adapta 3.2.4a477.dev3__tar.gz → 3.2.4a477.dev5__tar.gz

{adapta-3.2.4a477.dev3 → adapta-3.2.4a477.dev5}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: adapta
-Version: 3.2.4a477.dev3
+Version: 3.2.4a477.dev5
 Summary: Logging, data connectors, monitoring, secret handling and general lifehacks to make data people lives easier.
 Home-page: https://github.com/SneaksAndData/adapta
 License: Apache 2.0

adapta-3.2.4a477.dev5/adapta/_version.py

@@ -0,0 +1 @@
+__version__ = 'v3.2.4a477.dev5'

{adapta-3.2.4a477.dev3 → adapta-3.2.4a477.dev5}/adapta/storage/distributed_object_store/v3/datastax_astra/README.md

@@ -132,17 +132,19 @@ with AstraClient(
 1. Create a table in Astra and insert some rows:
 ```cassandraql
 CREATE TABLE IF NOT EXISTS tmp.test_entity_with_embeddings (
-    col_a TEXT,
+    col_a TEXT PRIMARY KEY,
     col_b TEXT,
     col_c VECTOR<FLOAT, 3>,
     col_d TEXT,
-    PRIMARY KEY ((col_a, col_b))
 );
 
 CREATE INDEX IF NOT EXISTS ann_index
   ON tmp.test_entity_with_embeddings(col_c)
   WITH OPTIONS = {'source_model': 'other'};
 
+CREATE INDEX IF NOT EXISTS col_b_index
+  ON tmp.test_entity_with_embeddings(col_b);
+
 INSERT INTO tmp.test_entity_with_embeddings (col_a, col_b, col_c, col_d)
 VALUES ('something1', 'different', [0.3, 0.4, 0.5], 'extra1');
 
@@ -164,10 +166,7 @@ class TestEntityWithEmbeddings:
         "is_primary_key": True,
         "is_partition_key": True
     })
-    col_b: str = field(metadata={
-        "is_primary_key": True,
-        "is_partition_key": True
-    })
+    col_b: str
     col_c: list[float] = field(metadata={
         "is_vector_enabled": True
     })
@@ -212,7 +211,7 @@ with astra_client:
 
 
 # Search with primary key filter in Astra (with Expression)
-filter_expression = (FilterField('col_a') == 'something2') & (FilterField('col_b') == 'different1')
+filter_expression = (FilterField('col_a') == 'something2') & (FilterField('col_b').isin(['different1', 'doesnt_exist']))
 with astra_client:
     print(astra_client.ann_search(
         entity_type=TestEntityWithEmbeddings,

{adapta-3.2.4a477.dev3 → adapta-3.2.4a477.dev5}/adapta/storage/distributed_object_store/v3/datastax_astra/_models.py

@@ -15,11 +15,17 @@
 #  See the License for the specific language governing permissions and
 #  limitations under the License.
 #
+import re
 
 from enum import Enum
 from typing import Optional, Union, List, Dict, Any
 
-from adapta.storage.models.filter_expression import Expression, compile_expression, AstraFilterExpression
+from adapta.storage.models.filter_expression import (
+    Expression,
+    compile_expression,
+    AstraFilterExpression,
+    FilterExpressionOperation,
+)
 
 
 class SimilarityFunction(Enum):
@@ -72,6 +78,35 @@ class VectorSearchQuery:
         if self._key_column_filter_values is None:
             return ""
 
+        def format_value_for_cql(val: Any) -> str:
+            if isinstance(val, (tuple, list, set)):
+                return f"({', '.join(format_value_for_cql(v) for v in val)})"
+            if isinstance(val, str):
+                return f"'{val}'"
+            return str(val)
+
+        def get_astra_operator(column_expression: str) -> str:
+            operator_map = {
+                FilterExpressionOperation.GT.value["astra"]: ">",
+                FilterExpressionOperation.GE.value["astra"]: ">=",
+                FilterExpressionOperation.LT.value["astra"]: "<",
+                FilterExpressionOperation.LE.value["astra"]: "<=",
+                FilterExpressionOperation.IN.value["astra"]: "IN",
+            }
+
+            for suffix, operator in operator_map.items():
+                if column_expression.endswith(suffix):
+                    return operator
+
+            return "="
+
+        def remove_operator_suffix(col: str) -> str:
+            """
+            Removes the double underscore and the subsequent operator suffix from a column expression.
+            Returns only the column name.
+            """
+            return re.sub(r"__\w+$", "", col)
+
         compiled_filter_values = (
             compile_expression(self._key_column_filter_values, AstraFilterExpression)
             if isinstance(self._key_column_filter_values, Expression)
@@ -80,10 +115,11 @@ class VectorSearchQuery:
         if len(compiled_filter_values) > 1:
             raise ValueError("Restriction on key columns must not be nested under OR operator")
 
-        compiled_filter_values = {
-            col: f"'{val}'" if isinstance(val, str) else val for col, val in compiled_filter_values[0].items()
-        }
-        return f"where {' and '.join([f'{col} = {val}' for col, val in compiled_filter_values.items()])}"
+        cql_filter_expressions = [
+            f"{remove_operator_suffix(col)} {get_astra_operator(col)} {format_value_for_cql(val)}"
+            for col, val in compiled_filter_values[0].items()
+        ]
+        return f"where {' and '.join(cql_filter_expressions)}"
 
     def __str__(self):
         return " ".join(

{adapta-3.2.4a477.dev3 → adapta-3.2.4a477.dev5}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "adapta"
-version = "v3.2.4a477.dev3"
+version = "v3.2.4a477.dev5"
 description = "Logging, data connectors, monitoring, secret handling and general lifehacks to make data people lives easier."
 authors = ["ECCO Sneaks & Data <esdsupport@ecco.com>"]
 maintainers = ['GZU <gzu@ecco.com>', 'JRB <ext-jrb@ecco.com>']

adapta-3.2.4a477.dev3/adapta/_version.py

@@ -1 +0,0 @@
-__version__ = 'v3.2.4a477.dev3'