diff --git a/semantic_router/index/pinecone.py b/semantic_router/index/pinecone.py
index 91726e44480d0a19fce17c078186c28d51792cdf..9042eafe4ad4b9360b60120030c92107fdeb4f2e 100644
--- a/semantic_router/index/pinecone.py
+++ b/semantic_router/index/pinecone.py
@@ -499,31 +499,16 @@ class PineconeIndex(BaseIndex):
return all_vector_ids, metadata
def get_routes(self) -> List[Tuple]:
- """
- Gets a list of route and utterance objects currently stored in the index, including additional metadata.
+ """Gets a list of route and utterance objects currently stored in the
+ index, including additional metadata.
- Returns:
- List[Tuple]: A list of tuples, each containing route, utterance, function schema and additional metadata.
+ :return: A list of tuples, each containing route, utterance, function
+ schema and additional metadata.
+ :rtype: List[Tuple]
"""
_, metadata = self._get_all(include_metadata=True)
- route_tuples = [
- (
- data.get("sr_route", ""),
- data.get("sr_utterance", ""),
- (
- json.loads(data["sr_function_schema"])
- if data.get("sr_function_schema", "")
- else {}
- ),
- {
- key: value
- for key, value in data.items()
- if key not in ["sr_route", "sr_utterance", "sr_function_schema"]
- },
- )
- for data in metadata
- ]
- return route_tuples # type: ignore
+ route_tuples = parse_route_info(metadata=metadata)
+ return route_tuples
def delete(self, route_name: str):
route_vec_ids = self._get_route_ids(route_name=route_name)
@@ -553,8 +538,7 @@ class PineconeIndex(BaseIndex):
route_filter: Optional[List[str]] = None,
**kwargs: Any,
) -> Tuple[np.ndarray, List[str]]:
- """
- Search the index for the query vector and return the top_k results.
+ """Search the index for the query vector and return the top_k results.
:param vector: The query vector to search for.
:type vector: np.ndarray
@@ -633,11 +617,11 @@ class PineconeIndex(BaseIndex):
return np.array(scores), route_names
async def aget_routes(self) -> list[tuple]:
- """
- Asynchronously get a list of route and utterance objects currently stored in the index.
+ """Asynchronously get a list of route and utterance objects currently
+ stored in the index.
- Returns:
- List[Tuple]: A list of (route_name, utterance) objects.
+ :return: A list of (route_name, utterance) objects.
+ :rtype: List[Tuple]
"""
if self.async_client is None or self.host is None:
raise ValueError("Async client or host are not initialized.")
@@ -703,8 +687,15 @@ class PineconeIndex(BaseIndex):
async def _async_get_all(
self, prefix: Optional[str] = None, include_metadata: bool = False
) -> tuple[list[str], list[dict]]:
- """
- Retrieves all vector IDs from the Pinecone index using pagination asynchronously.
+ """Retrieves all vector IDs from the Pinecone index using pagination
+ asynchronously.
+
+ :param prefix: The prefix to filter the vectors by.
+ :type prefix: Optional[str]
+ :param include_metadata: Whether to include metadata in the response.
+ :type include_metadata: bool
+ :return: A tuple containing a list of vector IDs and a list of metadata dictionaries.
+ :rtype: tuple[list[str], list[dict]]
"""
if self.index is None:
raise ValueError("Index is None, could not retrieve vector IDs.")
@@ -754,8 +745,13 @@ class PineconeIndex(BaseIndex):
return all_vector_ids, metadata
async def _async_fetch_metadata(self, vector_id: str) -> dict:
- """
- Fetch metadata for a single vector ID asynchronously using the async_client.
+ """Fetch metadata for a single vector ID asynchronously using the
+ async_client.
+
+ :param vector_id: The ID of the vector to fetch metadata for.
+ :type vector_id: str
+ :return: A dictionary containing the metadata for the vector.
+ :rtype: dict
"""
url = f"https://{self.host}/vectors/fetch"
@@ -786,31 +782,44 @@ class PineconeIndex(BaseIndex):
)
async def _async_get_routes(self) -> List[Tuple]:
- """
- Asynchronously gets a list of route and utterance objects currently stored in the index, including additional metadata.
+ """Asynchronously gets a list of route and utterance objects currently
+ stored in the index, including additional metadata.
- Returns:
- List[Tuple]: A list of tuples, each containing route, utterance, function schema and additional metadata.
+ :return: A list of tuples, each containing route, utterance, function
+ schema and additional metadata.
+ :rtype: List[Tuple]
"""
_, metadata = await self._async_get_all(include_metadata=True)
- route_info = [
- (
- data.get("sr_route", ""),
- data.get("sr_utterance", ""),
- (
- json.loads(data["sr_function_schema"])
- if data.get("sr_function_schema", "")
- else {}
- ),
- {
- key: value
- for key, value in data.items()
- if key not in ["sr_route", "sr_utterance", "sr_function_schema"]
- },
- )
- for data in metadata
- ]
+ route_info = parse_route_info(metadata=metadata)
return route_info # type: ignore
def __len__(self):
return self.index.describe_index_stats()["total_vector_count"]
+
+
+
+def parse_route_info(metadata: List[Dict[str, Any]]) -> List[Tuple]:
+ """Parses metadata from Pinecone index to extract route, utterance, function
+ schema and additional metadata.
+
+ :param metadata: List of metadata dictionaries.
+ :type metadata: List[Dict[str, Any]]
+ :return: A list of tuples, each containing route, utterance, function schema and additional metadata.
+ :rtype: List[Tuple]
+ """
+ route_info = []
+ for record in metadata:
+ sr_route = record.get("sr_route", "")
+ sr_utterance = record.get("sr_utterance", "")
+ sr_function_schema = json.loads(record.get("sr_function_schema", "{}"))
+ if sr_function_schema == {}:
+ sr_function_schema = None
+
+ additional_metadata = {
+ key: value
+ for key, value in record.items()
+ if key not in ["sr_route", "sr_utterance", "sr_function_schema"]
+ }
+ # TODO: Not a fan of tuple packing here
+ route_info.append((sr_route, sr_utterance, sr_function_schema, additional_metadata))
+ return route_info
diff --git a/semantic_router/layer.py b/semantic_router/layer.py
index 0bf7d99f4809e08ba490832153da1498a31f8a10..5991c4686cfefe4a014489b902cd63a4f366a0f3 100644
--- a/semantic_router/layer.py
+++ b/semantic_router/layer.py
@@ -219,6 +219,7 @@ class RouteLayer:
# if routes list has been passed, we initialize index now
if self.index.sync:
# initialize index now
+ logger.info(f"JB TEMP: {self.routes=}")
if len(self.routes) > 0:
self._add_and_sync_routes(routes=self.routes)
else:
@@ -544,15 +545,20 @@ class RouteLayer:
)
# Update local route layer state
- self.routes = [
- Route(
- name=route,
- utterances=data.get("utterances", []),
- function_schemas=[data.get("function_schemas", None)],
- metadata=data.get("metadata", {}),
+ logger.info([data.get("function_schemas", None) for _, data in layer_routes_dict.items()])
+ self.routes = []
+ for route, data in layer_routes_dict.items():
+ function_schemas = data.get("function_schemas", None)
+ if function_schemas is not None:
+ function_schemas = [function_schemas]
+ self.routes.append(
+ Route(
+ name=route,
+ utterances=data.get("utterances", []),
+ function_schemas=function_schemas,
+ metadata=data.get("metadata", {}),
+ )
)
- for route, data in layer_routes_dict.items()
- ]
def _extract_routes_details(
self, routes: List[Route], include_metadata: bool = False
diff --git a/semantic_router/llms/openai.py b/semantic_router/llms/openai.py
index f22f409e8c59e830d9c1f20059c4be8d3bc10a27..4c49974a52a78f73c9e099cee4cd048bf4fd3cd8 100644
--- a/semantic_router/llms/openai.py
+++ b/semantic_router/llms/openai.py
@@ -92,8 +92,11 @@ class OpenAILLM(BaseLLM):
raise ValueError("OpenAI client is not initialized.")
try:
tools: Union[List[Dict[str, Any]], NotGiven] = (
- function_schemas if function_schemas is not None else NOT_GIVEN
+ function_schemas if function_schemas else NOT_GIVEN
)
+ logger.info(f"{function_schemas=}")
+ logger.info(f"{function_schemas is None=}")
+ logger.info(f"{tools=}")
completion = self.client.chat.completions.create(
model=self.name,