Data Analysis: Support transient in- and output

This commit is a rebase of the following original commits:

Analysis execution methods supporting transient in- and output

---

Conflicting files:
bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/DataAnalysis_executeDataOperation.py
bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/DataAnalysis_executeDataOperation.xml
bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/ERP5Site_executeDataAnalysisList.py

processing of analysis with transient items

---

Conflicts:
bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/DataAnalysis_executeDataOperation.py
bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/ERP5Site_createDataAnalysisList.py

bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/DataAnalysis_executeDataOperation.py

 portal = context.getPortalObject()
 operation = None
-use = None
+use_list = []
 parameter_dict = {}
 
+transient_output_item = None
 context.checkConsistency(fixit=True)
 initial_product = context.getSpecialiseValue(portal_type="Data Transformation").getResourceValue()
 for analysis_line in sorted(context.objectValues(portal_type="Data Analysis Line"),
                            key=lambda x: x.getIntIndex()):
   resource = analysis_line.getResourceValue()
   if resource == initial_product:
-    use = analysis_line.getUse()
+    use_list = analysis_line.getUseList()
   if resource is not None:
     resource_portal_type = resource.getPortalType()
   else:
@@ -19,13 +20,20 @@ for analysis_line in sorted(context.objectValues(portal_type="Data Analysis Line
     operation = analysis_line.getResourceValue()
   else:
     parameter = {}
-    for portal_type in ["Data Array", "Progress Indicator"] + \
+    for portal_type in ["Data Array", "Data Array View", "Progress Indicator"] + \
                         list(portal.getPortalDataSinkTypeList()) + \
                         list(portal.getPortalDataDescriptorTypeList()):
       value = analysis_line.getAggregateValue(portal_type=portal_type)
       if value is not None:
         parameter[portal_type] = value
-
+    if analysis_line.getQuantity() < 0 and "big_data/analysis/transient" in analysis_line.getUseList():
+      # at the moment we only support transient data arrays
+      parameter['Data Array'] = transient_input_item
+    if analysis_line.getQuantity() > 0 and "big_data/analysis/transient" in analysis_line.getUseList():
+      # at the moment we only support transient data arrays
+      transient_output_item = portal.data_array_module.newContent(portal_type='Data Array',
+                                                                  temp_object=True)
+      parameter['Data Array'] = transient_output_item
     for base_category in analysis_line.getVariationRangeBaseCategoryList():
       parameter[base_category] = analysis_line.getVariationCategoryItemList(
                                    base_category_list=(base_category,))[0][0]
@@ -43,16 +51,21 @@ for analysis_line in sorted(context.objectValues(portal_type="Data Analysis Line
       parameter_dict[reference].append(parameter)
     else:
       parameter_dict[reference] = parameter
+      
+if transient_output_item is not None and not consuming_analysis_list:
+  return
 
 script_id = operation.getScriptId()
-
 out = getattr(operation_analysis_line, script_id)(**parameter_dict)
 
+for consuming_analysis in consuming_analysis_list:
+  portal.restrictedTraverse(consuming_analysis).DataAnalysis_executeDataOperation(transient_input_item = transient_output_item)
+
 if out == 1:
-  context.activate(serialization_tag=str(context.getUid())).DataAnalysis_executeDataOperation()
+  context.activate(serialization_tag=str(context.getUid())).DataAnalysis_executeDataOperation(consuming_analysis_list)
 else:
   # only stop batch ingestions
-  if use == "big_data/ingestion/batch":
+  if "big_data/ingestion/batch" in use_list:
     context.stop()
   # stop refresh
   if context.getRefreshState() == "refresh_started":

bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/DataAnalysis_executeDataOperation.xml

@@ -50,7 +50,7 @@
         </item>
         <item>
             <key> <string>_params</string> </key>
-            <value> <string></string> </value>
+            <value> <string>consuming_analysis_list=[], transient_input_item=None</string> </value>
         </item>
         <item>
             <key> <string>id</string> </key>

bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/ERP5Site_createDataAnalysisList.py

@@ -121,13 +121,13 @@ for movement in portal_catalog(query = query):
             resource_relative_url = resource.getRelativeUrl())
             
           for related_movement in related_movement_list:
-            #aggregate_set.update(related_movement.getAggregateSet())
-            related_movement.getParentValue().deliver()
-
-
+            if "big_data/ingestion/batch" in related_movement.getUseList():
+              related_movement.getParentValue().deliver()
+      
       # create new item based on item_type if it is not already aggregated
       aggregate_type_set = set(
         [portal.restrictedTraverse(a).getPortalType() for a in aggregate_set])
+      
       for item_type in transformation_line.getAggregatedPortalTypeList():
         # if item is not yet aggregated to this line, search it by related project
         # and source If the item is a data configuration or a device configuration
@@ -149,8 +149,13 @@ for movement in portal_catalog(query = query):
                 item_project_relative_url=delivery.getDestinationProject(),
                 item_source_relative_url=delivery.getSource())
   
-          elif item_type != "Data Array Line":
-            
+          elif all(
+            [
+              item_type != "Data Array Line",
+              not (item_type == "Data Array" and "big_data/analysis/transient" in transformation_line.getUseList()),
+              not (quantity < 0 and item_type == "Data Array")
+            ]
+          ):  
             item_query_dict = dict(
               portal_type=item_type,
               validation_state="validated",
@@ -159,6 +164,7 @@ for movement in portal_catalog(query = query):
               item_resource_uid=resource.getUid(),
               item_source_relative_url=data_analysis.getSource())
 
+
             if data_analysis.getDestinationProjectValue() is not None:
               item_query_dict["item_project_relative_url"] = data_analysis.getDestinationProject()
             item = portal.portal_catalog.getResultValue(**item_query_dict)
@@ -176,7 +182,9 @@ for movement in portal_catalog(query = query):
               pass
           aggregate_set.add(item.getRelativeUrl())
 
+      tag = "%s-%s" %(data_analysis.getUid(), transformation_line.getUid())
       data_analysis_line = data_analysis.newContent(
+        activate_kw={'tag': tag},
         portal_type = "Data Analysis Line",
         title = transformation_line.getTitle(),
         reference = transformation_line.getReference(),
@@ -185,7 +193,7 @@ for movement in portal_catalog(query = query):
         variation_category_list = transformation_line.getVariationCategoryList(),
         quantity = quantity,
         quantity_unit = transformation_line.getQuantityUnit(),
-        use = transformation_line.getUse(),
+        use_list = transformation_line.getUseList(),
         aggregate_set = aggregate_set)
       # for intput lines of first level analysis set causality and specialise
       if quantity < 0 and delivery.getPortalType() == "Data Ingestion":
@@ -193,7 +201,14 @@ for movement in portal_catalog(query = query):
           causality_value = delivery,
           specialise_value_list = data_supply_list)
     
-    data_analysis.checkConsistency(fixit=True)
+    # fix consistency of line and all affected items. Do it after reindexing
+    # activities of newly created Data Analysis Line finished, because check
+    # consistency script might need to find the newly created Data Analysis
+    # Line in catalog.
+    data_analysis_line.checkConsistency(fixit=True)
+    for item in data_analysis_line.getAggregateValueList():
+      item.activate(after_tag=tag).checkConsistency(fixit=True)
+
     try:
       data_analysis.start()
     except UnsupportedWorkflowMethod:

bt5/erp5_wendelin/SkinTemplateItem/portal_skins/erp5_wendelin/ERP5Site_executeDataAnalysisList.py

 portal = context.getPortalObject()
 
-#search_kw = {
-#  'simulation_state': 'started',
-#  'portal_type': 'Data Analysis',
-#}
+consuming_analysis_list_dict = {}
 
-#method_kw = {
-#  'active_process': this_portal_type_active_process,
-#}
-
-#activate_kw = {
-#  'tag': tag,
-#  'priority': priority,
-#}
-
-#portal.portal_catalog.searchAndActivate(
-#  method_id='DataAnalysis_executeDataOperation',
-#  method_kw=method_kw,
-#  activate_kw=activate_kw,
-#  **search_kw)
+def add_consuming_analysis(producing_analysis_relative_url, consuming_analysis_relative_url):
+  consuming_analysis_list = consuming_analysis_list_dict.setdefault(producing_analysis_relative_url, [])
+  consuming_analysis_list.append(consuming_analysis_relative_url)
 
+# First we split all started Data Analysis documents into documents with transient
+# inputs and without transient inputs. Documents without transient inputs
+# are added to 'data_analysis_list'.
+data_analysis_list = []
 for data_analysis in portal.portal_catalog(portal_type = "Data Analysis",
                                            simulation_state = "started"):
+  has_transient_input = False
+  for line in data_analysis.objectValues(portal_type="Data Analysis Line"):
+    if line.getUse() == "big_data/analysis/transient" and line.getQuantity() < 0:
+      has_transient_input = True
+      add_consuming_analysis(line.getParentValue().getCausality(), line.getParentRelativeUrl())
+  if not has_transient_input:
+    data_analysis_list.append(data_analysis)
+
+# Now we will activate `executeDataOperation` on given Data Analysis documents
+for data_analysis in data_analysis_list:
   if not data_analysis.hasActivity():
     if data_analysis.getRefreshState() == "current":
+      consuming_analysis_list = consuming_analysis_list_dict.get(data_analysis.getRelativeUrl(), [])
       data_analysis.activate(serialization_tag=str(data_analysis.getUid()))\
-        .DataAnalysis_executeDataOperation()
+        .DataAnalysis_executeDataOperation(consuming_analysis_list)
 
+# Finally we refresh specified Data Analysis documents
 for data_analysis in portal.portal_catalog(portal_type = "Data Analysis",
                                            refresh_state = "refresh_planned"):
   if data_analysis.getRefreshState() == "refresh_planned":