graphiti-core 0.30.0rc2__py3-none-any.whl → 0.30.0rc4__py3-none-any.whl

graphiti_core/prompts/dedupe_nodes.py

@@ -92,12 +92,23 @@ def node(context: dict[str, Any]) -> list[Message]:
 
          TASK:
          1. Compare `new_entity` against each item in `existing_entities`.
-         2. If it refers to the same real‐world object or concept, collect its index.
-         3. Let `duplicate_idx` = the *first* collected index, or –1 if none.
-         4. Let `duplicates` = the list of *all* collected indices (empty list if none).
-        
-        Also return the full name of the NEW ENTITY (whether it is the name of the NEW ENTITY, a node it
-        is a duplicate of, or a combination of the two).
+         2. If it refers to the same real-world object or concept, collect its index.
+         3. Let `duplicate_idx` = the smallest collected index, or -1 if none.
+         4. Let `duplicates` = the sorted list of all collected indices (empty list if none).
+
+        Respond with a JSON object containing an "entity_resolutions" array with a single entry:
+        {{
+            "entity_resolutions": [
+                {{
+                    "id": integer id from NEW ENTITY,
+                    "name": the best full name for the entity,
+                    "duplicate_idx": integer index of the best duplicate in EXISTING ENTITIES, or -1 if none,
+                    "duplicates": sorted list of all duplicate indices you collected (deduplicate the list, use [] when none)
+                }}
+            ]
+        }}
+
+        Only reference indices that appear in EXISTING ENTITIES, and return [] / -1 when unsure.
         """,
         ),
     ]
@@ -126,26 +137,26 @@ def nodes(context: dict[str, Any]) -> list[Message]:
         {{
             id: integer id of the entity,
             name: "name of the entity",
-            entity_type: "ontological classification of the entity",
-            entity_type_description: "Description of what the entity type represents",
-            duplication_candidates: [
-                {{
-                    idx: integer index of the candidate entity,
-                    name: "name of the candidate entity",
-                    entity_type: "ontological classification of the candidate entity",
-                    ...<additional attributes>
-                }}
-            ]
+            entity_type: ["Entity", "<optional additional label>", ...],
+            entity_type_description: "Description of what the entity type represents"
         }}
-        
+
         <ENTITIES>
         {to_prompt_json(context['extracted_nodes'], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
         </ENTITIES>
-        
+
         <EXISTING ENTITIES>
         {to_prompt_json(context['existing_nodes'], ensure_ascii=context.get('ensure_ascii', True), indent=2)}
         </EXISTING ENTITIES>
 
+        Each entry in EXISTING ENTITIES is an object with the following structure:
+        {{
+            idx: integer index of the candidate entity (use this when referencing a duplicate),
+            name: "name of the candidate entity",
+            entity_types: ["Entity", "<optional additional label>", ...],
+            ...<additional attributes such as summaries or metadata>
+        }}
+
         For each of the above ENTITIES, determine if the entity is a duplicate of any of the EXISTING ENTITIES.
 
         Entities should only be considered duplicates if they refer to the *same real-world object or concept*.
@@ -155,14 +166,19 @@ def nodes(context: dict[str, Any]) -> list[Message]:
         - They have similar names or purposes but refer to separate instances or concepts.
 
         Task:
-        Your response will be a list called entity_resolutions which contains one entry for each entity.
-        
-        For each entity, return the id of the entity as id, the name of the entity as name, and the duplicate_idx
-        as an integer.
-        
-        - If an entity is a duplicate of one of the EXISTING ENTITIES, return the idx of the candidate it is a 
-        duplicate of.
-        - If an entity is not a duplicate of one of the EXISTING ENTITIES, return the -1 as the duplication_idx
+        Respond with a JSON object that contains an "entity_resolutions" array with one entry for each entity in ENTITIES, ordered by the entity id.
+
+        For every entity, return an object with the following keys:
+        {{
+            "id": integer id from ENTITIES,
+            "name": the best full name for the entity (preserve the original name unless a duplicate has a more complete name),
+            "duplicate_idx": the idx of the EXISTING ENTITY that is the best duplicate match, or -1 if there is no duplicate,
+            "duplicates": a sorted list of all idx values from EXISTING ENTITIES that refer to duplicates (deduplicate the list, use [] when none or unsure)
+        }}
+
+        - Only use idx values that appear in EXISTING ENTITIES.
+        - Set duplicate_idx to the smallest idx you collected for that entity, or -1 if duplicates is empty.
+        - Never fabricate entities or indices.
         """,
         ),
     ]

graphiti_core/utils/maintenance/edge_operations.py

@@ -43,6 +43,8 @@ from graphiti_core.search.search_filters import SearchFilters
 from graphiti_core.utils.datetime_utils import ensure_utc, utc_now
 from graphiti_core.utils.maintenance.dedup_helpers import _normalize_string_exact
 
+DEFAULT_EDGE_NAME = 'RELATES_TO'
+
 logger = logging.getLogger(__name__)
 
 
@@ -310,6 +312,15 @@ async def resolve_extracted_edges(
 
         edge_types_lst.append(extracted_edge_types)
 
+    for extracted_edge, extracted_edge_types in zip(extracted_edges, edge_types_lst, strict=True):
+        allowed_type_names = set(extracted_edge_types)
+        if not allowed_type_names:
+            if extracted_edge.name != DEFAULT_EDGE_NAME:
+                extracted_edge.name = DEFAULT_EDGE_NAME
+            continue
+        if extracted_edge.name not in allowed_type_names:
+            extracted_edge.name = DEFAULT_EDGE_NAME
+
     # resolve edges with related edges in the graph and find invalidation candidates
     results: list[tuple[EntityEdge, list[EntityEdge], list[EntityEdge]]] = list(
         await semaphore_gather(
@@ -392,7 +403,7 @@ async def resolve_extracted_edge(
     related_edges: list[EntityEdge],
     existing_edges: list[EntityEdge],
     episode: EpisodicNode,
-    edge_types: dict[str, type[BaseModel]] | None = None,
+    edge_type_candidates: dict[str, type[BaseModel]] | None = None,
     ensure_ascii: bool = True,
 ) -> tuple[EntityEdge, list[EntityEdge], list[EntityEdge]]:
     if len(related_edges) == 0 and len(existing_edges) == 0:
@@ -429,9 +440,9 @@ async def resolve_extracted_edge(
                 'fact_type_name': type_name,
                 'fact_type_description': type_model.__doc__,
             }
-            for i, (type_name, type_model) in enumerate(edge_types.items())
+            for i, (type_name, type_model) in enumerate(edge_type_candidates.items())
         ]
-        if edge_types is not None
+        if edge_type_candidates is not None
         else []
     )
 
@@ -468,7 +479,8 @@ async def resolve_extracted_edge(
     ]
 
     fact_type: str = response_object.fact_type
-    if fact_type.upper() != 'DEFAULT' and edge_types is not None:
+    candidate_type_names = set(edge_type_candidates or {})
+    if candidate_type_names and fact_type in candidate_type_names:
         resolved_edge.name = fact_type
 
         edge_attributes_context = {
@@ -478,7 +490,7 @@ async def resolve_extracted_edge(
             'ensure_ascii': ensure_ascii,
         }
 
-        edge_model = edge_types.get(fact_type)
+        edge_model = edge_type_candidates.get(fact_type) if edge_type_candidates else None
         if edge_model is not None and len(edge_model.model_fields) != 0:
             edge_attributes_response = await llm_client.generate_response(
                 prompt_library.extract_edges.extract_attributes(edge_attributes_context),
@@ -487,6 +499,9 @@ async def resolve_extracted_edge(
             )
 
             resolved_edge.attributes = edge_attributes_response
+    elif fact_type.upper() != 'DEFAULT':
+        resolved_edge.name = DEFAULT_EDGE_NAME
+        resolved_edge.attributes = {}
 
     end = time()
     logger.debug(

{graphiti_core-0.30.0rc2.dist-info → graphiti_core-0.30.0rc4.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: graphiti-core
-Version: 0.30.0rc2
+Version: 0.30.0rc4
 Summary: A temporal graph building library
 Project-URL: Homepage, https://help.getzep.com/graphiti/graphiti/overview
 Project-URL: Repository, https://github.com/getzep/graphiti

{graphiti_core-0.30.0rc2.dist-info → graphiti_core-0.30.0rc4.dist-info}/RECORD

@@ -44,7 +44,7 @@ graphiti_core/models/nodes/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJ
 graphiti_core/models/nodes/node_db_queries.py,sha256=TCHZKG5bQNarV9C5k4hOFFqc-LwTVQ8Pnd6okVVNKbo,12826
 graphiti_core/prompts/__init__.py,sha256=EA-x9xUki9l8wnu2l8ek_oNf75-do5tq5hVq7Zbv8Kw,101
 graphiti_core/prompts/dedupe_edges.py,sha256=WRXQi7JQZdIfKDICWyU7Wbs5WyD_KBblLBSeKdbLyuk,5914
-graphiti_core/prompts/dedupe_nodes.py,sha256=eYDk0axHEKLjZS2tKlT4Zy1fW9EJkn6EnrJLSN0fvAY,8235
+graphiti_core/prompts/dedupe_nodes.py,sha256=H4sIzpi1gBwPedTMhdY175jnLj5JtnEeb_WNITitPLU,9171
 graphiti_core/prompts/eval.py,sha256=ijwxbE87G678imdhfPvRujepQMq_JZ3XHX4vOAcVnVI,5507
 graphiti_core/prompts/extract_edge_dates.py,sha256=3Drs3CmvP0gJN5BidWSxrNvLet3HPoTybU3BUIAoc0Y,4218
 graphiti_core/prompts/extract_edges.py,sha256=mnncxb6lyr3ufKajRAh09czmJawiEM54sSPNy9ukiio,6888
@@ -69,13 +69,13 @@ graphiti_core/utils/datetime_utils.py,sha256=J-zYSq7-H-2n9hYOXNIun12kM10vNX9mMAT
 graphiti_core/utils/maintenance/__init__.py,sha256=vW4H1KyapTl-OOz578uZABYcpND4wPx3Vt6aAPaXh78,301
 graphiti_core/utils/maintenance/community_operations.py,sha256=XMiokEemn96GlvjkOvbo9hIX04Fea3eVj408NHG5P4o,11042
 graphiti_core/utils/maintenance/dedup_helpers.py,sha256=B7k6KkB6Sii8PZCWNNTvsNiy4BNTNWpoLeGgrPLq6BE,9220
-graphiti_core/utils/maintenance/edge_operations.py,sha256=fvWKJWzz4_d2Y8bOfZFjJpLnGmsFwnrutFW25LX-S08,21287
+graphiti_core/utils/maintenance/edge_operations.py,sha256=LHud849E-w6_arUszhbCEdqVxWFk_Hm-jv60YIY51zo,22015
 graphiti_core/utils/maintenance/graph_data_operations.py,sha256=42icj3S_ELAJ-NK3jVS_rg_243dmnaZOyUitJj_uJ-M,6085
 graphiti_core/utils/maintenance/node_operations.py,sha256=TKpXPtnTVxxan8I1xQyVkGn3zyRdb_Q00cgUpLcloig,16860
 graphiti_core/utils/maintenance/temporal_operations.py,sha256=IIaVtShpVkOYe6haxz3a1x3v54-MzaEXG8VsxFUNeoY,3582
 graphiti_core/utils/maintenance/utils.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 graphiti_core/utils/ontology_utils/entity_types_utils.py,sha256=4eVgxLWY6Q8k9cRJ5pW59IYF--U4nXZsZIGOVb_yHfQ,1285
-graphiti_core-0.30.0rc2.dist-info/METADATA,sha256=eGlZ91MklPKsP1m7Me1RivW4qPnc9wff6uIGZ5XRVqI,26933
-graphiti_core-0.30.0rc2.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-graphiti_core-0.30.0rc2.dist-info/licenses/LICENSE,sha256=KCUwCyDXuVEgmDWkozHyniRyWjnWUWjkuDHfU6o3JlA,11325
-graphiti_core-0.30.0rc2.dist-info/RECORD,,
+graphiti_core-0.30.0rc4.dist-info/METADATA,sha256=W12RaBo2Hhk7ttsTVOIJbMZyj11BqH1B_4UTVJENqUw,26933
+graphiti_core-0.30.0rc4.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+graphiti_core-0.30.0rc4.dist-info/licenses/LICENSE,sha256=KCUwCyDXuVEgmDWkozHyniRyWjnWUWjkuDHfU6o3JlA,11325
+graphiti_core-0.30.0rc4.dist-info/RECORD,,