pgvector-haystack 1.2.1__py3-none-any.whl → 2.0.0__py3-none-any.whl

haystack_integrations/document_stores/pgvector/document_store.py

@@ -28,7 +28,6 @@ CREATE TABLE IF NOT EXISTS {schema_name}.{table_name} (
 id VARCHAR(128) PRIMARY KEY,
 embedding VECTOR({embedding_dimension}),
 content TEXT,
-dataframe JSONB,
 blob_data BYTEA,
 blob_meta JSONB,
 blob_mime_type VARCHAR(255),
@@ -37,15 +36,14 @@ meta JSONB)
 
 INSERT_STATEMENT = """
 INSERT INTO {schema_name}.{table_name}
-(id, embedding, content, dataframe, blob_data, blob_meta, blob_mime_type, meta)
-VALUES (%(id)s, %(embedding)s, %(content)s, %(dataframe)s, %(blob_data)s, %(blob_meta)s, %(blob_mime_type)s, %(meta)s)
+(id, embedding, content, blob_data, blob_meta, blob_mime_type, meta)
+VALUES (%(id)s, %(embedding)s, %(content)s, %(blob_data)s, %(blob_meta)s, %(blob_mime_type)s, %(meta)s)
 """
 
 UPDATE_STATEMENT = """
 ON CONFLICT (id) DO UPDATE SET
 embedding = EXCLUDED.embedding,
 content = EXCLUDED.content,
-dataframe = EXCLUDED.dataframe,
 blob_data = EXCLUDED.blob_data,
 blob_meta = EXCLUDED.blob_meta,
 blob_mime_type = EXCLUDED.blob_mime_type,
@@ -555,10 +553,18 @@ class PgvectorDocumentStore:
             db_document["blob_data"] = blob.data if blob else None
             db_document["blob_meta"] = Jsonb(blob.meta) if blob and blob.meta else None
             db_document["blob_mime_type"] = blob.mime_type if blob and blob.mime_type else None
-
-            db_document["dataframe"] = Jsonb(db_document["dataframe"]) if db_document["dataframe"] else None
             db_document["meta"] = Jsonb(db_document["meta"])
 
+            if "dataframe" in db_document:
+                dataframe = db_document.pop("dataframe", None)
+                if dataframe:
+                    logger.warning(
+                        "Document %s has the `dataframe` field set. "
+                        "PgvectorDocumentStore no longer supports dataframes and this field will be ignored. "
+                        "The `dataframe` field will soon be removed from Haystack Document.",
+                        db_document["id"],
+                    )
+
             if "sparse_embedding" in db_document:
                 sparse_embedding = db_document.pop("sparse_embedding", None)
                 if sparse_embedding:
@@ -585,12 +591,20 @@ class PgvectorDocumentStore:
             blob_data = haystack_dict.pop("blob_data")
             blob_meta = haystack_dict.pop("blob_meta")
             blob_mime_type = haystack_dict.pop("blob_mime_type")
+            dataframe = haystack_dict.pop("dataframe", None)
 
-            # postgresql returns the embedding as a string
-            # so we need to convert it to a list of floats
+            # convert the embedding to a list of floats
             if document.get("embedding") is not None:
                 haystack_dict["embedding"] = document["embedding"].tolist()
 
+            if dataframe:
+                logger.warning(
+                    "Document %s has the `dataframe` field set. "
+                    "PgvectorDocumentStore no longer supports dataframes and this field will be ignored. "
+                    "The `dataframe` field will soon be removed from Haystack Document.",
+                    haystack_dict["id"],
+                )
+
             haystack_document = Document.from_dict(haystack_dict)
 
             if blob_data:

haystack_integrations/document_stores/pgvector/filters.py

@@ -6,7 +6,6 @@ from itertools import chain
 from typing import Any, Dict, List, Literal, Tuple
 
 from haystack.errors import FilterError
-from pandas import DataFrame
 from psycopg.sql import SQL
 from psycopg.types.json import Jsonb
 
@@ -93,10 +92,6 @@ def _parse_comparison_condition(condition: Dict[str, Any]) -> Tuple[str, List[An
         raise FilterError(msg)
 
     value: Any = condition["value"]
-    if isinstance(value, DataFrame):
-        # DataFrames are stored as JSONB and we query them as such
-        value = Jsonb(value.to_json())
-        field = f"({field})::jsonb"
 
     if field.startswith("meta."):
         field = _treat_meta_field(field, value)
@@ -232,6 +227,20 @@ 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 +250,6 @@ COMPARISON_OPERATORS = {
     "<=": _less_than_equal,
     "in": _in,
     "not in": _not_in,
+    "like": _like,
+    "not like": _not_like,
 }

{pgvector_haystack-1.2.1.dist-info → pgvector_haystack-2.0.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pgvector-haystack
-Version: 1.2.1
+Version: 2.0.0
 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.dist-info → pgvector_haystack-2.0.0.dist-info}/RECORD

@@ -2,9 +2,9 @@ haystack_integrations/components/retrievers/pgvector/__init__.py,sha256=n4q4pUAp
 haystack_integrations/components/retrievers/pgvector/embedding_retriever.py,sha256=IEIdgKy_IWgmL5Qy_Hqkwa2mq2JxnQ_JOdjm38JPxmE,7530
 haystack_integrations/components/retrievers/pgvector/keyword_retriever.py,sha256=SYHqjTkDczobYZ2qq0z8yS9p_g8aINnzwA5AN_nKdHI,5632
 haystack_integrations/document_stores/pgvector/__init__.py,sha256=Ge-7ggarfr3x8UqNdFHUuFGiu52u49sOVe3l16WEwZU,197
-haystack_integrations/document_stores/pgvector/document_store.py,sha256=Ni9Cg9sXdsVBhuTNFhtSbJMkMNYORAlYLqcKM1lR96M,31440
-haystack_integrations/document_stores/pgvector/filters.py,sha256=751_Z1QnXJCG0GWC2jcfeZ25NNYfJewQEIUCeJS70PA,8212
-pgvector_haystack-1.2.1.dist-info/METADATA,sha256=qD2SeLMhjHrsrPvJRCojsCKg4TMl4HHM0xVg-87Qp14,2226
-pgvector_haystack-1.2.1.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
-pgvector_haystack-1.2.1.dist-info/licenses/LICENSE.txt,sha256=B05uMshqTA74s-0ltyHKI6yoPfJ3zYgQbvcXfDVGFf8,10280
-pgvector_haystack-1.2.1.dist-info/RECORD,,
+haystack_integrations/document_stores/pgvector/document_store.py,sha256=rirjSAMaPLADimwPcj-WQohrQbnXpTWgOCSIJy3qpPg,32148
+haystack_integrations/document_stores/pgvector/filters.py,sha256=M-_wmM1iKS1QdKhnRk27N8t2W1st2YZWAwVSFUyyTtk,8508
+pgvector_haystack-2.0.0.dist-info/METADATA,sha256=NhUGkaP8KzfX8qrlU5LrI-pjEQ3B4UDCfqN-WvlXsGY,2226
+pgvector_haystack-2.0.0.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+pgvector_haystack-2.0.0.dist-info/licenses/LICENSE.txt,sha256=B05uMshqTA74s-0ltyHKI6yoPfJ3zYgQbvcXfDVGFf8,10280
+pgvector_haystack-2.0.0.dist-info/RECORD,,