fix: Introduced new index_document() to fix chunking related issue (#96)

* Introduced new index_document() to fix chunking related issue

* Minor deprecated version correction

Author: chandrasekharan-zipstack

src/unstract/sdk/__init__.py

@@ -1,4 +1,4 @@
-__version__ = "0.47.0"
+__version__ = "0.48.0"
 
 
 def get_sdk_version():

src/unstract/sdk/index.py

@@ -4,7 +4,7 @@
 
 from deprecated import deprecated
 from llama_index.core import Document
-from llama_index.core.node_parser import SimpleNodeParser
+from llama_index.core.node_parser import SentenceSplitter
 from llama_index.core.vector_stores import (
     FilterOperator,
     MetadataFilter,
@@ -199,7 +199,8 @@ def index(
                     self.tool.stream_log(f"No nodes found for {doc_id}")
             except Exception as e:
                 self.tool.stream_log(
-                    f"Error querying {vector_db_instance_id}: {e}", level=LogLevel.ERROR
+                    f"Error querying {vector_db_instance_id}: {e}, proceeding to index",
+                    level=LogLevel.ERROR,
                 )
 
             if doc_id_found and reindex:
@@ -288,7 +289,7 @@ def index(
 
             try:
                 if chunk_size == 0:
-                    parser = SimpleNodeParser.from_defaults(
+                    parser = SentenceSplitter.from_defaults(
                         chunk_size=len(documents[0].text) + 10,
                         chunk_overlap=0,
                         callback_manager=embedding.get_callback_manager(),
@@ -301,12 +302,6 @@ def index(
                     vector_db.add(doc_id, nodes=[node])
                     self.tool.stream_log("Added node to vector db")
                 else:
-                    storage_context = vector_db.get_storage_context()
-                    parser = SimpleNodeParser.from_defaults(
-                        chunk_size=chunk_size,
-                        chunk_overlap=chunk_overlap,
-                        callback_manager=embedding.get_callback_manager(),
-                    )
                     self.tool.stream_log("Adding nodes to vector db...")
                     # TODO: Phase 2:
                     # Post insertion to VDB, use query using doc_id and
@@ -318,13 +313,11 @@ def index(
                     # Once this is in place, the overridden implementation
                     # of prefixing ids with doc_id before adding to VDB
                     # can be removed
-                    vector_db.get_vector_store_index_from_storage_context(
+                    vector_db.index_document(
                         documents,
-                        storage_context=storage_context,
+                        chunk_size=chunk_size,
+                        chunk_overlap=chunk_overlap,
                         show_progress=True,
-                        embed_model=embedding,
-                        node_parser=parser,
-                        callback_manager=embedding.get_callback_manager(),
                     )
             except Exception as e:
                 self.tool.stream_log(

src/unstract/sdk/vector_db.py

@@ -5,6 +5,7 @@
 from deprecated import deprecated
 from llama_index.core import StorageContext, VectorStoreIndex
 from llama_index.core.indices.base import IndexType
+from llama_index.core.node_parser import SentenceSplitter
 from llama_index.core.schema import BaseNode, Document
 from llama_index.core.vector_stores.types import (
     BasePydanticVectorStore,
@@ -119,6 +120,33 @@ def _get_vector_db(self) -> Union[BasePydanticVectorStore, VectorStore]:
             )
             raise VectorDBError(f"Error getting vectorDB instance: {e}") from e
 
+    def index_document(
+        self,
+        documents: Sequence[Document],
+        chunk_size: int = 1024,
+        chunk_overlap: int = 128,
+        show_progress: bool = False,
+        **index_kwargs,
+    ) -> IndexType:
+        if not self._embedding_instance:
+            raise VectorDBError(self.EMBEDDING_INSTANCE_ERROR)
+        storage_context = self.get_storage_context()
+        parser = SentenceSplitter.from_defaults(
+            chunk_size=chunk_size,
+            chunk_overlap=chunk_overlap,
+            callback_manager=self._embedding_instance.callback_manager,
+        )
+        return VectorStoreIndex.from_documents(
+            documents,
+            storage_context=storage_context,
+            show_progress=show_progress,
+            embed_model=self._embedding_instance,
+            transformations=[parser],
+            callback_manager=self._embedding_instance.callback_manager,
+            **index_kwargs,
+        )
+
+    @deprecated(version="0.47.0", reason="Use index_document() instead")
     def get_vector_store_index_from_storage_context(
         self,
         documents: Sequence[Document],