refactor merge

src/caoscrawler/sync_graph.py

@@ -506,13 +506,18 @@ class SyncGraph:
         """A path or an ID is sufficiently identifying. Thus, those entities can be marked as
         checked
 
+        When this function returns, there is only one node for each ID (i.e. no two nodes with the
+        same ID). The same is true for paths.
+
+        This function also updates _id_look_up and _path_look_up
+
         Last review by Alexander Schlemmer on 2024-05-29.
         """
         for node in list(self.nodes):
             if node.id is not None:
                 eq_node = self.get_equivalent(node)
                 if eq_node is not None:
-                    self._merge_into(node, eq_node)
+                    self._basic_merge_into(node, eq_node)
                 else:
                     self._id_look_up[node.id] = node
                     self._mark_existing(node)
@@ -521,7 +526,7 @@ class SyncGraph:
             if node.path is not None:
                 eq_node = self.get_equivalent(node)
                 if eq_node is not None:
-                    self._merge_into(node, eq_node)
+                    self._basic_merge_into(node, eq_node)
                 else:
                     self._path_look_up[node.path] = node
                     try:
@@ -531,39 +536,19 @@ class SyncGraph:
                     remote_id = None
                     if existing is not None:
                         remote_id = existing.id
-                    self._path_look_up[node.path] = node
                     self.set_id_of_node(node, remote_id)
 
-    def _merge_into(self, source: SyncNode, target: SyncNode):
-        """tries to merge source into target and performs the necessary updates:
-        - update the member variables of target using source (``target.update(source)``).
-        - replaces reference values to source by target
-        - updates the reference map
-        - updates lookup tables
-        - removes source from node lists
-        - marks target as missing/existing if source was marked that way
-        - adds an identifiable if now possible (e.g. merging based on ID might allow create an
-          identifiable when none of the two nodes had the sufficient properties on its own before)
-        - check whether dependent nodes can now get an identifiable (the merge might have set the
-          ID such that dependent nodes can now create an identifiable)
+    def _basic_merge_into(self, source: SyncNode, target: SyncNode):
+        """tries to merge source into target and updates member variables
 
-        Last review by Alexander Schlemmer on 2024-05-29.
+        - reference maps are updated
+        - self.nodes is updated
+        - self.unchecked is updated
+        - lookups are being updated
         """
         # sanity checks
         if source is target:
             raise ValueError("source must not be target")
-        if target.id is None and source.id is not None:
-            if self._id_look_up[source.id] != source:
-                raise ValueError(
-                    "It is assumed that always only one node exists with a certain ID and that "
-                    "node is in the look up"
-                )
-        if target.path is None and source.path is not None:
-            if self._id_look_up[source.path] != source:
-                raise ValueError(
-                    "It is assumed that always only one node exists with a certain path and that"
-                    " node is in the look up"
-                )
 
         target.update(source)
 
@@ -575,12 +560,13 @@ class SyncGraph:
 
         # update reference mappings
         for setA, setB in (
-              (self.forward_references, self.backward_references),
-              (self.backward_references, self.forward_references),
-              (self.forward_references_id_props, self.backward_references_id_props),
-              (self.backward_references_id_props, self.forward_references_id_props),
-              (self.forward_references_backref, self.backward_references_backref),
-              (self.backward_references_backref, self.forward_references_backref),):
+            (self.forward_references, self.backward_references),
+            (self.backward_references, self.forward_references),
+            (self.forward_references_id_props, self.backward_references_id_props),
+            (self.backward_references_id_props, self.forward_references_id_props),
+            (self.forward_references_backref, self.backward_references_backref),
+            (self.backward_references_backref, self.forward_references_backref),
+        ):
             for node in setA.pop(id(source)):
                 setA[id(target)].add(node)
                 setB[id(node)].remove(source)
@@ -596,9 +582,24 @@ class SyncGraph:
         if target.path is not None:
             self._path_look_up[target.path] = target
         if target.identifiable is not None:
-            self._identifiable_look_up[target.identifiable.get_representation()] = (
-                target
-            )
+            self._identifiable_look_up[target.identifiable.get_representation()] = target
+
+    def _merge_into(self, source: SyncNode, target: SyncNode):
+        """tries to merge source into target and performs the necessary updates:
+        - update the member variables of target using source (``target.update(source)``).
+        - replaces reference values to source by target
+        - updates the reference map
+        - updates lookup tables
+        - removes source from node lists
+        - marks target as missing/existing if source was marked that way
+        - adds an identifiable if now possible (e.g. merging based on ID might allow create an
+          identifiable when none of the two nodes had the sufficient properties on its own before)
+        - check whether dependent nodes can now get an identifiable (the merge might have set the
+          ID such that dependent nodes can now create an identifiable)
+
+        Last review by Alexander Schlemmer on 2024-05-29.
+        """
+        self._basic_merge_into(source, target)
 
         if (id(source) in self._existing and id(target) in self._missing) or (
             id(target) in self._existing and id(source) in self._missing