
+    if len(model_str):
+        idref2['obj'] = _obj(self.pool, cr, uid, model_str, context=context)
+    try:
+        import pytz
+    except:
+        logger = netsvc.Logger()
+        logger.notifyChannel("init", netsvc.LOG_WARNING, 'could not find pytz library')
+        class pytzclass(object):
+            all_timezones = []
+        pytz = pytzclass()
+    idref2['pytz'] = pytz
+    return idref2
 
 def _eval_xml(self,node, pool, cr, uid, idref, context=None):
-       if context is None:
-               context = {}
-       if node.nodeType == node.TEXT_NODE:
-               return node.data.encode("utf8")
-       elif node.nodeType == node.ELEMENT_NODE:
-               if node.nodeName in ('field','value'):
-                       t = node.getAttribute('type') or 'char'
-                       f_model = node.getAttribute("model").encode('ascii')
-                       if len(node.getAttribute('search')):
-                               f_search = node.getAttribute("search").encode('utf-8')
-                               f_use = node.getAttribute("use").encode('ascii')
-                               f_name = node.getAttribute("name").encode('utf-8')
-                               if len(f_use)==0:
-                                       f_use = "id"
-                               q = eval(f_search, idref)
-                               ids = pool.get(f_model).search(cr, uid, q)
-                               if f_use<>'id':
-                                       ids = map(lambda x: x[f_use], pool.get(f_model).read(cr, uid, ids, [f_use]))
-                               _cols = pool.get(f_model)._columns
-                               if (f_name in _cols) and _cols[f_name]._type=='many2many':
-                                       return ids
-                               f_val = False
-                               if len(ids):
-                                       f_val = ids[0]
-                                       if isinstance(f_val, tuple):
-                                               f_val = f_val[0]
-                               return f_val
-                       a_eval = node.getAttribute('eval')
-                       if len(a_eval):
-                               import time
-                               idref['time'] = time
-                               import release
-                               idref['version'] = release.version.rsplit('.', 1)[0]
-                               idref['ref'] = lambda x: self.id_get(cr, False, x)
-                               if len(f_model):
-                                       idref['obj'] = _obj(self.pool, cr, uid, f_model, context=context)
-                               try:
-                                       import pytz
-                               except:
-                                       logger = netsvc.Logger()
-                                       logger.notifyChannel("init", netsvc.LOG_INFO, 'could not find pytz library')
-                                       class pytzclass(object):
-                                               all_timezones=[]
-                                       pytz=pytzclass()
-                               idref['pytz'] = pytz
-                               return eval(a_eval, idref)
-                       if t == 'xml':
-                               def _process(s, idref):
-                                       m = re.findall('[^%]%\((.*?)\)[ds]', s)
-                                       for id in m:
-                                               if not id in idref:
-                                                       idref[id]=self.id_get(cr, False, id)
-                                       return s % idref
-                               txt = '<?xml version="1.0"?>\n'+_process("".join([i.toxml().encode("utf8") for i in node.childNodes]), idref)
-#                              txt = '<?xml version="1.0"?>\n'+"".join([i.toxml().encode("utf8") for i in node.childNodes]) % idref
-
-                               return txt
-                       if t in ('char', 'int', 'float'):
-                               d = ""
-                               for n in [i for i in node.childNodes]:
-                                       d+=str(_eval_xml(self,n,pool,cr,uid,idref))
-                               if t == 'int':
-                                       d = d.strip()
-                                       if d=='None':
-                                               return None
-                                       else:
-                                               d=int(d.strip())
-                               elif t=='float':
-                                       d=float(d.strip())
-                               return d
-                       elif t in ('list','tuple'):
-                               res=[]
-                               for n in [i for i in node.childNodes if (i.nodeType == i.ELEMENT_NODE and i.nodeName=='value')]:
-                                       res.append(_eval_xml(self,n,pool,cr,uid,idref))
-                               if t=='tuple':
-                                       return tuple(res)
-                               return res
-               elif node.nodeName=="getitem":
-                       for n in [i for i in node.childNodes if (i.nodeType == i.ELEMENT_NODE)]:
-                               res=_eval_xml(self,n,pool,cr,uid,idref)
-                       if not res:
-                               raise LookupError
-                       elif node.getAttribute('type') in ("int", "list"):
-                               return res[int(node.getAttribute('index'))]
-                       else:
-                               return res[node.getAttribute('index').encode("utf8")]
-               elif node.nodeName=="function":
-                       args = []
-                       a_eval = node.getAttribute('eval')
-                       if len(a_eval):
-                               idref['ref'] = lambda x: self.id_get(cr, False, x)
-                               args = eval(a_eval, idref)
-                       for n in [i for i in node.childNodes if (i.nodeType == i.ELEMENT_NODE)]:
-                               args.append(_eval_xml(self,n, pool, cr, uid, idref, context))
-                       model = pool.get(node.getAttribute('model'))
-                       method = node.getAttribute('name')
-                       res = getattr(model, method)(cr, uid, *args)
-                       return res
-               elif node.nodeName=="test":
-                       d = ""
-                       for n in [i for i in node.childNodes]:
-                               d+=str(_eval_xml(self,n,pool,cr,uid,idref, context=context))
-                       return d
-
+    if context is None:
+        context = {}
+    if node.tag in ('field','value'):
+        t = node.get('type','char')
+        f_model = node.get('model', '').encode('ascii')
+        if node.get('search'):
+            f_search = node.get("search",'').encode('utf-8')
+            f_use = node.get("use",'id').encode('ascii')
+            f_name = node.get("name",'').encode('utf-8')
+            idref2 = {}
+            if f_search:
+                idref2 = _get_idref(self, cr, uid, f_model, context, idref)
+            q = unsafe_eval(f_search, idref2)
+            ids = pool.get(f_model).search(cr, uid, q)
+            if f_use != 'id':
+                ids = map(lambda x: x[f_use], pool.get(f_model).read(cr, uid, ids, [f_use]))
+            _cols = pool.get(f_model)._columns
+            if (f_name in _cols) and _cols[f_name]._type=='many2many':
+                return ids
+            f_val = False
+            if len(ids):
+                f_val = ids[0]
+                if isinstance(f_val, tuple):
+                    f_val = f_val[0]
+            return f_val
+        a_eval = node.get('eval','')
+        idref2 = {}
+        if a_eval:
+            idref2 = _get_idref(self, cr, uid, f_model, context, idref)
+            return unsafe_eval(a_eval, idref2)
+        if t == 'xml':
+            def _process(s, idref):
+                m = re.findall('[^%]%\((.*?)\)[ds]', s)
+                for id in m:
+                    if not id in idref:
+                        idref[id]=self.id_get(cr, False, id)
+                return s % idref
+            return '<?xml version="1.0"?>\n'\
+                +_process("".join([etree.tostring(n, encoding='utf-8')
+                                   for n in node]),
+                          idref)
+        if t in ('char', 'int', 'float'):
+            d = node.text
+            if t == 'int':
+                d = d.strip()
+                if d == 'None':
+                    return None
+                else:
+                    return int(d.strip())
+            elif t == 'float':
+                return float(d.strip())
+            return d
+        elif t in ('list','tuple'):
+            res=[]
+            for n in node.findall('./value'):
+                res.append(_eval_xml(self,n,pool,cr,uid,idref))
+            if t=='tuple':
+                return tuple(res)
+            return res
+    elif node.tag == "getitem":
+        for n in node:
+            res=_eval_xml(self,n,pool,cr,uid,idref)
+        if not res:
+            raise LookupError
+        elif node.get('type') in ("int", "list"):
+            return res[int(node.get('index'))]
+        else:
+            return res[node.get('index','').encode("utf8")]
+    elif node.tag == "function":
+        args = []
+        a_eval = node.get('eval','')
+        if a_eval:
+            idref['ref'] = lambda x: self.id_get(cr, False, x)
+            args = unsafe_eval(a_eval, idref)
+        for n in node:
+            return_val = _eval_xml(self,n, pool, cr, uid, idref, context)
+            if return_val is not None:
+                args.append(return_val)
+        model = pool.get(node.get('model',''))
+        method = node.get('name','')
+        res = getattr(model, method)(cr, uid, *args)
+        return res
+    elif node.tag == "test":
+        return node.text
 
 escape_re = re.compile(r'(?<!\\)/')
 def escape(x):
-       return x.replace('\\/', '/')
+    return x.replace('\\/', '/')
 
 class assertion_report(object):
-       def __init__(self):
-               self._report = {}
-
-       def record_assertion(self, success, severity):
-               """
-                       Records the result of an assertion for the failed/success count
-                       retrurns success
-               """
-               if severity in self._report:
-                       self._report[severity][success] += 1
-               else:
-                       self._report[severity] = {success:1, not success: 0}
-               return success
-
-       def get_report(self):
-               return self._report
-
-       def __str__(self):
-               res = '\nAssertions report:\nLevel\tsuccess\tfailed\n'
-               success = failed = 0
-               for sev in self._report:
-                       res += sev + '\t' + str(self._report[sev][True]) + '\t' + str(self._report[sev][False]) + '\n'
-                       success += self._report[sev][True]
-                       failed += self._report[sev][False]
-               res += 'total\t' + str(success) + '\t' + str(failed) + '\n'
-               res += 'end of report (' + str(success + failed) + ' assertion(s) checked)'
-               return res
+    def __init__(self):
+        self._report = {}
+
+    def record_assertion(self, success, severity):
+        """
+            Records the result of an assertion for the failed/success count
+            returns success
+        """
+        if severity in self._report:
+            self._report[severity][success] += 1
+        else:
+            self._report[severity] = {success:1, not success: 0}
+        return success
+
+    def get_report(self):
+        return self._report
+
+    def __str__(self):
+        res = '\nAssertions report:\nLevel\tsuccess\tfailed\n'
+        success = failed = 0
+        for sev in self._report:
+            res += sev + '\t' + str(self._report[sev][True]) + '\t' + str(self._report[sev][False]) + '\n'
+            success += self._report[sev][True]
+            failed += self._report[sev][False]
+        res += 'total\t' + str(success) + '\t' + str(failed) + '\n'
+        res += 'end of report (' + str(success + failed) + ' assertion(s) checked)'
+        return res
 
 class xml_import(object):
 
-       def isnoupdate(self, data_node = None):
-               return self.noupdate or (data_node and data_node.getAttribute('noupdate'))
-
-       def get_context(self, data_node, node, eval_dict):
-               data_node_context = (data_node and data_node.getAttribute('context').encode('utf8'))
-               if data_node_context:
-                       context = eval(data_node_context, eval_dict)
-               else:
-                       context = {}
-
-               node_context = node.getAttribute("context").encode('utf8')
-               if len(node_context):
-                       context.update(eval(node_context, eval_dict))
-
-               return context
-
-       def get_uid(self, cr, uid, data_node, node):
-               node_uid = node.getAttribute('uid') or (data_node and data_node.getAttribute('uid'))
-               if len(node_uid):
-                       return self.id_get(cr, None, node_uid)
-               return uid
-
-       def _test_xml_id(self, xml_id):
-               id = xml_id
-               if '.' in xml_id:
-                       assert len(id.split('.'))==2, """The ID reference '"%s" must contains
+    @staticmethod
+    def nodeattr2bool(node, attr, default=False):
+        if not node.get(attr):
+            return default
+        val = node.get(attr).strip()
+        if not val:
+            return default
+        return val.lower() not in ('0', 'false', 'off')
+
+    def isnoupdate(self, data_node=None):
+        return self.noupdate or (len(data_node) and self.nodeattr2bool(data_node, 'noupdate', False))
+
+    def get_context(self, data_node, node, eval_dict):
+        data_node_context = (len(data_node) and data_node.get('context','').encode('utf8'))
+        if data_node_context:
+            context = unsafe_eval(data_node_context, eval_dict)
+        else:
+            context = {}
+
+        node_context = node.get("context",'').encode('utf8')
+        if node_context:
+            context.update(unsafe_eval(node_context, eval_dict))
+
+        return context
+
+    def get_uid(self, cr, uid, data_node, node):
+        node_uid = node.get('uid','') or (len(data_node) and data_node.get('uid',''))
+        if node_uid:
+            return self.id_get(cr, None, node_uid)
+        return uid
+
+    def _test_xml_id(self, xml_id):
+        id = xml_id
+        if '.' in xml_id:
+            module, id = xml_id.split('.', 1)
+            assert '.' not in id, """The ID reference "%s" must contains
 maximum one dot. They are used to refer to other modules ID, in the
 form: module.record_id""" % (xml_id,)
-                       base, id = xml_id.split('.')
-               if len(id) > 64:
-                       self.logger.notifyChannel('init', netsvc.LOG_ERROR, 'id: %s is to long (max: 64)'%xml_id)
-
-       def _tag_delete(self, cr, rec, data_node=None):
-               d_model = rec.getAttribute("model")
-               d_search = rec.getAttribute("search")
-               d_id = rec.getAttribute("id")
-               ids = []
-               if len(d_search):
-                       ids = self.pool.get(d_model).search(cr,self.uid,eval(d_search))
-               if len(d_id):
-                       ids.append(self.id_get(cr, d_model, d_id))
-               if len(ids):
-                       self.pool.get(d_model).unlink(cr, self.uid, ids)
-                       #self.pool.get('ir.model.data')._unlink(cr, self.uid, d_model, ids, direct=True)
-               return False
-
-       def _tag_report(self, cr, rec, data_node=None):
-               res = {}
-               for dest,f in (('name','string'),('model','model'),('report_name','name')):
-                       res[dest] = rec.getAttribute(f).encode('utf8')
-                       assert res[dest], "Attribute %s of report is empty !" % (f,)
-               for field,dest in (('rml','report_rml'),('xml','report_xml'),('xsl','report_xsl')):
-                       if rec.hasAttribute(field):
-                               res[dest] = rec.getAttribute(field).encode('utf8')
-               if rec.hasAttribute('auto'):
-                       res['auto'] = eval(rec.getAttribute('auto'))
-               if rec.hasAttribute('sxw'):
-                       sxw_content = misc.file_open(rec.getAttribute('sxw')).read()
-                       res['report_sxw_content'] = sxw_content
-               if rec.hasAttribute('header'):
-                       res['header'] = eval(rec.getAttribute('header'))
-               res['multi'] = rec.hasAttribute('multi') and  eval(rec.getAttribute('multi'))
-               xml_id = rec.getAttribute('id').encode('utf8')
-               self._test_xml_id(xml_id)
-               id = self.pool.get('ir.model.data')._update(cr, self.uid, "ir.actions.report.xml", self.module, res, xml_id, mode=self.mode)
-               self.idref[xml_id] = int(id)
-               if not rec.hasAttribute('menu') or eval(rec.getAttribute('menu')):
-                       keyword = str(rec.getAttribute('keyword') or 'client_print_multi')
-                       keys = [('action',keyword),('res_model',res['model'])]
-                       value = 'ir.actions.report.xml,'+str(id)
-                       replace = rec.hasAttribute('replace') and rec.getAttribute("replace")
-                       self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', keyword, res['name'], [res['model']], value, replace=replace, isobject=True, xml_id=xml_id)
-               return False
-
-       def _tag_function(self, cr, rec, data_node=None):
-               if self.isnoupdate(data_node) and self.mode != 'init':
-                       return
-               context = self.get_context(data_node, rec, {'ref': _ref(self, cr)})
-               uid = self.get_uid(cr, self.uid, data_node, rec)
-               _eval_xml(self,rec, self.pool, cr, uid, self.idref, context=context)
-               return False
-
-       def _tag_wizard(self, cr, rec, data_node=None):
-               string = rec.getAttribute("string").encode('utf8')
-               model = rec.getAttribute("model").encode('utf8')
-               name = rec.getAttribute("name").encode('utf8')
-               xml_id = rec.getAttribute('id').encode('utf8')
-               self._test_xml_id(xml_id)
-               multi = rec.hasAttribute('multi') and  eval(rec.getAttribute('multi'))
-               res = {'name': string, 'wiz_name': name, 'multi':multi}
-
-               id = self.pool.get('ir.model.data')._update(cr, self.uid, "ir.actions.wizard", self.module, res, xml_id, mode=self.mode)
-               self.idref[xml_id] = int(id)
-               # ir_set
-               if (not rec.hasAttribute('menu') or eval(rec.getAttribute('menu'))) and id:
-                       keyword = str(rec.getAttribute('keyword') or 'client_action_multi')
-                       keys = [('action',keyword),('res_model',model)]
-                       value = 'ir.actions.wizard,'+str(id)
-                       replace = rec.hasAttribute('replace') and \
-                                       rec.getAttribute("replace") or True
-                       self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', keyword, string, [model], value, replace=replace, isobject=True, xml_id=xml_id)
-               return False
-
-       def _tag_url(self, cr, rec, data_node=None):
-               url = rec.getAttribute("string").encode('utf8')
-               target = rec.getAttribute("target").encode('utf8')
-               name = rec.getAttribute("name").encode('utf8')
-               xml_id = rec.getAttribute('id').encode('utf8')
-               self._test_xml_id(xml_id)
-
-               res = {'name': name, 'url': url, 'target':target}
-
-               id = self.pool.get('ir.model.data')._update(cr, self.uid, "ir.actions.url", self.module, res, xml_id, mode=self.mode)
-               self.idref[xml_id] = int(id)
-               # ir_set
-               if (not rec.hasAttribute('menu') or eval(rec.getAttribute('menu'))) and id:
-                       keyword = str(rec.getAttribute('keyword') or 'client_action_multi')
-                       keys = [('action',keyword)]
-                       value = 'ir.actions.url,'+str(id)
-                       replace = rec.hasAttribute('replace') and \
-                                       rec.getAttribute("replace") or True
-                       self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', keyword, url, ["ir.actions.url"], value, replace=replace, isobject=True, xml_id=xml_id)
-               return False
-
-       def _tag_act_window(self, cr, rec, data_node=None):
-               name = rec.hasAttribute('name') and rec.getAttribute('name').encode('utf-8')
-               xml_id = rec.getAttribute('id').encode('utf8')
-               self._test_xml_id(xml_id)
-               type = rec.hasAttribute('type') and rec.getAttribute('type').encode('utf-8') or 'ir.actions.act_window'
-               view_id = False
-               if rec.hasAttribute('view'):
-                       view_id = self.id_get(cr, 'ir.actions.act_window', rec.getAttribute('view').encode('utf-8'))
-               domain = rec.hasAttribute('domain') and rec.getAttribute('domain').encode('utf-8')
-               context = rec.hasAttribute('context') and rec.getAttribute('context').encode('utf-8') or '{}'
-               res_model = rec.getAttribute('res_model').encode('utf-8')
-               src_model = rec.hasAttribute('src_model') and rec.getAttribute('src_model').encode('utf-8')
-               view_type = rec.hasAttribute('view_type') and rec.getAttribute('view_type').encode('utf-8') or 'form'
-               view_mode = rec.hasAttribute('view_mode') and rec.getAttribute('view_mode').encode('utf-8') or 'tree,form'
-               usage = rec.hasAttribute('usage') and rec.getAttribute('usage').encode('utf-8')
-               limit = rec.hasAttribute('limit') and rec.getAttribute('limit').encode('utf-8')
-               auto_refresh = rec.hasAttribute('auto_refresh') \
-                               and rec.getAttribute('auto_refresh').encode('utf-8')
-
-               res = {
-                       'name': name,
-                       'type': type,
-                       'view_id': view_id,
-                       'domain': domain,
-                       'context': context,
-                       'res_model': res_model,
-                       'src_model': src_model,
-                       'view_type': view_type,
-                       'view_mode': view_mode,
-                       'usage': usage,
-                       'limit': limit,
-                       'auto_refresh': auto_refresh,
-               }
-               if rec.hasAttribute('target'):
-                       res['target'] = rec.getAttribute('target')
-               id = self.pool.get('ir.model.data')._update(cr, self.uid, 'ir.actions.act_window', self.module, res, xml_id, mode=self.mode)
-               self.idref[xml_id] = int(id)
-
-               if src_model:
-                       keyword = 'client_action_relate'
-                       keys = [('action', keyword), ('res_model', res_model)]
-                       value = 'ir.actions.act_window,'+str(id)
-                       replace = rec.hasAttribute('replace') and rec.getAttribute('replace')
-                       self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', keyword, xml_id, [src_model], value, replace=replace, isobject=True, xml_id=xml_id)
-               # TODO add remove ir.model.data
-               return False
-
-       def _tag_ir_set(self, cr, rec, data_node=None):
-               if not self.mode=='init':
-                       return False
-               res = {}
-               for field in [i for i in rec.childNodes if (i.nodeType == i.ELEMENT_NODE and i.nodeName=="field")]:
-                       f_name = field.getAttribute("name").encode('utf-8')
-                       f_val = _eval_xml(self,field,self.pool, cr, self.uid, self.idref)
-                       res[f_name] = f_val
-               self.pool.get('ir.model.data').ir_set(cr, self.uid, res['key'], res['key2'], res['name'], res['models'], res['value'], replace=res.get('replace',True), isobject=res.get('isobject', False), meta=res.get('meta',None))
-               return False
-
-       def _tag_workflow(self, cr, rec, data_node=None):
-               if self.isnoupdate(data_node) and self.mode != 'init':
-                       return
-               model = str(rec.getAttribute('model'))
-               w_ref = rec.getAttribute('ref')
-               if len(w_ref):
-                       id = self.id_get(cr, model, w_ref)
-               else:
-                       assert rec.childNodes, 'You must define a child node if you dont give a ref'
-                       element_childs = [i for i in rec.childNodes if i.nodeType == i.ELEMENT_NODE]
-                       assert len(element_childs) == 1, 'Only one child node is accepted (%d given)' % len(rec.childNodes)
-                       id = _eval_xml(self, element_childs[0], self.pool, cr, self.uid, self.idref)
-
-               uid = self.get_uid(cr, self.uid, data_node, rec)
-               wf_service = netsvc.LocalService("workflow")
-               wf_service.trg_validate(uid, model,
-                       id,
-                       str(rec.getAttribute('action')), cr)
-               return False
-
-       #
-       # Support two types of notation:
-       #   name="Inventory Control/Sending Goods"
-       # or
-       #   action="action_id"
-       #   parent="parent_id"
-       #
-       def _tag_menuitem(self, cr, rec, data_node=None):
-               rec_id = rec.getAttribute("id").encode('ascii')
-               self._test_xml_id(rec_id)
-               m_l = map(escape, escape_re.split(rec.getAttribute("name").encode('utf8')))
-
-               values = {'parent_id': False}
-               if not rec.hasAttribute('parent'):
-                       pid = False
-                       for idx, menu_elem in enumerate(m_l):
-                               if pid:
-                                       cr.execute('select id from ir_ui_menu where parent_id=%d and name=%s', (pid, menu_elem))
-                               else:
-                                       cr.execute('select id from ir_ui_menu where parent_id is null and name=%s', (menu_elem,))
-                               res = cr.fetchone()
-                               if idx==len(m_l)-1:
-                                       values = {'parent_id': pid,'name':menu_elem}
-                               elif res:
-                                       pid = res[0]
-                                       xml_id = idx==len(m_l)-1 and rec.getAttribute('id').encode('utf8')
-                                       try:
-                                               npid = self.pool.get('ir.model.data')._update_dummy(cr, self.uid, 'ir.ui.menu', self.module, xml_id, idx==len(m_l)-1)
-                                       except:
-                                               print 'Menu Error', self.module, xml_id, idx==len(m_l)-1
-                               else:
-                                       # the menuitem does't exist but we are in branch (not a leaf)
-                                       self.logger.notifyChannel("init", netsvc.LOG_INFO, 'Warning no ID for submenu %s of menu %s !' % (menu_elem, str(m_l)))
-                                       pid = self.pool.get('ir.ui.menu').create(cr, self.uid, {'parent_id' : pid, 'name' : menu_elem})
-               else:
-                       menu_parent_id = self.id_get(cr, 'ir.ui.menu', rec.getAttribute('parent'))
-                       values = {'parent_id': menu_parent_id}
-                       if rec.hasAttribute('name'):
-                               values['name'] = rec.getAttribute('name')
-                       try:
-                               res = [ self.id_get(cr, 'ir.ui.menu', rec.getAttribute('id')) ]
-                       except:
-                               res = None
-
-               if rec.hasAttribute('action'):
-                       a_action = rec.getAttribute('action').encode('utf8')
-                       a_type = rec.getAttribute('type').encode('utf8') or 'act_window'
-                       icons = {
-                               "act_window": 'STOCK_NEW',
-                               "report.xml": 'STOCK_PASTE',
-                               "wizard": 'STOCK_EXECUTE',
-                               "url": 'STOCK_JUMP_TO'
-                       }
-                       values['icon'] = icons.get(a_type,'STOCK_NEW')
-                       if a_type=='act_window':
-                               a_id = self.id_get(cr, 'ir.actions.%s'% a_type, a_action)
-                               cr.execute('select view_type,view_mode,name,view_id,target from ir_act_window where id=%d', (int(a_id),))
-                               action_type,action_mode,action_name,view_id,target = cr.fetchone()
-                               if view_id:
-                                       cr.execute('SELECT type FROM ir_ui_view WHERE id=%d', (int(view_id),))
-                                       action_mode, = cr.fetchone()
-                               cr.execute('SELECT view_mode FROM ir_act_window_view WHERE act_window_id=%d ORDER BY sequence LIMIT 1', (int(a_id),))
-                               if cr.rowcount:
-                                       action_mode, = cr.fetchone()
-                               if action_type=='tree':
-                                       values['icon'] = 'STOCK_INDENT'
-                               elif action_mode and action_mode.startswith('tree'):
-                                       values['icon'] = 'STOCK_JUSTIFY_FILL'
-                               elif action_mode and action_mode.startswith('graph'):
-                                       values['icon'] = 'terp-graph'
-                               elif action_mode and action_mode.startswith('calendar'):
-                                       values['icon'] = 'terp-calendar'
-                               if target=='new':
-                                       values['icon'] = 'STOCK_EXECUTE'
-                               if not values.get('name', False):
-                                       values['name'] = action_name
-                       elif a_type=='act_wizard':
-                               a_id = self.id_get(cr, 'ir.actions.%s'% a_type, a_action)
-                               cr.execute('select name from ir_act_wizard where id=%d', (int(a_id),))
-                               resw = cr.fetchone()
-                               if (not values.get('name', False)) and resw:
-                                       values['name'] = resw[0]
-               if rec.hasAttribute('sequence'):
-                       values['sequence'] = int(rec.getAttribute('sequence'))
-               if rec.hasAttribute('icon'):
-                       values['icon'] = str(rec.getAttribute('icon'))
-               if rec.hasAttribute('groups'):
-                       g_names = rec.getAttribute('groups').split(',')
-                       groups_value = []
-                       groups_obj = self.pool.get('res.groups')
-                       for group in g_names:
-                               if group.startswith('-'):
-                                       group_id = self.id_get(cr, 'res.groups', group[1:])
-                                       groups_value.append((3, group_id))
-                               else:
-                                       group_id = self.id_get(cr, 'res.groups', group)
-                                       groups_value.append((4, group_id))
-                       values['groups_id'] = groups_value
-
-               xml_id = rec.getAttribute('id').encode('utf8')
-               self._test_xml_id(xml_id)
-               pid = self.pool.get('ir.model.data')._update(cr, self.uid, 'ir.ui.menu', self.module, values, xml_id, True, mode=self.mode, res_id=res and res[0] or False)
-
-               if rec_id and pid:
-                       self.idref[rec_id] = int(pid)
-
-               if rec.hasAttribute('action') and pid:
-                       a_action = rec.getAttribute('action').encode('utf8')
-                       a_type = rec.getAttribute('type').encode('utf8') or 'act_window'
-                       a_id = self.id_get(cr, 'ir.actions.%s' % a_type, a_action)
-                       action = "ir.actions.%s,%d" % (a_type, a_id)
-                       self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', 'tree_but_open', 'Menuitem', [('ir.ui.menu', int(pid))], action, True, True, xml_id=rec_id)
-               return ('ir.ui.menu', pid)
-
-       def _assert_equals(self, f1, f2, prec = 4):
-               return not round(f1 - f2, prec)
-
-       def _tag_assert(self, cr, rec, data_node=None):
-               if self.isnoupdate(data_node) and self.mode != 'init':
-                       return
-
-               rec_model = rec.getAttribute("model").encode('ascii')
-               model = self.pool.get(rec_model)
-               assert model, "The model %s does not exist !" % (rec_model,)
-               rec_id = rec.getAttribute("id").encode('ascii')
-               self._test_xml_id(rec_id)
-               rec_src = rec.getAttribute("search").encode('utf8')
-               rec_src_count = rec.getAttribute("count")
-
-               severity = rec.getAttribute("severity").encode('ascii') or 'info'
-
-               rec_string = rec.getAttribute("string").encode('utf8') or 'unknown'
-
-               ids = None
-               eval_dict = {'ref': _ref(self, cr)}
-               context = self.get_context(data_node, rec, eval_dict)
-               uid = self.get_uid(cr, self.uid, data_node, rec)
-               if len(rec_id):
-                       ids = [self.id_get(cr, rec_model, rec_id)]
-               elif len(rec_src):
-                       q = eval(rec_src, eval_dict)
-                       ids = self.pool.get(rec_model).search(cr, uid, q, context=context)
-                       if len(rec_src_count):
-                               count = int(rec_src_count)
-                               if len(ids) != count:
-                                       self.assert_report.record_assertion(False, severity)
-                                       self.logger.notifyChannel('init', severity, 'assertion "' + rec_string + '" failed ! (search count is incorrect: ' + str(len(ids)) + ')' )
-                                       sevval = getattr(logging, severity.upper())
-                                       if sevval > config['assert_exit_level']:
-                                               # TODO: define a dedicated exception
-                                               raise Exception('Severe assertion failure')
-                                       return
-
-               assert ids != None, 'You must give either an id or a search criteria'
-
-               ref = _ref(self, cr)
-               for id in ids:
-                       brrec =  model.browse(cr, uid, id, context)
-                       class d(dict):
-                               def __getitem__(self2, key):
-                                       if key in brrec:
-                                               return brrec[key]
-                                       return dict.__getitem__(self2, key)
-                       globals = d()
-                       globals['floatEqual'] = self._assert_equals
-                       globals['ref'] = ref
-                       globals['_ref'] = ref
-                       for test in [i for i in rec.childNodes if (i.nodeType == i.ELEMENT_NODE and i.nodeName=="test")]:
-                               f_expr = test.getAttribute("expr").encode('utf-8')
-                               f_val = _eval_xml(self, test, self.pool, cr, uid, self.idref, context=context) or True
-                               if eval(f_expr, globals) != f_val: # assertion failed
-                                       self.assert_report.record_assertion(False, severity)
-                                       self.logger.notifyChannel('init', severity, 'assertion "' + rec_string + '" failed ! (tag ' + test.toxml() + ')' )
-                                       sevval = getattr(logging, severity.upper())
-                                       if sevval > config['assert_exit_level']:
-                                               # TODO: define a dedicated exception
-                                               raise Exception('Severe assertion failure')
-                                       return
-               else: # all tests were successful for this assertion tag (no break)
-                       self.assert_report.record_assertion(True, severity)
-
-       def _tag_record(self, cr, rec, data_node=None):
-               rec_model = rec.getAttribute("model").encode('ascii')
-               model = self.pool.get(rec_model)
-               assert model, "The model %s does not exist !" % (rec_model,)
-               rec_id = rec.getAttribute("id").encode('ascii')
-               self._test_xml_id(rec_id)
-
-#              if not rec_id and not self.isnoupdate(data_node):
-#                      print "Warning", rec_model
-
-               if self.isnoupdate(data_node) and not self.mode == 'init':
-                       # check if the xml record has an id string
-                       if rec_id:
-                               id = self.pool.get('ir.model.data')._update_dummy(cr, self.uid, rec_model, self.module, rec_id)
-                               # check if the resource already existed at the last update
-                               if id:
-                                       # if it existed, we don't update the data, but we need to
-                                       # know the id of the existing record anyway
-                                       self.idref[rec_id] = int(id)
-                                       return None
-                               else:
-                                       # if the resource didn't exist
-                                       if rec.getAttribute("forcecreate"):
-                                               # we want to create it, so we let the normal "update" behavior happen
-                                               pass
-                                       else:
-                                               # otherwise do nothing
-                                               return None
-                       else:
-                               # otherwise it is skipped
-                               return None
-
-               res = {}
-               for field in [i for i in rec.childNodes if (i.nodeType == i.ELEMENT_NODE and i.nodeName=="field")]:
-#TODO: most of this code is duplicated above (in _eval_xml)...
-                       f_name = field.getAttribute("name").encode('utf-8')
-                       f_ref = field.getAttribute("ref").encode('ascii')
-                       f_search = field.getAttribute("search").encode('utf-8')
-                       f_model = field.getAttribute("model").encode('ascii')
-                       if not f_model and model._columns.get(f_name,False):
-                               f_model = model._columns[f_name]._obj
-                       f_use = field.getAttribute("use").encode('ascii') or 'id'
-                       f_val = False
-
-                       if len(f_search):
-                               q = eval(f_search, self.idref)
-                               field = []
-                               assert f_model, 'Define an attribute model="..." in your .XML file !'
-                               f_obj = self.pool.get(f_model)
-                               # browse the objects searched
-                               s = f_obj.browse(cr, self.uid, f_obj.search(cr, self.uid, q))
-                               # column definitions of the "local" object
-                               _cols = self.pool.get(rec_model)._columns
-                               # if the current field is many2many
-                               if (f_name in _cols) and _cols[f_name]._type=='many2many':
-                                       f_val = [(6, 0, map(lambda x: x[f_use], s))]
-                               elif len(s):
-                                       # otherwise (we are probably in a many2one field),
-                                       # take the first element of the search
-                                       f_val = s[0][f_use]
-                       elif len(f_ref):
-                               if f_ref=="null":
-                                       f_val = False
-                               else:
-                                       f_val = self.id_get(cr, f_model, f_ref)
-                       else:
-                               f_val = _eval_xml(self,field, self.pool, cr, self.uid, self.idref)
-                               if model._columns.has_key(f_name):
-                                       if isinstance(model._columns[f_name], osv.fields.integer):
-                                               f_val = int(f_val)
-                       res[f_name] = f_val
-               id = self.pool.get('ir.model.data')._update(cr, self.uid, rec_model, self.module, res, rec_id or False, not self.isnoupdate(data_node), noupdate=self.isnoupdate(data_node), mode=self.mode )
-               if rec_id:
-                       self.idref[rec_id] = int(id)
-               if config.get('commit_mode', False):
-                       cr.commit()
-               return rec_model, id
-
-       def id_get(self, cr, model, id_str):
-               if id_str in self.idref:
-                       return self.idref[id_str]
-               mod = self.module
-               if '.' in id_str:
-                       mod,id_str = id_str.split('.')
-               result = self.pool.get('ir.model.data')._get_id(cr, self.uid, mod, id_str)
-               return int(self.pool.get('ir.model.data').read(cr, self.uid, [result], ['res_id'])[0]['res_id'])
-
-       def parse(self, xmlstr):
-               d = xml.dom.minidom.parseString(xmlstr)
-               de = d.documentElement
-               for n in [i for i in de.childNodes if (i.nodeType == i.ELEMENT_NODE and i.nodeName=="data")]:
-                       for rec in n.childNodes:
-                               if rec.nodeType == rec.ELEMENT_NODE:
-                                       if rec.nodeName in self._tags:
-                                               try:
-                                                       self._tags[rec.nodeName](self.cr, rec, n)
-                                               except:
-                                                       self.logger.notifyChannel("init", netsvc.LOG_INFO, '\n'+rec.toxml())
-                                                       self.cr.rollback()
-                                                       raise
-               return True
-
-       def __init__(self, cr, module, idref, mode, report=assertion_report(), noupdate = False):
-               self.logger = netsvc.Logger()
-               self.mode = mode
-               self.module = module
-               self.cr = cr
-               self.idref = idref
-               self.pool = pooler.get_pool(cr.dbname)
-#              self.pool = osv.osv.FakePool(module)
-               self.uid = 1
-               self.assert_report = report
-               self.noupdate = noupdate
-               self._tags = {
-                       'menuitem': self._tag_menuitem,
-                       'record': self._tag_record,
-                       'assert': self._tag_assert,
-                       'report': self._tag_report,
-                       'wizard': self._tag_wizard,
-                       'delete': self._tag_delete,
-                       'ir_set': self._tag_ir_set,
-                       'function': self._tag_function,
-                       'workflow': self._tag_workflow,
-                       'act_window': self._tag_act_window,
-                       'url': self._tag_url
-               }
+            if module != self.module:
+                modcnt = self.pool.get('ir.module.module').search_count(self.cr, self.uid, ['&', ('name', '=', module), ('state', 'in', ['installed'])])
+                assert modcnt == 1, """The ID "%s" refer to an uninstalled module""" % (xml_id,)
+
+        if len(id) > 64:
+            self.logger.notifyChannel('init', netsvc.LOG_ERROR, 'id: %s is to long (max: 64)'% (id,))
+
+    def _tag_delete(self, cr, rec, data_node=None):
+        d_model = rec.get("model",'')
+        d_search = rec.get("search",'').encode('utf-8')
+        d_id = rec.get("id",'')
+        ids = []
+        
+        if d_search:
+            idref = _get_idref(self, cr, self.uid, d_model, context={}, idref={})
+            ids = self.pool.get(d_model).search(cr,self.uid, unsafe_eval(d_search, idref))
+        if d_id:
+            try:
+                ids.append(self.id_get(cr, d_model, d_id))
+            except:
+                # d_id cannot be found. doesn't matter in this case
+                pass
+        if ids:
+            self.pool.get(d_model).unlink(cr, self.uid, ids)
+            self.pool.get('ir.model.data')._unlink(cr, self.uid, d_model, ids, direct=True)
+    
+    def _remove_ir_values(self, cr, name, value, model):
+        ir_value_ids = self.pool.get('ir.values').search(cr, self.uid, [('name','=',name),('value','=',value),('model','=',model)])
+        if ir_value_ids:
+            self.pool.get('ir.values').unlink(cr, self.uid, ir_value_ids)
+            self.pool.get('ir.model.data')._unlink(cr, self.uid, 'ir.values', ir_value_ids, direct=True)
+        
+        return True
+     
+    def _tag_report(self, cr, rec, data_node=None):
+        res = {}
+        for dest,f in (('name','string'),('model','model'),('report_name','name')):
+            res[dest] = rec.get(f,'').encode('utf8')
+            assert res[dest], "Attribute %s of report is empty !" % (f,)
+        for field,dest in (('rml','report_rml'),('xml','report_xml'),('xsl','report_xsl'),('attachment','attachment'),('attachment_use','attachment_use')):
+            if rec.get(field):
+                res[dest] = rec.get(field).encode('utf8')
+        if rec.get('auto'):
+            res['auto'] = eval(rec.get('auto','False'))
+        if rec.get('sxw'):
+            sxw_content = misc.file_open(rec.get('sxw')).read()
+            res['report_sxw_content'] = sxw_content
+        if rec.get('header'):
+            res['header'] = eval(rec.get('header','False'))
+        res['multi'] = rec.get('multi') and eval(rec.get('multi','False'))
+        xml_id = rec.get('id','').encode('utf8')
+        self._test_xml_id(xml_id)
+
+        if rec.get('groups'):
+            g_names = rec.get('groups','').split(',')
+            groups_value = []
+            groups_obj = self.pool.get('res.groups')
+            for group in g_names:
+                if group.startswith('-'):
+                    group_id = self.id_get(cr, 'res.groups', group[1:])
+                    groups_value.append((3, group_id))
+                else:
+                    group_id = self.id_get(cr, 'res.groups', group)
+                    groups_value.append((4, group_id))
+            res['groups_id'] = groups_value
+
+        id = self.pool.get('ir.model.data')._update(cr, self.uid, "ir.actions.report.xml", self.module, res, xml_id, noupdate=self.isnoupdate(data_node), mode=self.mode)
+        self.idref[xml_id] = int(id)
+
+        if not rec.get('menu') or eval(rec.get('menu','False')):
+            keyword = str(rec.get('keyword', 'client_print_multi'))
+            keys = [('action',keyword),('res_model',res['model'])]
+            value = 'ir.actions.report.xml,'+str(id)
+            replace = rec.get('replace', True)
+            self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', keyword, res['name'], [res['model']], value, replace=replace, isobject=True, xml_id=xml_id)
+        elif self.mode=='update' and eval(rec.get('menu','False'))==False:
+            # Special check for report having attribute menu=False on update
+            value = 'ir.actions.report.xml,'+str(id)
+            self._remove_ir_values(cr, res['name'], value, res['model'])
+        return False
+
+    def _tag_function(self, cr, rec, data_node=None):
+        if self.isnoupdate(data_node) and self.mode != 'init':
+            return
+        context = self.get_context(data_node, rec, {'ref': _ref(self, cr)})
+        uid = self.get_uid(cr, self.uid, data_node, rec)
+        _eval_xml(self,rec, self.pool, cr, uid, self.idref, context=context)
+        return
+
+    def _tag_wizard(self, cr, rec, data_node=None):
+        string = rec.get("string",'').encode('utf8')
+        model = rec.get("model",'').encode('utf8')
+        name = rec.get("name",'').encode('utf8')
+        xml_id = rec.get('id','').encode('utf8')
+        self._test_xml_id(xml_id)
+        multi = rec.get('multi','') and eval(rec.get('multi','False'))
+        res = {'name': string, 'wiz_name': name, 'multi': multi, 'model': model}
+
+        if rec.get('groups'):
+            g_names = rec.get('groups','').split(',')
+            groups_value = []
+            groups_obj = self.pool.get('res.groups')
+            for group in g_names:
+                if group.startswith('-'):
+                    group_id = self.id_get(cr, 'res.groups', group[1:])
+                    groups_value.append((3, group_id))
+                else:
+                    group_id = self.id_get(cr, 'res.groups', group)
+                    groups_value.append((4, group_id))
+            res['groups_id'] = groups_value
+
+        id = self.pool.get('ir.model.data')._update(cr, self.uid, "ir.actions.wizard", self.module, res, xml_id, noupdate=self.isnoupdate(data_node), mode=self.mode)
+        self.idref[xml_id] = int(id)
+        # ir_set
+        if (not rec.get('menu') or eval(rec.get('menu','False'))) and id:
+            keyword = str(rec.get('keyword','') or 'client_action_multi')
+            keys = [('action',keyword),('res_model',model)]
+            value = 'ir.actions.wizard,'+str(id)
+            replace = rec.get("replace",'') or True
+            self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', keyword, string, [model], value, replace=replace, isobject=True, xml_id=xml_id)
+        elif self.mode=='update' and (rec.get('menu') and eval(rec.get('menu','False'))==False):
+            # Special check for wizard having attribute menu=False on update
+            value = 'ir.actions.wizard,'+str(id)
+            self._remove_ir_values(cr, string, value, model)
+        
+    def _tag_url(self, cr, rec, data_node=None):
+        url = rec.get("string",'').encode('utf8')
+        target = rec.get("target",'').encode('utf8')
+        name = rec.get("name",'').encode('utf8')
+        xml_id = rec.get('id','').encode('utf8')
+        self._test_xml_id(xml_id)
+
+        res = {'name': name, 'url': url, 'target':target}
+
+        id = self.pool.get('ir.model.data')._update(cr, self.uid, "ir.actions.url", self.module, res, xml_id, noupdate=self.isnoupdate(data_node), mode=self.mode)
+        self.idref[xml_id] = int(id)
+        # ir_set
+        if (not rec.get('menu') or eval(rec.get('menu','False'))) and id:
+            keyword = str(rec.get('keyword','') or 'client_action_multi')
+            keys = [('action',keyword)]
+            value = 'ir.actions.url,'+str(id)
+            replace = rec.get("replace",'') or True
+            self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', keyword, url, ["ir.actions.url"], value, replace=replace, isobject=True, xml_id=xml_id)
+        elif self.mode=='update' and (rec.get('menu') and eval(rec.get('menu','False'))==False):
+            # Special check for URL having attribute menu=False on update
+            value = 'ir.actions.url,'+str(id)
+            self._remove_ir_values(cr, url, value, "ir.actions.url")
+
+    def _tag_act_window(self, cr, rec, data_node=None):
+        name = rec.get('name','').encode('utf-8')
+        xml_id = rec.get('id','').encode('utf8')
+        self._test_xml_id(xml_id)
+        type = rec.get('type','').encode('utf-8') or 'ir.actions.act_window'
+        view_id = False
+        if rec.get('view'):
+            view_id = self.id_get(cr, 'ir.actions.act_window', rec.get('view','').encode('utf-8'))
+        domain = rec.get('domain','').encode('utf-8') or '[]'
+        context = rec.get('context','').encode('utf-8') or '{}'
+        res_model = rec.get('res_model','').encode('utf-8')
+        src_model = rec.get('src_model','').encode('utf-8')
+        view_type = rec.get('view_type','').encode('utf-8') or 'form'
+        view_mode = rec.get('view_mode','').encode('utf-8') or 'tree,form'
+        usage = rec.get('usage','').encode('utf-8')
+        limit = rec.get('limit','').encode('utf-8')
+        auto_refresh = rec.get('auto_refresh','').encode('utf-8')
+
+        # def ref() added because , if context has ref('id') eval wil use this ref
+
+        active_id=str("active_id") # for further reference in client/bin/tools/__init__.py
+
+        def ref(str_id):
+            return self.id_get(cr, None, str_id)
+        context=eval(context, {}, {'active_id' : active_id})
+        #domain=unsafe_eval(domain) # XXX need to test this line -> uid, active_id, active_ids, ...
+
+        res = {
+            'name': name,
+            'type': type,
+            'view_id': view_id,
+            'domain': domain,
+            'context': context,
+            'res_model': res_model,
+            'src_model': src_model,
+            'view_type': view_type,
+            'view_mode': view_mode,
+            'usage': usage,
+            'limit': limit,
+            'auto_refresh': auto_refresh,
+#            'groups_id':groups_id,
+        }
+
+        if rec.get('groups'):
+            g_names = rec.get('groups','').split(',')
+            groups_value = []
+            groups_obj = self.pool.get('res.groups')
+            for group in g_names:
+                if group.startswith('-'):
+                    group_id = self.id_get(cr, 'res.groups', group[1:])
+                    groups_value.append((3, group_id))
+                else:
+                    group_id = self.id_get(cr, 'res.groups', group)
+                    groups_value.append((4, group_id))
+            res['groups_id'] = groups_value
+
+        if rec.get('target'):
+            res['target'] = rec.get('target','')
+        id = self.pool.get('ir.model.data')._update(cr, self.uid, 'ir.actions.act_window', self.module, res, xml_id, noupdate=self.isnoupdate(data_node), mode=self.mode)
+        self.idref[xml_id] = int(id)
+
+        if src_model:
+            keyword = 'client_action_relate'
+            keys = [('action', keyword), ('res_model', res_model)]
+            value = 'ir.actions.act_window,'+str(id)
+            replace = rec.get('replace','') or True
+            self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', keyword, xml_id, [src_model], value, replace=replace, isobject=True, xml_id=xml_id)
+        # TODO add remove ir.model.data
+
+    def _tag_ir_set(self, cr, rec, data_node=None):
+        if self.mode != 'init':
+            return
+        res = {}
+        for field in rec.findall('./field'):
+            f_name = field.get("name",'').encode('utf-8')
+            f_val = _eval_xml(self,field,self.pool, cr, self.uid, self.idref)
+            res[f_name] = f_val
+        self.pool.get('ir.model.data').ir_set(cr, self.uid, res['key'], res['key2'], res['name'], res['models'], res['value'], replace=res.get('replace',True), isobject=res.get('isobject', False), meta=res.get('meta',None))
+
+    def _tag_workflow(self, cr, rec, data_node=None):
+        if self.isnoupdate(data_node) and self.mode != 'init':
+            return
+        model = str(rec.get('model',''))
+        w_ref = rec.get('ref','')
+        if w_ref:
+            id = self.id_get(cr, model, w_ref)
+        else:
+            number_children = len(rec)
+            assert number_children > 0,\
+                'You must define a child node if you dont give a ref'
+            assert number_children == 1,\
+                'Only one child node is accepted (%d given)' % number_children
+            id = _eval_xml(self, rec[0], self.pool, cr, self.uid, self.idref)
+
+        uid = self.get_uid(cr, self.uid, data_node, rec)
+        wf_service = netsvc.LocalService("workflow")
+        wf_service.trg_validate(uid, model,
+            id,
+            str(rec.get('action','')), cr)
+
+    #
+    # Support two types of notation:
+    #   name="Inventory Control/Sending Goods"
+    # or
+    #   action="action_id"
+    #   parent="parent_id"
+    #
+    def _tag_menuitem(self, cr, rec, data_node=None):
+        rec_id = rec.get("id",'').encode('ascii')
+        self._test_xml_id(rec_id)
+        m_l = map(escape, escape_re.split(rec.get("name",'').encode('utf8')))
+
+        values = {'parent_id': False}
+        if not rec.get('parent'):
+            pid = False
+            for idx, menu_elem in enumerate(m_l):
+                if pid:
+                    cr.execute('select id from ir_ui_menu where parent_id=%s and name=%s', (pid, menu_elem))
+                else:
+                    cr.execute('select id from ir_ui_menu where parent_id is null and name=%s', (menu_elem,))
+                res = cr.fetchone()
+                if idx==len(m_l)-1:
+                    values = {'parent_id': pid,'name':menu_elem}
+                elif res:
+                    pid = res[0]
+                    xml_id = idx==len(m_l)-1 and rec.get('id','').encode('utf8')
+                    try:
+                        npid = self.pool.get('ir.model.data')._update_dummy(cr, self.uid, 'ir.ui.menu', self.module, xml_id, idx==len(m_l)-1)
+                    except:
+                        self.logger.notifyChannel('init', netsvc.LOG_ERROR, "module: %s xml_id: %s" % (self.module, xml_id))
+                else:
+                    # the menuitem does't exist but we are in branch (not a leaf)
+                    self.logger.notifyChannel("init", netsvc.LOG_WARNING, 'Warning no ID for submenu %s of menu %s !' % (menu_elem, str(m_l)))
+                    pid = self.pool.get('ir.ui.menu').create(cr, self.uid, {'parent_id' : pid, 'name' : menu_elem})
+        else:
+            menu_parent_id = self.id_get(cr, 'ir.ui.menu', rec.get('parent',''))
+            values = {'parent_id': menu_parent_id}
+            if rec.get('name'):
+                values['name'] = rec.get('name')
+            try:
+                res = [ self.id_get(cr, 'ir.ui.menu', rec.get('id','')) ]
+            except:
+                res = None
+
+        if rec.get('action'):
+            a_action = rec.get('action','').encode('utf8')
+            a_type = rec.get('type','').encode('utf8') or 'act_window'
+            icons = {
+                "act_window": 'STOCK_NEW',
+                "report.xml": 'STOCK_PASTE',
+                "wizard": 'STOCK_EXECUTE',
+                "url": 'STOCK_JUMP_TO'
+            }
+            values['icon'] = icons.get(a_type,'STOCK_NEW')
+            if a_type=='act_window':
+                a_id = self.id_get(cr, 'ir.actions.%s'% a_type, a_action)
+                cr.execute('select view_type,view_mode,name,view_id,target from ir_act_window where id=%s', (int(a_id),))
+                rrres = cr.fetchone()
+                assert rrres, "No window action defined for this id %s !\n" \
+                    "Verify that this is a window action or add a type argument." % (a_action,)
+                action_type,action_mode,action_name,view_id,target = rrres
+                if view_id:
+                    cr.execute('SELECT type FROM ir_ui_view WHERE id=%s', (int(view_id),))
+                    action_mode, = cr.fetchone()
+                cr.execute('SELECT view_mode FROM ir_act_window_view WHERE act_window_id=%s ORDER BY sequence LIMIT 1', (int(a_id),))
+                if cr.rowcount:
+                    action_mode, = cr.fetchone()
+                if action_type=='tree':
+                    values['icon'] = 'STOCK_INDENT'
+                elif action_mode and action_mode.startswith('tree'):
+                    values['icon'] = 'STOCK_JUSTIFY_FILL'
+                elif action_mode and action_mode.startswith('graph'):
+                    values['icon'] = 'terp-graph'
+                elif action_mode and action_mode.startswith('calendar'):
+                    values['icon'] = 'terp-calendar'
+                if target=='new':
+                    values['icon'] = 'STOCK_EXECUTE'
+                if not values.get('name', False):
+                    values['name'] = action_name
+            elif a_type=='wizard':
+                a_id = self.id_get(cr, 'ir.actions.%s'% a_type, a_action)
+                cr.execute('select name from ir_act_wizard where id=%s', (int(a_id),))
+                resw = cr.fetchone()
+                if (not values.get('name', False)) and resw:
+                    values['name'] = resw[0]
+        if rec.get('sequence'):
+            values['sequence'] = int(rec.get('sequence'))
+        if rec.get('icon'):
+            values['icon'] = str(rec.get('icon'))
+
+        if rec.get('groups'):
+            g_names = rec.get('groups','').split(',')
+            groups_value = []
+            groups_obj = self.pool.get('res.groups')
+            for group in g_names:
+                if group.startswith('-'):
+                    group_id = self.id_get(cr, 'res.groups', group[1:])
+                    groups_value.append((3, group_id))
+                else:
+                    group_id = self.id_get(cr, 'res.groups', group)
+                    groups_value.append((4, group_id))
+            values['groups_id'] = groups_value
+
+        xml_id = rec.get('id','').encode('utf8')
+        self._test_xml_id(xml_id)
+        pid = self.pool.get('ir.model.data')._update(cr, self.uid, 'ir.ui.menu', self.module, values, xml_id, noupdate=self.isnoupdate(data_node), mode=self.mode, res_id=res and res[0] or False)
+
+        if rec_id and pid:
+            self.idref[rec_id] = int(pid)
+
+        if rec.get('action') and pid:
+            a_action = rec.get('action').encode('utf8')
+            a_type = rec.get('type','').encode('utf8') or 'act_window'
+            a_id = self.id_get(cr, 'ir.actions.%s' % a_type, a_action)
+            action = "ir.actions.%s,%d" % (a_type, a_id)
+            self.pool.get('ir.model.data').ir_set(cr, self.uid, 'action', 'tree_but_open', 'Menuitem', [('ir.ui.menu', int(pid))], action, True, True, xml_id=rec_id)
+        return ('ir.ui.menu', pid)
+
+    def _assert_equals(self, f1, f2, prec = 4):
+        return not round(f1 - f2, prec)
+
+    def _tag_assert(self, cr, rec, data_node=None):
+        if self.isnoupdate(data_node) and self.mode != 'init':
+            return
+
+        rec_model = rec.get("model",'').encode('ascii')
+        model = self.pool.get(rec_model)
+        assert model, "The model %s does not exist !" % (rec_model,)
+        rec_id = rec.get("id",'').encode('ascii')
+        self._test_xml_id(rec_id)
+        rec_src = rec.get("search",'').encode('utf8')
+        rec_src_count = rec.get("count")
+
+        severity = rec.get("severity",'').encode('ascii') or netsvc.LOG_ERROR
+
+        rec_string = rec.get("string",'').encode('utf8') or 'unknown'
+
+        ids = None
+        eval_dict = {'ref': _ref(self, cr)}
+        context = self.get_context(data_node, rec, eval_dict)
+        uid = self.get_uid(cr, self.uid, data_node, rec)
+        if rec_id:
+            ids = [self.id_get(cr, rec_model, rec_id)]
+        elif rec_src:
+            q = unsafe_eval(rec_src, eval_dict)
+            ids = self.pool.get(rec_model).search(cr, uid, q, context=context)
+            if rec_src_count:
+                count = int(rec_src_count)
+                if len(ids) != count:
+                    self.assert_report.record_assertion(False, severity)
+                    msg = 'assertion "%s" failed!\n'    \
+                          ' Incorrect search count:\n'  \
+                          ' expected count: %d\n'       \
+                          ' obtained count: %d\n'       \
+                          % (rec_string, count, len(ids))
+                    self.logger.notifyChannel('init', severity, msg)
+                    sevval = getattr(logging, severity.upper())
+                    if sevval >= config['assert_exit_level']:
+                        # TODO: define a dedicated exception
+                        raise Exception('Severe assertion failure')
+                    return
+
+        assert ids is not None,\
+            'You must give either an id or a search criteria'
+        ref = _ref(self, cr)
+        for id in ids:
+            brrec =  model.browse(cr, uid, id, context)
+            class d(dict):
+                def __getitem__(self2, key):
+                    if key in brrec:
+                        return brrec[key]
+                    return dict.__getitem__(self2, key)
+            globals_dict = d()
+            globals_dict['floatEqual'] = self._assert_equals
+            globals_dict['ref'] = ref
+            globals_dict['_ref'] = ref
+            for test in rec.findall('./test'):
+                f_expr = test.get("expr",'').encode('utf-8')
+                expected_value = _eval_xml(self, test, self.pool, cr, uid, self.idref, context=context) or True
+                expression_value = unsafe_eval(f_expr, globals_dict)
+                if expression_value != expected_value: # assertion failed
+                    self.assert_report.record_assertion(False, severity)
+                    msg = 'assertion "%s" failed!\n'    \
+                          ' xmltag: %s\n'               \
+                          ' expected value: %r\n'       \
+                          ' obtained value: %r\n'       \
+                          % (rec_string, etree.tostring(test), expected_value, expression_value)
+                    self.logger.notifyChannel('init', severity, msg)
+                    sevval = getattr(logging, severity.upper())
+                    if sevval >= config['assert_exit_level']:
+                        # TODO: define a dedicated exception
+                        raise Exception('Severe assertion failure')
+                    return
+        else: # all tests were successful for this assertion tag (no break)
+            self.assert_report.record_assertion(True, severity)
+
+    def _tag_record(self, cr, rec, data_node=None):
+        rec_model = rec.get("model").encode('ascii')
+        model = self.pool.get(rec_model)
+        assert model, "The model %s does not exist !" % (rec_model,)
+        rec_id = rec.get("id",'').encode('ascii')
+        self._test_xml_id(rec_id)
+        if self.isnoupdate(data_node) and self.mode != 'init':
+            # check if the xml record has an id string
+            if rec_id:
+                if '.' in rec_id:
+                    module,rec_id2 = rec_id.split('.')
+                else:
+                    module = self.module
+                    rec_id2 = rec_id
+                id = self.pool.get('ir.model.data')._update_dummy(cr, self.uid, rec_model, module, rec_id2)
+                # check if the resource already existed at the last update
+                if id:
+                    # if it existed, we don't update the data, but we need to
+                    # know the id of the existing record anyway
+                    self.idref[rec_id] = int(id)
+                    return None
+                else:
+                    # if the resource didn't exist
+                    if not self.nodeattr2bool(rec, 'forcecreate', True):
+                        # we don't want to create it, so we skip it
+                        return None
+                    # else, we let the record to be created
+
+            else:
+                # otherwise it is skipped
+                return None
+        res = {}
+        for field in rec.findall('./field'):
+            #TODO: most of this code is duplicated above (in _eval_xml)...
+            f_name = field.get("name",'').encode('utf-8')
+            f_ref = field.get("ref",'').encode('ascii')
+            f_search = field.get("search",'').encode('utf-8')
+            f_model = field.get("model",'').encode('ascii')
+            if not f_model and model._columns.get(f_name,False):
+                f_model = model._columns[f_name]._obj
+            f_use = field.get("use",'').encode('ascii') or 'id'
+            f_val = False
+
+            if f_search:
+                q = unsafe_eval(f_search, self.idref)
+                field = []
+                assert f_model, 'Define an attribute model="..." in your .XML file !'
+                f_obj = self.pool.get(f_model)
+                # browse the objects searched
+                s = f_obj.browse(cr, self.uid, f_obj.search(cr, self.uid, q))
+                # column definitions of the "local" object
+                _cols = self.pool.get(rec_model)._columns
+                # if the current field is many2many
+                if (f_name in _cols) and _cols[f_name]._type=='many2many':
+                    f_val = [(6, 0, map(lambda x: x[f_use], s))]
+                elif len(s):
+                    # otherwise (we are probably in a many2one field),
+                    # take the first element of the search
+                    f_val = s[0][f_use]
+            elif f_ref:
+                if f_ref=="null":
+                    f_val = False
+                else:
+                    f_val = self.id_get(cr, f_model, f_ref)
+            else:
+                f_val = _eval_xml(self,field, self.pool, cr, self.uid, self.idref)
+                if model._columns.has_key(f_name):
+                    if isinstance(model._columns[f_name], osv.fields.integer):
+                        f_val = int(f_val)
+            res[f_name] = f_val
+
+        id = self.pool.get('ir.model.data')._update(cr, self.uid, rec_model, self.module, res, rec_id or False, not self.isnoupdate(data_node), noupdate=self.isnoupdate(data_node), mode=self.mode )
+        if rec_id:
+            self.idref[rec_id] = int(id)
+        if config.get('import_partial', False):
+            cr.commit()
+        return rec_model, id
+
+    def id_get(self, cr, model, id_str):
+        if id_str in self.idref:
+            return self.idref[id_str]
+        mod = self.module
+        if '.' in id_str:
+            mod,id_str = id_str.split('.')
+        result = self.pool.get('ir.model.data')._get_id(cr, self.uid, mod, id_str)
+        return int(self.pool.get('ir.model.data').read(cr, self.uid, [result], ['res_id'])[0]['res_id'])
+
+    def parse(self, de):
+        if not de.tag in ['terp', 'openerp']:
+            self.logger.notifyChannel("init", netsvc.LOG_ERROR, "Mismatch xml format" )
+            raise Exception( "Mismatch xml format: only terp or openerp as root tag" )
+
+        if de.tag == 'terp':
+            self.logger.notifyChannel("init", netsvc.LOG_WARNING, "The tag <terp/> is deprecated, use <openerp/>")
+
+        for n in de.findall('./data'):
+            for rec in n:
+                    if rec.tag in self._tags:
+                        try:
+                            self._tags[rec.tag](self.cr, rec, n)
+                        except:
+                            self.logger.notifyChannel("init", netsvc.LOG_ERROR, '\n'+etree.tostring(rec))
+                            self.cr.rollback()
+                            raise
+        return True
+
+    def __init__(self, cr, module, idref, mode, report=None, noupdate=False):
+
+        self.logger = netsvc.Logger()
+        self.mode = mode
+        self.module = module
+        self.cr = cr
+        self.idref = idref
+        self.pool = pooler.get_pool(cr.dbname)
+        self.uid = 1
+        if report is None:
+            report = assertion_report()
+        self.assert_report = report
+        self.noupdate = noupdate
+        self._tags = {
+            'menuitem': self._tag_menuitem,
+            'record': self._tag_record,
+            'assert': self._tag_assert,
+            'report': self._tag_report,
+            'wizard': self._tag_wizard,
+            'delete': self._tag_delete,
+            'ir_set': self._tag_ir_set,
+            'function': self._tag_function,
+            'workflow': self._tag_workflow,
+            'act_window': self._tag_act_window,
+            'url': self._tag_url
+        }
 
 def convert_csv_import(cr, module, fname, csvcontent, idref=None, mode='init',
-               noupdate=False):
-       '''Import csv file :
-               quote: "
-               delimiter: ,
-               encoding: utf-8'''
-       if not idref:
-               idref={}
-       model = ('.'.join(fname.split('.')[:-1]).split('-'))[0]
-       #remove folder path from model
-       head, model = os.path.split(model)
-
-       pool = pooler.get_pool(cr.dbname)
-
-       input = StringIO.StringIO(csvcontent)
-       reader = csv.reader(input, quotechar='"', delimiter=',')
-       fields = reader.next()
-
-       if not (mode == 'init' or 'id' in fields):
-               return
-
-       uid = 1
-       datas = []
-       for line in reader:
-               if (not line) or not reduce(lambda x,y: x or y, line) :
-                       continue
-               datas.append( map(lambda x:x.decode('utf8').encode('utf8'), line))
-       pool.get(model).import_data(cr, uid, fields, datas,mode, module,noupdate)
+        noupdate=False):
+    '''Import csv file :
+        quote: "
+        delimiter: ,
+        encoding: utf-8'''
+    if not idref:
+        idref={}
+    model = ('.'.join(fname.split('.')[:-1]).split('-'))[0]
+    #remove folder path from model
+    head, model = os.path.split(model)
+
+    pool = pooler.get_pool(cr.dbname)
+
+    input = cStringIO.StringIO(csvcontent)
+    reader = csv.reader(input, quotechar='"', delimiter=',')
+    fields = reader.next()
+    fname_partial = ""
+    if config.get('import_partial'):
+        fname_partial = module + '/'+ fname
+        if not os.path.isfile(config.get('import_partial')):
+            pickle.dump({}, file(config.get('import_partial'),'w+'))
+        else:
+            data = pickle.load(file(config.get('import_partial')))
+            if fname_partial in data:
+                if not data[fname_partial]:
+                    return
+                else:
+                    for i in range(data[fname_partial]):
+                        reader.next()
+
+    if not (mode == 'init' or 'id' in fields):
+        return
+
+    uid = 1
+    datas = []
+    for line in reader:
+        if (not line) or not reduce(lambda x,y: x or y, line) :
+            continue
+        try:
+            datas.append(map(lambda x: misc.ustr(x), line))
+        except:
+            logger = netsvc.Logger()
+            logger.notifyChannel("init", netsvc.LOG_ERROR, "Cannot import the line: %s" % line)
+    result = pool.get(model).import_data(cr, uid, fields, datas,mode, module,noupdate,filename=fname_partial)
+    if result[0] == -1:
+        tools.debug(result)
+    if config.get('import_partial'):
+        data = pickle.load(file(config.get('import_partial')))
+        data[fname_partial] = 0
+        pickle.dump(data, file(config.get('import_partial'),'wb'))
+        cr.commit()
 
 #
 # xml import/export
 #
-def convert_xml_import(cr, module, xmlfile, idref=None, mode='init', noupdate = False, report=None):
-       xmlstr = xmlfile.read()
-       xmlfile.seek(0)
-       relaxng_doc = etree.parse(file('import_xml.rng'))
-       relaxng = etree.RelaxNG(relaxng_doc)
-
-       doc = etree.parse(xmlfile)
-       try:
-               relaxng.assert_(doc)
-       except Exception, e:
-               logger = netsvc.Logger()
-               logger.notifyChannel('init', netsvc.LOG_ERROR, 'The XML file do not fit the required schema !')
-               logger.notifyChannel('init', netsvc.LOG_ERROR, relaxng.error_log.last_error)
-               raise
-
-       if not idref:
-               idref={}
-       if report is None:
-               report=assertion_report()
-       obj = xml_import(cr, module, idref, mode, report=report, noupdate = noupdate)
-       obj.parse(xmlstr)
-       del obj
-       return True
+def convert_xml_import(cr, module, xmlfile, idref=None, mode='init', noupdate=False, report=None):
+    doc = etree.parse(xmlfile)
+    relaxng = etree.RelaxNG(
+        etree.parse(os.path.join(config['root_path'],'import_xml.rng' )))
+    try:
+        relaxng.assert_(doc)
+    except Exception, e:
+        logger = netsvc.Logger()
+        logger.notifyChannel('init', netsvc.LOG_ERROR, 'The XML file does not fit the required schema !')
+        logger.notifyChannel('init', netsvc.LOG_ERROR, misc.ustr(relaxng.error_log.last_error))
+        raise
+
+    if idref is None:
+        idref={}
+    obj = xml_import(cr, module, idref, mode, report=report, noupdate=noupdate)
+    obj.parse(doc.getroot())
+    return True
 
 def convert_xml_export(res):
-       uid=1
-       pool=pooler.get_pool(cr.dbname)
-       cr=pooler.db.cursor()
-       idref = {}
-       d = xml.dom.minidom.getDOMImplementation().createDocument(None, "terp", None)
-       de = d.documentElement
-       data=d.createElement("data")
-       de.appendChild(data)
-       de.appendChild(d.createTextNode('Some textual content.'))
-       cr.commit()
-       cr.close()
+    uid=1
+    pool=pooler.get_pool(cr.dbname)
+    cr=pooler.db.cursor()
+    idref = {}
+
+    page = etree.Element ( 'terp' )
+    doc = etree.ElementTree ( page )
+    data = etree.SubElement ( page, 'data' )
+    text_node = etree.SubElement ( page, 'text' )
+    text_node.text = 'Some textual content.'
+    
+    cr.commit()
+    cr.close()
+
+
+# vim:expandtab:smartindent:tabstop=4:softtabstop=4:shiftwidth=4: