Rather than read Data Streams' properties from ZODB object do read them from...

Rather than read Data Streams' properties from ZODB object do read them from ZSQL's brain object - much faster when data sets contain hunreds of thousands of files.

bt5/erp5_wendelin_data_lake_ingestion/SkinTemplateItem/portal_skins/erp5_wendelin_data_lake/ERP5Site_getDataStreamList.py

 """
 This script is called from ebulk client to get list of Data Streams for a Data set.
 """
-
 import json
 from erp5.component.module.Log import log
 
@@ -9,28 +8,41 @@ portal = context.getPortalObject()
 
 try:
   data_set = portal.data_set_module.get(data_set_reference)
-  if data_set is None or portal.ERP5Site_checkReferenceInvalidated(data_set):
+  # XXX: why does we need reference= "something_invalidated" when we have Data Set's state ?
+  if data_set is None or data_set.getReference().endswith("_invalid"):
     return { "status_code": 0, "result": [] }
 except Exception as e: # fails because unauthorized access
   log("Unauthorized access to getDataStreamList: " + str(e))
   return { "status_code": 1, "error_message": "401 - Unauthorized access. Please check your user credentials and try again." }
 
 data_stream_dict = {}
-for stream in data_set.DataSet_getDataStreamList():
-  if stream and not portal.ERP5Site_checkReferenceInvalidated(stream) and stream.getValidationState() != "draft":
-    data_stream_info_dict = { 'id': 'data_stream_module/'+stream.getId(),
-                              'size': stream.getSize(),
-                              'hash': stream.getVersion() }
-    if stream.getReference() in data_stream_dict:
-      data_stream_dict[stream.getReference()]['data-stream-list'].append(data_stream_info_dict)
-      data_stream_dict[stream.getReference()]['large-hash'] = data_stream_dict[stream.getReference()]['large-hash'] + str(stream.getVersion())
-      data_stream_dict[stream.getReference()]['full-size'] = int(data_stream_dict[stream.getReference()]['full-size']) + int(stream.getSize())
-    else:
-      data_stream_dict[stream.getReference()] = { 'data-stream-list': [data_stream_info_dict],
-                                                  'id': 'data_stream_module/'+stream.getId(),
-                                                  'reference': stream.getReference(),
-                                                  'large-hash': stream.getVersion(),
-                                                  'full-size': stream.getSize() }
+# XXX: reference NOT ending with "_invalidated" -> why is that needed when we can invalidate Data Stream ???
+# XXX: state != draft
+catalog_kw = dict(portal_type = "Data Stream", 
+                  set_uid = data_set.getUid(),
+                  validation_state = ['published', 'validated'])
+data_stream_brain_list = portal.portal_catalog(**catalog_kw)
+context.log("Data Streams found=%s" %len(data_stream_brain_list))
+
+for stream_brain in data_stream_brain_list:
+  reference = stream_brain.getReference()
+  version = stream_brain.version
+  size = stream_brain.size
+  data_stream_id = "data_stream_module/%s" %stream_brain.id
+  #context.log("id=%s, version=%s, reference=%s, size=%s" %(data_stream_id, version, reference, size))
+  data_stream_info_dict = {'id': data_stream_id,
+                           'size': size,
+                           'hash': version}
+  if reference in data_stream_dict:
+    data_stream_dict[reference]['data-stream-list'].append(data_stream_info_dict)
+    data_stream_dict[reference]['large-hash'] = data_stream_dict[reference]['large-hash'] + str(version)
+    data_stream_dict[reference]['full-size'] = int(data_stream_dict[reference]['full-size']) + int(size)
+  else:
+    data_stream_dict[reference] = { 'data-stream-list': [data_stream_info_dict],
+                                    'id': data_stream_id,
+                                    'reference': reference,
+                                    'large-hash': version,
+                                    'full-size': size}
 
 result_dict = { 'status_code': 0, 'result': data_stream_dict.values()}
 return json.dumps(result_dict)