Set use_rcts in build_index helper used by local index build. (#34110)

* Set `use_rcts` in `build_index` helper used by local index build.

Recursive Text Splitter is enabled by default for cloud index builds and effects how Markdown files are chunked. It preferences chunk length being close to the given `chunk_size` over splitting by heading sections.

* Udpate changelog.

* Move changelog to latest unreleased version.

Author: tot0

sdk/ai/azure-ai-generative/CHANGELOG.md

@@ -4,6 +4,7 @@
 
 ### Other Changes
 
+- Recursive Text Splitter is now enabled by default for local index builds which mimics behavior of cloud index builds
 - Bug fixes
 
 ## 1.0.0b5 (2024-02-02)

sdk/ai/azure-ai-generative/azure/ai/generative/index/_build_mlindex.py

@@ -4,14 +4,14 @@
 
 from pathlib import Path
 from typing import Dict, Optional, Union
-from packaging import version
 
 import yaml  # type: ignore[import]
+from packaging import version
 
+from azure.ai.resources._utils._open_ai_utils import build_open_ai_protocol
 from azure.ai.resources.entities.mlindex import Index
-from azure.ai.resources.operations._index_data_source import ACSSource, LocalSource
 from azure.ai.resources.operations._acs_output_config import ACSOutputConfig
-from azure.ai.resources._utils._open_ai_utils import build_open_ai_protocol
+from azure.ai.resources.operations._index_data_source import ACSSource, LocalSource
 
 
 def build_index(
@@ -60,6 +60,7 @@ def build_index(
     splitter_args= {
         'chunk_size': chunk_size,
         'chunk_overlap': chunk_overlap,
+        'use_rcts': True
     }
     if max_sample_files is not None:
         splitter_args["max_sample_files"] = max_sample_files