pgvector-haystack 1.2.1__tar.gz → 1.3.0__tar.gz

{pgvector_haystack-1.2.1 → pgvector_haystack-1.3.0}/CHANGELOG.md

@@ -1,5 +1,12 @@
 # Changelog
 
+## [integrations/pgvector-v1.2.1] - 2025-01-10
+
+### 🐛 Bug Fixes
+
+- PgvectorDocumentStore - use appropriate schema name if dropping index (#1277)
+
+
 ## [integrations/pgvector-v1.2.0] - 2024-11-22
 
 ### 🚀 Features

{pgvector_haystack-1.2.1 → pgvector_haystack-1.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pgvector-haystack
-Version: 1.2.1
+Version: 1.3.0
 Summary: An integration of pgvector (vector search extension for Postgres) with Haystack
 Project-URL: Source, https://github.com/deepset-ai/haystack-core-integrations
 Project-URL: Documentation, https://github.com/deepset-ai/haystack-core-integrations/blob/main/integrations/pgvector/README.md

{pgvector_haystack-1.2.1 → pgvector_haystack-1.3.0}/src/haystack_integrations/document_stores/pgvector/filters.py

@@ -232,6 +232,20 @@ def _in(field: str, value: Any) -> Tuple[str, List]:
     return f"{field} = ANY(%s)", [value]
 
 
+def _like(field: str, value: Any) -> Tuple[str, Any]:
+    if not isinstance(value, str):
+        msg = f"{field}'s value must be a str when using 'LIKE' "
+        raise FilterError(msg)
+    return f"{field} LIKE %s", value
+
+
+def _not_like(field: str, value: Any) -> Tuple[str, Any]:
+    if not isinstance(value, str):
+        msg = f"{field}'s value must be a str when using 'LIKE' "
+        raise FilterError(msg)
+    return f"{field} NOT LIKE %s", value
+
+
 COMPARISON_OPERATORS = {
     "==": _equal,
     "!=": _not_equal,
@@ -241,4 +255,6 @@ COMPARISON_OPERATORS = {
     "<=": _less_than_equal,
     "in": _in,
     "not in": _not_in,
+    "like": _like,
+    "not like": _not_like,
 }

{pgvector_haystack-1.2.1 → pgvector_haystack-1.3.0}/tests/test_filters.py

@@ -2,7 +2,7 @@ from typing import List
 
 import pytest
 from haystack.dataclasses.document import Document
-from haystack.testing.document_store import FilterDocumentsTest
+from haystack.testing.document_store import FilterDocumentsTest, FilterDocumentsTestWithDataframe
 from pandas import DataFrame
 from psycopg.sql import SQL
 from psycopg.types.json import Jsonb
@@ -17,7 +17,7 @@ from haystack_integrations.document_stores.pgvector.filters import (
 
 
 @pytest.mark.integration
-class TestFilters(FilterDocumentsTest):
+class TestFilters(FilterDocumentsTest, FilterDocumentsTestWithDataframe):
     def assert_documents_are_equal(self, received: List[Document], expected: List[Document]):
         """
         This overrides the default assert_documents_are_equal from FilterDocumentsTest.
@@ -40,6 +40,62 @@ class TestFilters(FilterDocumentsTest):
     @pytest.mark.skip(reason="NOT operator is not supported in PgvectorDocumentStore")
     def test_not_operator(self, document_store, filterable_docs): ...
 
+    def test_like_operator(self, document_store, filterable_docs):
+        document_store.write_documents(filterable_docs)
+        filters = {"field": "content", "operator": "like", "value": "%Foo%"}
+        result = document_store.filter_documents(filters=filters)
+
+        self.assert_documents_are_equal(result, [d for d in filterable_docs if "Foo" in d.content])
+
+    def test_like_operator_startswith(self, document_store, filterable_docs):
+        document_store.write_documents(filterable_docs)
+        filters = {"field": "content", "operator": "like", "value": "Foo%"}
+        result = document_store.filter_documents(filters=filters)
+
+        self.assert_documents_are_equal(result, [d for d in filterable_docs if d.content.startswith("Foo")])
+
+    def test_like_operator_nb_chars(self, document_store, filterable_docs):
+        document_store.write_documents(filterable_docs)
+        filters = {"field": "content", "operator": "like", "value": "A Foobar Document__"}
+        result = document_store.filter_documents(filters=filters)
+
+        self.assert_documents_are_equal(
+            result,
+            [
+                d
+                for d in filterable_docs
+                if (d.content.startswith("A Foobar Document") and len(d.content) == (len("A Foobar Document") + 2))
+            ],
+        )
+
+    def test_not_like_operator(self, document_store, filterable_docs):
+        document_store.write_documents(filterable_docs)
+        filters = {"field": "content", "operator": "not like", "value": "%Foo%"}
+        result = document_store.filter_documents(filters=filters)
+
+        self.assert_documents_are_equal(result, [d for d in filterable_docs if "Foo" not in d.content])
+
+    def test_not_like_operator_startswith(self, document_store, filterable_docs):
+        document_store.write_documents(filterable_docs)
+        filters = {"field": "content", "operator": "not like", "value": "Foo%"}
+        result = document_store.filter_documents(filters=filters)
+
+        self.assert_documents_are_equal(result, [d for d in filterable_docs if not d.content.startswith("Foo")])
+
+    def test_not_like_operator_nb_chars(self, document_store, filterable_docs):
+        document_store.write_documents(filterable_docs)
+        filters = {"field": "content", "operator": "not like", "value": "A Foobar Document__"}
+        result = document_store.filter_documents(filters=filters)
+
+        self.assert_documents_are_equal(
+            result,
+            [
+                d
+                for d in filterable_docs
+                if not (d.content.startswith("A Foobar Document") and len(d.content) == (len("A Foobar Document") + 2))
+            ],
+        )
+
     def test_complex_filter(self, document_store, filterable_docs):
         document_store.write_documents(filterable_docs)
         filters = {