changed the ingestion procedure - in some cases some properties are read from...

changed the ingestion procedure - in some cases some properties are read from the content, and sometimes we do not check for consistency

git-svn-id: https://svn.erp5.org/repos/public/erp5/trunk@11253 20353a03-c40f-0410-a6d1-a30d3c3de9de

bt5/erp5_dms/ExtensionTemplateItem/documentUtils.py

@@ -18,8 +18,34 @@ def extractContent(data):
   z.close()
   return s
 
-def getAttrFromContent(data):
-  return dict(rx_atr.findall(extractContent(data)))
+rx_fileno=re.compile('File No.: *(?P<reference>(?P<group>[A-Za-z-]+)-(?P<project>\d+)-(?P<number>\d+)\.(?P<year>\d{2}))')
+rx_classif=re.compile('([A-Z]{1}[a-z]+/[A-Z]{1}[a-z]+)')
+
+def getAttrFromContent(self,data,ptype):
+  if ptype!='Memo':return {}
+  atrs={}
+  fileno=rx_fileno.search(data)
+  if fileno:
+    dic=fileno.groupdict()
+    atrs['source_project']='project_module/'+dic['project']
+    atrs['reference']=dic['reference']
+  classif=rx_classif.search(data)
+  log=[]
+  if classif:
+    classif=classif.groups()[0].split('/')
+    classif.reverse()
+    res=self.portal_catalog(portal_type='Category',title=classif[0])
+    for r in res:
+      c=r.getObject()
+      for x,t in enumerate(classif):
+        c=c.aq_parent
+        if c.getId()=='classification':
+          atrs['classification']='/'.join(r.getRelativeUrl().split('/')[1:])
+          break
+        if c.getTitle()!=classif[x+1]:
+          break
+  self.log(atrs)
+  return atrs
 
 def getDoctypeFromContent(data):
   atrs=getAttrFromContent(data)

bt5/erp5_dms/SkinTemplateItem/portal_skins/erp5_dms/DMS_createObjectFromFile.xml

@@ -70,8 +70,8 @@
             <key> <string>_body</string> </key>
             <value> <string># determine content type if not given\n
 # first try from content\n
-if not doctype:\n
-  doctype=context.Document_getDoctypeFromContent(data)\n
+#if not doctype:\n
+#  doctype=context.Document_getDoctypeFromContent(data)\n
 # then from filename\n
 if not doctype:\n
   doctype=context.content_type_registry.findTypeName(fname,None,None)\n

bt5/erp5_dms/SkinTemplateItem/portal_skins/erp5_dms/DMS_ingestFile.xml

@@ -78,8 +78,8 @@ if data is None:\n
   else:\n
     if hasattr(context,\'data\'):\n
       data=data\n
-if data is not None:\n
-  context.setPropertyListFromContent(data)\n
+#if data is not None:\n
+#  context.setPropertyListFromContent(data)\n
 \n
 if fname is None:\n
   fname=context.getSourceReference()\n

bt5/erp5_dms/SkinTemplateItem/portal_skins/erp5_dms/DMS_notifyByEmail.xml

@@ -76,8 +76,11 @@ if pref is not None and len(pref)==0:\n
   if pref==\'never\':return\n
 \n
 ob=kw[\'object\']\n
-res=ob.checkConsistency()\n
-res=[c for c in res if c[1]==\'DocumentCoordinatesConstraint inconsistency\']\n
+if ob.getPortalType()==\'Memo\': # we don\'t check constistency for Memo XXX use type groups, or what?\n
+  res=\'\'\n
+else:\n
+  res=ob.checkConsistency()\n
+  res=[c for c in res if c[1]==\'DocumentCoordinatesConstraint inconsistency\']\n
 \n
 # shall we send?\n
 send=(pref==\'always\' or len(res)>0)\n

bt5/erp5_dms/SkinTemplateItem/portal_skins/erp5_dms/Document_getDoctypeFromContent.xml

@@ -38,7 +38,7 @@
         </item>
         <item>
             <key> <string>title</string> </key>
-            <value> <string></string> </value>
+            <value> <string>(not used)</string> </value>
         </item>
       </dictionary>
     </pickle>

bt5/erp5_dms/SkinTemplateItem/portal_skins/erp5_dms/Document_getPropertyListFromContent.xml

+<?xml version="1.0"?>
+<ZopeData>
+  <record id="1" aka="AAAAAAAAAAE=">
+    <pickle>
+      <tuple>
+        <tuple>
+          <string>Products.PythonScripts.PythonScript</string>
+          <string>PythonScript</string>
+        </tuple>
+        <none/>
+      </tuple>
+    </pickle>
+    <pickle>
+      <dictionary>
+        <item>
+            <key> <string>Python_magic</string> </key>
+            <value>
+              <none/>
+            </value>
+        </item>
+        <item>
+            <key> <string>Script_magic</string> </key>
+            <value> <int>3</int> </value>
+        </item>
+        <item>
+            <key> <string>__ac_local_roles__</string> </key>
+            <value>
+              <none/>
+            </value>
+        </item>
+        <item>
+            <key> <string>_bind_names</string> </key>
+            <value>
+              <object>
+                <klass>
+                  <global name="NameAssignments" module="Shared.DC.Scripts.Bindings"/>
+                </klass>
+                <tuple/>
+                <state>
+                  <dictionary>
+                    <item>
+                        <key> <string>_asgns</string> </key>
+                        <value>
+                          <dictionary>
+                            <item>
+                                <key> <string>name_container</string> </key>
+                                <value> <string>container</string> </value>
+                            </item>
+                            <item>
+                                <key> <string>name_context</string> </key>
+                                <value> <string>context</string> </value>
+                            </item>
+                            <item>
+                                <key> <string>name_m_self</string> </key>
+                                <value> <string>script</string> </value>
+                            </item>
+                            <item>
+                                <key> <string>name_subpath</string> </key>
+                                <value> <string>traverse_subpath</string> </value>
+                            </item>
+                          </dictionary>
+                        </value>
+                    </item>
+                  </dictionary>
+                </state>
+              </object>
+            </value>
+        </item>
+        <item>
+            <key> <string>_body</string> </key>
+            <value> <string>return {}\n
+\n
+# write your own method (probably External Method, since it is most likely\n
+# to use re) that would analyze text content of the doc\n
+# and return a dictionary of properties\n
+</string> </value>
+        </item>
+        <item>
+            <key> <string>_code</string> </key>
+            <value>
+              <none/>
+            </value>
+        </item>
+        <item>
+            <key> <string>_filepath</string> </key>
+            <value>
+              <none/>
+            </value>
+        </item>
+        <item>
+            <key> <string>_owner</string> </key>
+            <value>
+              <none/>
+            </value>
+        </item>
+        <item>
+            <key> <string>_params</string> </key>
+            <value> <string>ob,content,ptype</string> </value>
+        </item>
+        <item>
+            <key> <string>errors</string> </key>
+            <value>
+              <tuple/>
+            </value>
+        </item>
+        <item>
+            <key> <string>func_code</string> </key>
+            <value>
+              <object>
+                <klass>
+                  <global name="FuncCode" module="Shared.DC.Scripts.Signature"/>
+                </klass>
+                <tuple/>
+                <state>
+                  <dictionary>
+                    <item>
+                        <key> <string>co_argcount</string> </key>
+                        <value> <int>3</int> </value>
+                    </item>
+                    <item>
+                        <key> <string>co_varnames</string> </key>
+                        <value>
+                          <tuple>
+                            <string>ob</string>
+                            <string>content</string>
+                            <string>ptype</string>
+                          </tuple>
+                        </value>
+                    </item>
+                  </dictionary>
+                </state>
+              </object>
+            </value>
+        </item>
+        <item>
+            <key> <string>func_defaults</string> </key>
+            <value>
+              <none/>
+            </value>
+        </item>
+        <item>
+            <key> <string>id</string> </key>
+            <value> <string>Document_getPropertyListFromContent</string> </value>
+        </item>
+        <item>
+            <key> <string>warnings</string> </key>
+            <value>
+              <tuple/>
+            </value>
+        </item>
+      </dictionary>
+    </pickle>
+  </record>
+</ZopeData>

bt5/erp5_dms/SkinTemplateItem/portal_skins/erp5_dms/OOoDocument_convertAfter.xml

@@ -69,9 +69,12 @@
         <item>
             <key> <string>_body</string> </key>
             <value> <string>ob=context\n
-if ob.hasFile() and not ob.hasOOFile(): #doublecheck\n
+context.log(kw)\n
+if (ob.hasFile() and not ob.hasOOFile()) or context.REQUEST.get(\'force\'): #doublecheck\n
+  context.log(\'we go ahead\')\n
   try:\n
     res=ob.convert(force=1)\n
+    ob.setPropertyListFromContent()\n
     ob.setExternalProcessingStatusMessage(\'converted\')\n
   except Exception,e:\n
     ob.log(\'Exception\',e)\n

bt5/erp5_dms/bt/revision

-327
\ No newline at end of file
+330
\ No newline at end of file