Switch to agent arch

src/rag_eval/vector_search/base.py

@@ -0,0 +1,68 @@
+"""Abstract base class for vector search providers."""
+
+from abc import ABC, abstractmethod
+from typing import Any, TypedDict
+
+
+class SearchResult(TypedDict):
+    """A single vector search result."""
+
+    id: str
+    distance: float
+    content: str
+
+
+class BaseVectorSearch(ABC):
+    """Abstract base class for a vector search provider.
+
+    This class defines the standard interface for creating a vector search
+    index and running queries against it.
+    """
+
+    @abstractmethod
+    def create_index(
+        self, name: str, content_path: str, **kwargs: Any  # noqa: ANN401
+    ) -> None:
+        """Create a new vector search index with the provided content.
+
+        Args:
+            name: The desired name for the new index.
+            content_path: Path to the data used to populate the index.
+            **kwargs: Additional provider-specific arguments.
+
+        """
+        ...
+
+    @abstractmethod
+    def update_index(
+        self, index_name: str, content_path: str, **kwargs: Any  # noqa: ANN401
+    ) -> None:
+        """Update an existing vector search index with new content.
+
+        Args:
+            index_name: The name of the index to update.
+            content_path: Path to the data used to populate the index.
+            **kwargs: Additional provider-specific arguments.
+
+        """
+        ...
+
+    @abstractmethod
+    def run_query(
+        self,
+        deployed_index_id: str,
+        query: list[float],
+        limit: int,
+    ) -> list[SearchResult]:
+        """Run a similarity search query against the index.
+
+        Args:
+            deployed_index_id: The ID of the deployed index.
+            query: The embedding vector for the search query.
+            limit: Maximum number of nearest neighbors to return.
+
+        Returns:
+            A list of matched items with id, distance, and content.
+
+        """
+        ...