improve the OCR pipeline processing and outputs formatting

setup.py

@@ -1,15 +1,9 @@
-from setuptools import setup, find_packages
+"""Legacy setuptools compatibility shim.
 
-setup(
-    name="kg_ocr",
-    packages=find_packages(),
-    python_requires=">=3.10",
-    install_requires=[
-        "pytesseract",
-        "Pillow",
-        "txtai",
-        "sentence-transformers",
-        "litellm",
-        "python-dotenv"
-    ],
-)
+Project metadata, dependencies, and the src package layout are maintained solely in
+pyproject.toml. Install with `pip install .` or `uv sync`.
+"""
+
+from setuptools import setup
+
+setup()