Upload folder using huggingface_hub

requirements.txt

@@ -13,5 +13,4 @@ pdfplumber
 pymupdf
 timm
 transformers
-img2table
-poppler-utils
+img2table

utils.py

@@ -18,14 +18,28 @@ def draw_boxes(image_path, boxes):
     return image
 
 
-def pdf_to_images(pdf_path):
-    images = convert_from_path(pdf_path)
+def pdf_to_images(
+    pdf_path, output_dir="extract_tables/table_outputs", output_format="png"
+):
+    if not os.path.exists(output_dir):
+        os.makedirs(output_dir)
+
+    pdf_document = fitz.open(pdf_path)
 
     image_paths = []
-    for idx, image in enumerate(images):
-        image_file_path = f"extract_tables/table_outputs/pdf-image-{idx + 1}.png"
-        image.save(image_file_path, format="PNG")
-        image_paths.append(image_file_path)
+    for page_num in range(len(pdf_document)):
+        page = pdf_document.load_page(page_num)
+        pix = page.get_pixmap(dpi=300)
+
+        image_file_path = os.path.join(
+            output_dir, f"pdf-image-{page_num + 1}.{output_format}"
+        )
+
+        try:
+            pix.save(image_file_path)
+            image_paths.append(image_file_path)
+        except Exception as e:
+            print(f"Error saving image {image_file_path}: {e}")
 
     return image_paths