corvic-engine 0.3.0rc64__tar.gz → 0.3.0rc66__tar.gz

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/Cargo.lock

@@ -16,7 +16,7 @@ checksum = "baf1de4339761588bc0619e3cbc0120ee582ebb74b53b4efbf79117bd2da40fd"
 
 [[package]]
 name = "corvic-engine"
-version = "0.3.0-rc.64"
+version = "0.3.0-rc.66"
 dependencies = [
  "numpy",
  "pyo3",
@@ -57,7 +57,7 @@ checksum = "48c757948c5ede0e46177b7add2e67155f70e33c07fea8284df6576da70b3719"
 
 [[package]]
 name = "engine-walk"
-version = "0.3.0-rc.64"
+version = "0.3.0-rc.66"
 dependencies = [
  "version_check",
 ]
@@ -76,9 +76,9 @@ checksum = "f4c7245a08504955605670dbf141fceab975f15ca21570696aebe9d2e71576bd"
 
 [[package]]
 name = "libc"
-version = "0.2.171"
+version = "0.2.172"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c19937216e9d3aa9956d9bb8dfc0b0c8beb6058fc4f7a4dc4d850edf86a237d6"
+checksum = "d750af042f7ef4f724306de029d18836c26c1765a54a6a3f094cbd23a7267ffa"
 
 [[package]]
 name = "matrixmultiply"
@@ -180,9 +180,9 @@ dependencies = [
 
 [[package]]
 name = "proc-macro2"
-version = "1.0.94"
+version = "1.0.95"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a31971752e70b8b2686d7e46ec17fb38dad4051d94024c88df49b667caea9c84"
+checksum = "02b3e5e68a3a1a02aad3ec490a98007cbc13c37cbe84a3cd7b8e406d76e7f778"
 dependencies = [
  "unicode-ident",
 ]

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/Cargo.toml

@@ -3,7 +3,7 @@ resolver = "2"
 members = ["python"]
 
 [workspace.package]
-version = "0.3.0-rc.64"
+version = "0.3.0-rc.66"
 edition = "2021"
 publish = false
 

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: corvic-engine
-Version: 0.3.0rc64
+Version: 0.3.0rc66
 Classifier: Environment :: Console
 Classifier: License :: Other/Proprietary License
 Classifier: Programming Language :: Python :: Implementation :: CPython

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/pyproject.toml

@@ -20,7 +20,7 @@ authors = [
 ]
 readme = "README.md"
 description = "Seamless embedding generation and retrieval."
-version = "0.3.0-rc.64" # sync this with tool.poetry.version below
+version = "0.3.0-rc.66" # sync this with tool.poetry.version below
 
 # TODO(ddn): Pull dependencies from tool.poetry.dependencies. We use maturin as
 # the build system but poetry for the development environment.
@@ -74,7 +74,7 @@ packages = [
 ]
 description = ""
 authors = []
-version = "0.3.0-rc.64"
+version = "0.3.0-rc.66"
 
 [tool.poetry.scripts]
 check = "corvic_check.cli:main"
@@ -162,8 +162,8 @@ optional = true
 
 [tool.poetry.group.ml.dependencies]
 torch = [
-  { platform = "linux", version = "^2.4.0", source = "torch-cpu-linux" },
-  { platform = "darwin", version = "^2.4.0", source = "pypi" },
+  { platform = "linux", version = "^2.6.0", source = "torch-cpu-linux" },
+  { platform = "darwin", version = "^2.6.0", source = "pypi" },
 ]
 transformers = "^4.48.0"
 umap-learn = "^0.5.5"

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/python/corvic/op_graph/row_filters/_row_filters.py

@@ -118,8 +118,11 @@ def _make_compare_to_literal(
     )
 
 
+LiteralType = struct_pb2.Value | float | str | bool
+
+
 def eq(
-    column_name: str, literal: struct_pb2.Value | float | str | bool, dtype: pa.DataType
+    column_name: str, literal: LiteralType | None, dtype: pa.DataType
 ) -> CompareColumnToLiteral:
     """Include rows where column is equal to a literal."""
     return _make_compare_to_literal(
@@ -128,7 +131,7 @@ def eq(
 
 
 def ne(
-    column_name: str, literal: struct_pb2.Value | float | str | bool, dtype: pa.DataType
+    column_name: str, literal: LiteralType | None, dtype: pa.DataType
 ) -> CompareColumnToLiteral:
     """Include rows where column is not equal to a literal."""
     return _make_compare_to_literal(
@@ -137,7 +140,7 @@ def ne(
 
 
 def lt(
-    column_name: str, literal: struct_pb2.Value | float | str | bool, dtype: pa.DataType
+    column_name: str, literal: LiteralType, dtype: pa.DataType
 ) -> CompareColumnToLiteral:
     """Include rows where column is less than to a literal."""
     return _make_compare_to_literal(
@@ -146,7 +149,7 @@ def lt(
 
 
 def gt(
-    column_name: str, literal: struct_pb2.Value | float | str | bool, dtype: pa.DataType
+    column_name: str, literal: LiteralType, dtype: pa.DataType
 ) -> CompareColumnToLiteral:
     """Include rows where column is greater than to a literal."""
     return _make_compare_to_literal(
@@ -155,7 +158,7 @@ def gt(
 
 
 def le(
-    column_name: str, literal: struct_pb2.Value | float | str | bool, dtype: pa.DataType
+    column_name: str, literal: LiteralType, dtype: pa.DataType
 ) -> CompareColumnToLiteral:
     """Include rows where column is less than or equal to than a literal."""
     return _make_compare_to_literal(
@@ -164,7 +167,7 @@ def le(
 
 
 def ge(
-    column_name: str, literal: struct_pb2.Value | float | str | bool, dtype: pa.DataType
+    column_name: str, literal: LiteralType, dtype: pa.DataType
 ) -> CompareColumnToLiteral:
     """Include rows where column is greater than or equal to to a literal."""
     return _make_compare_to_literal(

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/python/corvic/system/in_memory_executor.py

@@ -467,24 +467,39 @@ class InMemoryExecutor(OpGraphExecutor):
         # not pl.PythonLiterals (e.g., struct, map) which means
         # the below can fail at runtime with a cryptic error.
         lit = cast(Any, row_filter.literal_as_py)
-        match row_filter.comparison_type:
-            case table_pb2.COMPARISON_TYPE_EQ:
-                comp = pl.col(row_filter.column_name) == lit
-            case table_pb2.COMPARISON_TYPE_NE:
-                comp = pl.col(row_filter.column_name) != lit
-            case table_pb2.COMPARISON_TYPE_LT:
-                comp = pl.col(row_filter.column_name) < lit
-            case table_pb2.COMPARISON_TYPE_GT:
-                comp = pl.col(row_filter.column_name) > lit
-            case table_pb2.COMPARISON_TYPE_LE:
-                comp = pl.col(row_filter.column_name) <= lit
-            case table_pb2.COMPARISON_TYPE_GE:
-                comp = pl.col(row_filter.column_name) >= lit
-            case _:
-                return op_graph.OpParseError(
-                    "unknown comparison type value in row filter",
-                    value=row_filter.comparison_type,
-                )
+        col = row_filter.column_name
+
+        # Handle comparison with None/null separately
+        if lit is None:
+            match row_filter.comparison_type:
+                case table_pb2.COMPARISON_TYPE_EQ:
+                    comp = pl.col(col).is_null()
+                case table_pb2.COMPARISON_TYPE_NE:
+                    comp = pl.col(col).is_not_null()
+                case _:
+                    return op_graph.OpParseError(
+                        "Unsupported literal None and comparison type",
+                        value=row_filter.comparison_type,
+                    )
+        else:
+            match row_filter.comparison_type:
+                case table_pb2.COMPARISON_TYPE_EQ:
+                    comp = pl.col(col) == lit
+                case table_pb2.COMPARISON_TYPE_NE:
+                    comp = pl.col(col) != lit
+                case table_pb2.COMPARISON_TYPE_LT:
+                    comp = pl.col(col) < lit
+                case table_pb2.COMPARISON_TYPE_GT:
+                    comp = pl.col(col) > lit
+                case table_pb2.COMPARISON_TYPE_LE:
+                    comp = pl.col(col) <= lit
+                case table_pb2.COMPARISON_TYPE_GE:
+                    comp = pl.col(col) >= lit
+                case _:
+                    return op_graph.OpParseError(
+                        "unknown comparison type value in row filter",
+                        value=row_filter.comparison_type,
+                    )
         return Ok(comp)
 
     def _row_filter_combination_to_condition(

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/python/corvic/system_sqlite/client.py

@@ -104,7 +104,7 @@ class StagingDBMaker(Protocol):
 class Client(corvic.system.Client):
     """Client for the sqlite system implementation."""
 
-    def __init__(
+    def __init__(  # noqa: PLR0913
         self,
         sqlite_file: pathlib.Path,
         *,
@@ -115,6 +115,8 @@ class Client(corvic.system.Client):
         staging_db_maker: StagingDBMaker | None = None,
         bucket_name: str | None = None,
         fs_blob_data_dir_name: str | None = None,
+        text_embedder: corvic.system.TextEmbedder | None = None,
+        image_embedder: corvic.system.ImageEmbedder | None = None,
     ):
         sqlite_file_exists = sqlite_file.exists()
 
@@ -152,8 +154,12 @@ class Client(corvic.system.Client):
                 vector_column_names_to_sizes
                 or corvic.system.DEFAULT_VECTOR_COLUMN_NAMES_TO_SIZES,
             )
-        self._text_embedder = corvic.system.RandomTextEmbedder()
-        self._image_embedder = corvic.system.CombinedImageEmbedder()
+        self._text_embedder = (
+            text_embedder if text_embedder else corvic.system.RandomTextEmbedder()
+        )
+        self._image_embedder = (
+            image_embedder if image_embedder else corvic.system.CombinedImageEmbedder()
+        )
         self._executor = corvic.system.ValidateFirstExecutor(
             corvic.system.InMemoryExecutor(
                 self._staging_db,

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/python/corvic/system_sqlite/staging.py

@@ -226,6 +226,9 @@ class DuckDBStaging(StagingDB):
         similarity_metric: VectorSimilarityMetric,
         input_vector: list[float],
     ):
+        if similarity_metric == "none":
+            return None
+
         distance_metric = "array_distance"
         descending = False
         if similarity_metric == "cosine":
@@ -386,9 +389,10 @@ class DuckDBStaging(StagingDB):
             vector_column, similarity_metric, input_vector
         )
 
-        vector_search_query = (
-            sqlglot.select(*columns).from_(vector_union_table).order_by(similarity_func)
-        )
+        vector_search_query = sqlglot.select(*columns).from_(vector_union_table)
+        if similarity_func:
+            vector_search_query = vector_search_query.order_by(similarity_func)
+
         return vector_search_query.with_(vector_union_table, as_=union)
 
     def run_select_query(

{corvic_engine-0.3.0rc64 → corvic_engine-0.3.0rc66}/python/corvic/well_known_types/__init__.py

@@ -39,7 +39,7 @@ def to_json(val: JSONAble) -> JSONExpressable:
     return val
 
 
-VectorSimilarityMetric: TypeAlias = Literal["cosine", "euclidean"]
+VectorSimilarityMetric: TypeAlias = Literal["cosine", "euclidean", "none"]
 
 
 __all__ = [