Update app.py

app.py

@@ -1,40 +1,47 @@
 import gradio as gr
-import os
-from docx import Document
-from utils import extract_text_from_pdf, parse_pdf_to_dict
-
-from docx.shared import RGBColor
-
 from docx import Document
 from docx.shared import RGBColor
 from difflib import get_close_matches
+from utils import extract_text_from_pdf, parse_pdf_to_dict
 
+# Improved red color detection
 def is_red_color(run):
     color = run.font.color
     if color is None:
         return False
     if color.rgb:
-        r, g, b = color.rgb[0], color.rgb[1], color.rgb[2]
-        return r > 180 and g < 100 and b < 100  # any strong red tint
+        # Try detecting red tones even if slightly off
+        red_shades = [
+            RGBColor(255, 0, 0),
+            RGBColor(200, 0, 0),
+            RGBColor(255, 20, 20),
+            RGBColor(192, 0, 0),
+            RGBColor(220, 20, 60),
+        ]
+        return any(color.rgb == shade for shade in red_shades)
     return False
-    
 
 def replace_red_text_with_data(doc_path, data_dict):
     doc = Document(doc_path)
+    matched = 0
 
     for para in doc.paragraphs:
         for run in para.runs:
             if is_red_color(run):
                 red_text = run.text.strip()
-                # Try fuzzy match
-                key_match = get_close_matches(red_text.lower(), [k.lower() for k in data_dict.keys()], n=1, cutoff=0.6)
+                print(f"Found red text: {red_text}")
+                if not red_text:
+                    continue
+                # Use fuzzy matching
+                key_match = get_close_matches(red_text.lower(), [k.lower() for k in data_dict], n=1, cutoff=0.6)
                 if key_match:
-                    # Find original key with matching text
                     for key in data_dict:
                         if key.lower() == key_match[0]:
+                            print(f"Replacing '{red_text}' with '{data_dict[key]}'")
                             run.text = data_dict[key]
+                            matched += 1
                             break
-
+    print(f"Total replacements: {matched}")
     return doc
 
 def process_files(pdf_file, template_docx):
@@ -42,14 +49,11 @@ def process_files(pdf_file, template_docx):
     doc_path = template_docx
     output_path = "filled_output.docx"
 
-    # Extract and parse PDF
     raw_text = extract_text_from_pdf(pdf_path)
     data_dict = parse_pdf_to_dict(raw_text)
 
-    # Replace red text with data
     final_doc = replace_red_text_with_data(doc_path, data_dict)
 
-    # Save final document
     final_doc.save(output_path)
     return output_path
 
@@ -61,7 +65,7 @@ demo = gr.Interface(
     ],
     outputs=gr.File(label="Download Filled Report (.docx)"),
     title="Audit Report Generator",
-    description="Upload a PDF and a Word template. This tool will auto-fill red-highlighted fields with data from the PDF."
+    description="Upload a PDF and a Word template. This tool will auto-fill red-colored outdated text with data from the PDF."
 )
 
 if __name__ == "__main__":