FEAT: Starting on medical information repository.

src/data/connection.py

@@ -16,6 +16,7 @@ class Collections:
 	SESSION = "sessions"
 	MEDICAL_RECORDS = "medical_records"
 	MEDICAL_MEMORY = "medical_memory"
+	INFORMATION = "chunks"
 
 class ActionFailed(Exception):
 	"""

src/data/repositories/information.py

@@ -0,0 +1,50 @@
+# src/data/repositories/information.py
+
+"""
+Contains all saved medical reference information to be used with RAG.
+This may need to be renamed.
+
+## Fields
+	chunk_id:
+	content:
+	embedding:
+	embedding_model:
+	embedding_dim:
+	metadata:
+		metadata.parent_id:
+		metadata.source:
+		metadata.task:
+		metadata.sequence:
+		metadata.total_chunks:
+		metadata.content_type:
+		metadata.related_chunks:
+		metadata.chunk_length:
+		metadata.created_timestamp:
+"""
+
+from pymongo.errors import ConnectionFailure, PyMongoError
+
+from src.data.connection import ActionFailed, Collections, get_collection
+from src.models.information import InfoChunk, InfoChunkMetadata
+from src.utils.logger import logger
+
+
+def get_chunk(
+	chunk_id: str,
+	*,
+	collection_name: str = Collections.INFORMATION
+) -> InfoChunk:
+	try:
+		collection = get_collection(collection_name)
+		cursor = collection.find(
+			{"chunk_id": chunk_id}
+		)
+
+		return InfoChunk.model_validate(cursor[0])
+	except (ConnectionFailure, PyMongoError) as e:
+		logger().error(f"Database error listing sessions for chunk '{chunk_id}': {e}")
+		raise ActionFailed("A database error occurred while retrieving chunk.") from e
+
+# TODO Embedding search
+
+# TODO Everything else :(

src/models/information.py

@@ -0,0 +1,39 @@
+# src/models/information.py
+
+from datetime import datetime
+
+from pydantic import BaseModel, ConfigDict, Field
+
+
+class InfoChunkMetadata(BaseModel):
+	"""Pydantic model for the nested metadata object."""
+	parent_id: str
+	source: str
+	task: str
+	sequence: int
+	total_chunks: int
+	content_type: str
+	related_chunks: list[str] | None = None
+	chunk_length: int | None = None
+	created_timestamp: datetime | None = None
+
+	model_config = ConfigDict(
+		frozen=True,
+		from_attributes=True,
+		populate_by_name=True
+	)
+
+class InfoChunk(BaseModel):
+	"""Pydantic model for the MongoDB collection."""
+	chunk_id: str
+	content: str
+	embedding: list[float]
+	embedding_model: str
+	embedding_dim: int = Field(..., gt=0)
+	metadata: InfoChunkMetadata
+
+	model_config = ConfigDict(
+		frozen=True,
+		from_attributes=True,
+		populate_by_name=True
+	)