add: pypdf2 loader text loader

medrag_multi_modal/__init__.py

@@ -2,12 +2,14 @@ from .document_loader import (
     ImageLoader,
     MarkerTextLoader,
     PyMuPDF4LLMTextLoader,
+    PyPDF2TextLoader,
     TextImageLoader,
 )
 from .retrieval import MultiModalRetriever
 
 __all__ = [
     "PyMuPDF4LLMTextLoader",
+    "PyPDF2TextLoader",
     "MarkerTextLoader",
     "ImageLoader",
     "TextImageLoader",

medrag_multi_modal/document_loader/__init__.py

@@ -1,12 +1,10 @@
 from .load_image import ImageLoader
 from .load_text_image import TextImageLoader
-from .text_loader import (
-    MarkerTextLoader,
-    PyMuPDF4LLMTextLoader,
-)
+from .text_loader import MarkerTextLoader, PyMuPDF4LLMTextLoader, PyPDF2TextLoader
 
 __all__ = [
     "PyMuPDF4LLMTextLoader",
+    "PyPDF2TextLoader",
     "MarkerTextLoader",
     "ImageLoader",
     "TextImageLoader",

medrag_multi_modal/document_loader/text_loader/__init__.py

@@ -1,7 +1,9 @@
 from .marker_text_loader import MarkerTextLoader
 from .pymupdf4llm_text_loader import PyMuPDF4LLMTextLoader
+from .pypdf2_text_loader import PyPDF2TextLoader
 
 __all__ = [
     "PyMuPDF4LLMTextLoader",
+    "PyPDF2TextLoader",
     "MarkerTextLoader",
 ]

medrag_multi_modal/document_loader/text_loader/pypdf2_text_loader.py

@@ -0,0 +1,21 @@
+from typing import Dict
+
+import PyPDF2
+
+from .base_text_loader import BaseTextLoader
+
+
+class PyPDF2TextLoader(BaseTextLoader):
+    async def _process_page(self, page_idx: int) -> Dict[str, str]:
+        with open(self.document_file_path, "rb") as file:
+            pdf_reader = PyPDF2.PdfReader(file)
+            page = pdf_reader.pages[page_idx]
+            text = page.extract_text()
+
+        return {
+            "text": text,
+            "page_idx": page_idx,
+            "document_name": self.document_name,
+            "file_path": self.document_file_path,
+            "file_url": self.url,
+        }