ENH: refactored confusing names of identifiable adapter

src/newcrawler/crawl.py

@@ -355,7 +355,7 @@ class Crawler(object):
 
         # TODO: can the following be removed at some point
         for ent in flat:
-            if len(ent.parents) == 0:
+            if ent.role == "Record" and len(ent.parents) == 0:
                 raise RuntimeError("Records must have a parent.")
 
         resolved_references = True
@@ -398,7 +398,7 @@ class Crawler(object):
                     
                     # Check remotely
                     # TODO: remove deepcopy?
-                    identified_record = self.identifiableAdapter.retrieve_identifiable(
+                    identified_record = self.identifiableAdapter.retrieve_identifiable_for_record(
                         deepcopy(record))
                     if identified_record is None:
                         # identifiable does not exist remotely
@@ -512,7 +512,7 @@ class Crawler(object):
         for el in to_be_updated:
             self.replace_entities_by_ids(el)
 
-        identified_records = [self.identifiableAdapter.retrieve_identifiable(record) for record
+        identified_records = [self.identifiableAdapter.retrieve_identifiable_for_record(record) for record
                               in to_be_updated]
         self.remove_unnecessary_updates(to_be_updated, identified_records)
 

src/newcrawler/identifiable_adapters.py

@@ -128,11 +128,23 @@ class IdentifiableAdapter(metaclass=ABCMeta):
     def resolve_reference(self, record: db.Record):
         pass
 
+    @abstractmethod
+    def get_file(self, record: db.File):
+        pass
+
     def get_identifiable(self, record: db.Record):
         """
         retrieve the registred identifiable and fill the property values to create an
         identifiable
         """
+
+        # TODO: this is incorrect
+        # The identifiable of a File is a file object with a path set.
+        # This needs to be taken into account also for retrieving the
+        # actually identified record.
+        if record.role == "File":
+            return self.get_file(record)
+        
         registered_identifiable = self.get_registered_identifiable(record)
 
         if registered_identifiable is None:
@@ -182,7 +194,7 @@ class IdentifiableAdapter(metaclass=ABCMeta):
         return identifiable
 
     @abstractmethod
-    def retrieve_identified_record(self, identifiable: db.Record):
+    def retrieve_identified_record_for_identifiable(self, identifiable: db.Record):
         """
         Retrieve identifiable record for a given identifiable.
 
@@ -191,9 +203,8 @@ class IdentifiableAdapter(metaclass=ABCMeta):
         """
         pass
 
-    # TODO: the name is confusing. it returns the identified record
     # TODO: remove side effect
-    def retrieve_identifiable(self, record: db.Record):
+    def retrieve_identifiable_for_record(self, record: db.Record):
         """
         This function combines all functionality of the IdentifierAdapter by
         returning the identifiable after having checked for an appropriate
@@ -207,7 +218,7 @@ class IdentifiableAdapter(metaclass=ABCMeta):
         if identifiable is None:
             return None
 
-        return self.retrieve_identified_record(identifiable)
+        return self.retrieve_identified_record_for_identifiable(identifiable)
 
 
 class LocalStorageIdentifiableAdapter(IdentifiableAdapter):
@@ -345,6 +356,17 @@ class CaosDBIdentifiableAdapter(IdentifiableAdapter):
     def register_identifiable(self, name: str, definition: db.RecordType):
         self._registered_identifiables[name] = definition
 
+    def get_file(self, record: db.File):
+        if record.path is None:
+            raise RuntimeError("Path must not be None for File retrieval.")
+        candidates = db.execute_query("FIND File which is stored at {}".format(
+            record.path))
+        if len(candidates) > 1:
+            raise RuntimeError("Identifiable was not defined unambigiously.")
+        if len(candidates) == 0:
+            return None
+        return candidates[0]
+
     def get_registered_identifiable(self, record: db.Record):
         """
         returns the registred identifiable for the given Record

unittests/test_tool.py

@@ -203,7 +203,7 @@ def test_crawler_update_list(crawler, ident):
     assert len(id_r0.properties) == 2
 
     idr_r0_test = ident.retrieve_identified_record(id_r0)
-    idr_r0 = ident.retrieve_identifiable(r_cur)
+    idr_r0 = ident.retrieve_identifiable_for_record(r_cur)
     assert idr_r0 == idr_r0_test
 
     # take the first measurement in the list of records: