faiss-haystack 2.1.0__tar.gz → 3.0.1__tar.gz

faiss_haystack-3.0.1/CHANGELOG.md

@@ -0,0 +1,69 @@
+# Changelog
+
+## [integrations/faiss-v3.0.0] - 2026-10-16
+
+### 🧪 Testing
+
+- Use Python 3.11 in tests instead of 3.10 (#4103)
+- Test compatible integrations with Python 3.15; update pyproject (#4104)
+
+### ⚙️ CI
+
+- Improve changelog generation; fix existing changelogs (#3883)
+
+### 🧹 Chores
+
+- Switch remaining document stores to haystack.logging (#4011)
+- [**breaking**] Drop Python 3.10 support (#4105)
+
+
+## [integrations/faiss-v2.1.0] - 2026-09-01
+
+### 🐛 Bug Fixes
+
+- Fix new issues raised by ruff 0.16.0 (#3670)
+- `FAISSDocumentStore` with `search_term`, pagination and returning `(values, total_count)` (#3682)
+- `FAISSDocumentStore` get_metadata_field_unique_values setting list of values to return to `Any` (#3704)
+- `FAISSDocumentStore` adding filter to unique metadata values (#3753)
+- Standardize license files (#3771)
+
+### 📚 Documentation
+
+- Replace old haystack core imports with haystack_integrations paths (#3545)
+
+### 🧪 Testing
+
+- Track test coverage for all integrations (#3065)
+- Mark some tests as integration in Faiss and Qdrant (#3076)
+- Faiss - add unit tests (#3213)
+- `FAISSDocumentStore` removing duplicated tests already covered by haystack Mixin tests (#3778)
+
+### 🧹 Chores
+
+- Improve consistency of integrations folder structure (#3430)
+- Re-standardize READMEs (#3768)
+
+
+## [integrations/faiss-v2.0.0] - 2026-03-23
+
+### 🧪 Testing
+
+- Test compatible integrations with python 3.14; update pyproject (#3001)
+- `FaissDocumentStore` use Mixin tests (#3024)
+
+### ⚙️ CI
+
+- Adding FAISS github workflow (#2879)
+
+### 🧹 Chores
+
+- FAISS - small changes to improve consistency with the codebase (#2890)
+- Standardize author mentions (#2897)
+- Add ANN type annotations to faiss, fastembed, firecrawl, github, google_ai (#2988)
+- Enforce ruff docstring rules (D102/D103/D205/D209/D213/D417/D419) in integrations 11-20 (#3009)
+
+### 🌀 Miscellaneous
+
+- Add FAISS DocumentStore integration (#2844)
+
+<!-- generated by git-cliff -->

{faiss_haystack-2.1.0 → faiss_haystack-3.0.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: faiss-haystack
-Version: 2.1.0
+Version: 3.0.1
 Summary: An integration of FAISS (library for efficient similarity search of dense vectors) with Haystack
 Project-URL: Documentation, https://github.com/deepset-ai/haystack-core-integrations/tree/main/integrations/faiss#readme
 Project-URL: Issues, https://github.com/deepset-ai/haystack-core-integrations/issues
@@ -11,17 +11,17 @@ License-File: LICENSE.txt
 Classifier: Development Status :: 4 - Beta
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Programming Language :: Python
-Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
 Classifier: Programming Language :: Python :: 3.14
+Classifier: Programming Language :: Python :: 3.15
 Classifier: Programming Language :: Python :: Implementation :: CPython
 Classifier: Programming Language :: Python :: Implementation :: PyPy
-Requires-Python: >=3.10
+Requires-Python: >=3.11
 Requires-Dist: faiss-cpu>=1.8.0
 Requires-Dist: haystack-ai>=2.26.1
-Requires-Dist: numpy<2,>=1.22; python_version < '3.13'
+Requires-Dist: numpy<2,>=1.23.2; python_version < '3.13'
 Description-Content-Type: text/markdown
 
 # faiss-haystack

{faiss_haystack-2.1.0 → faiss_haystack-3.0.1}/pyproject.toml

@@ -7,7 +7,7 @@ name = "faiss-haystack"
 dynamic = ["version"]
 description = 'An integration of FAISS (library for efficient similarity search of dense vectors) with Haystack'
 readme = "README.md"
-requires-python = ">=3.10"
+requires-python = ">=3.11"
 license = "Apache-2.0"
 keywords = []
 authors = [{ name = "deepset GmbH", email = "info@deepset.ai" }]
@@ -15,18 +15,18 @@ classifiers = [
   "License :: OSI Approved :: Apache Software License",
   "Development Status :: 4 - Beta",
   "Programming Language :: Python",
-  "Programming Language :: Python :: 3.10",
   "Programming Language :: Python :: 3.11",
   "Programming Language :: Python :: 3.12",
   "Programming Language :: Python :: 3.13",
   "Programming Language :: Python :: 3.14",
+  "Programming Language :: Python :: 3.15",
   "Programming Language :: Python :: Implementation :: CPython",
   "Programming Language :: Python :: Implementation :: PyPy",
 ]
 dependencies = [
   "haystack-ai>=2.26.1",
   "faiss-cpu>=1.8.0",
-  "numpy>=1.22,<2; python_version < '3.13'",
+  "numpy>=1.23.2,<2; python_version < '3.13'",
 ]
 
 [project.urls]

{faiss_haystack-2.1.0 → faiss_haystack-3.0.1}/src/haystack_integrations/document_stores/faiss/document_store.py

@@ -3,14 +3,13 @@
 # SPDX-License-Identifier: Apache-2.0
 
 import json
-import logging
 from collections.abc import Iterable
 from dataclasses import replace
 from pathlib import Path
 from typing import Any
 
 import numpy as np
-from haystack import default_from_dict, default_to_dict
+from haystack import default_from_dict, default_to_dict, logging
 from haystack.dataclasses import Document
 from haystack.document_stores.errors import DocumentStoreError, DuplicateDocumentError
 from haystack.document_stores.types import DuplicatePolicy
@@ -153,9 +152,9 @@ class FAISSDocumentStore:
                     msg = f"Document with id '{doc.id}' already exists."
                     raise DuplicateDocumentError(msg)
 
-        # Process documents
-        ids_to_add_to_index = []
-        vectors_to_add = []
+        # Process documents. Pending vectors are keyed by document ID so that a later duplicate in the same
+        # batch replaces the earlier one instead of leaving an orphaned vector in the FAISS index.
+        pending_vectors: dict[str, tuple[int, list[float]]] = {}
 
         docs_written = 0
 
@@ -169,6 +168,7 @@ class FAISSDocumentStore:
                 # FAISS doesn't support easy update. We'd have to remove and add.
                 # For MVP, let's implement remove then add for overwrite.
                 self.delete_documents([doc.id])
+                pending_vectors.pop(doc.id, None)
 
             self.documents[doc.id] = doc
 
@@ -180,15 +180,14 @@ class FAISSDocumentStore:
                 self.id_map[int_id] = doc.id
                 self.inverse_id_map[doc.id] = int_id
 
-                ids_to_add_to_index.append(int_id)
-                vectors_to_add.append(doc.embedding)
+                pending_vectors[doc.id] = (int_id, doc.embedding)
 
             docs_written += 1
 
         # Add to FAISS
-        if vectors_to_add:
-            vectors = np.array(vectors_to_add, dtype="float32")
-            ids = np.array(ids_to_add_to_index, dtype="int64")
+        if pending_vectors:
+            vectors = np.array([vector for _, vector in pending_vectors.values()], dtype="float32")
+            ids = np.array([int_id for int_id, _ in pending_vectors.values()], dtype="int64")
             index = self._get_index_or_raise()
             index.add_with_ids(vectors, ids)
 
@@ -599,7 +598,7 @@ class FAISSDocumentStore:
         # Verify sync
         if len(self.documents) != len(self.id_map):
             logger.warning(
-                "Loaded %d documents but %d ID mappings. Index might be out of sync.",
-                len(self.documents),
-                len(self.id_map),
+                "Loaded {documents_count} documents but {mappings_count} ID mappings. Index might be out of sync.",
+                documents_count=len(self.documents),
+                mappings_count=len(self.id_map),
             )

{faiss_haystack-2.1.0 → faiss_haystack-3.0.1}/tests/test_document_store.py

@@ -55,6 +55,21 @@ class TestFAISSDocumentStoreUnit:
         with pytest.raises(DuplicateDocumentError, match="already exists"):
             ds.write_documents([Document(id="dup", content="second")], policy=DuplicatePolicy.FAIL)
 
+    def test_write_documents_overwrite_duplicate_in_same_batch_leaves_no_orphan_vector(self):
+        ds = FAISSDocumentStore(embedding_dim=2)
+        docs = [
+            Document(id="a", content="old", embedding=[0.0, 0.0]),
+            Document(id="a", content="new", embedding=[0.1, 0.1]),
+            Document(id="b", content="b", embedding=[5.0, 5.0]),
+        ]
+        ds.write_documents(docs, policy=DuplicatePolicy.OVERWRITE)
+
+        assert ds.count_documents() == 2
+        assert ds.index.ntotal == 2
+        results = ds.search([0.0, 0.0], top_k=2)
+        assert [doc.id for doc in results] == ["a", "b"]
+        assert results[0].content == "new"
+
     @pytest.mark.parametrize(
         ("filters", "error_match"),
         [

faiss_haystack-2.1.0/CHANGELOG.md

@@ -1,25 +0,0 @@
-# Changelog
-
-## [integrations/faiss-v2.0.0] - 2026-03-23
-
-### 🧪 Testing
-
-- Test compatible integrations with python 3.14; update pyproject (#3001)
-- `FaissDocumentStore` use Mixin tests (#3024)
-
-### ⚙️ CI
-
-- Adding FAISS github workflow (#2879)
-
-### 🧹 Chores
-
-- FAISS - small changes to improve consistency with the codebase (#2890)
-- Standardize author mentions (#2897)
-- Add ANN type annotations to faiss, fastembed, firecrawl, github, google_ai (#2988)
-- Enforce ruff docstring rules (D102/D103/D205/D209/D213/D417/D419) in integrations 11-20 (#3009)
-
-### 🌀 Miscellaneous
-
-- Add FAISS DocumentStore integration (#2844)
-
-<!-- generated by git-cliff -->