Update app.py

app.py

@@ -1,55 +1,81 @@
 import gradio as gr
+import tempfile
+import shutil
+from pdf_extractor import extract_text_pdf_raw
+from word_extractor import extract_red_text_with_labels, is_red_font
 from docx import Document
 from docx.shared import RGBColor
-from difflib import get_close_matches
-from pdf_parser import extract_text_from_pdf, parse_data_blocks
+import difflib
 
-def is_red_color(run):
-    color = run.font.color
-    if not color or not color.rgb:
-        return False
-    r, g, b = color.rgb[0], color.rgb[1], color.rgb[2]
-    return r >= 200 and g <= 100 and b <= 100  # red-dominant
 
-def replace_red_text_with_data(doc_path, data_dict):
+def find_best_match(target, candidates):
+    match = difflib.get_close_matches(target, candidates, n=1, cutoff=0.5)
+    return match[0] if match else None
+
+
+def replace_red_text_in_doc(doc_path, replacements):
     doc = Document(doc_path)
 
     for para in doc.paragraphs:
         for run in para.runs:
-            if is_red_color(run):
-                original_text = run.text.strip()
-                # Try exact or close match
-                match = get_close_matches(original_text.lower(), [k.lower() for k in data_dict.keys()], n=1, cutoff=0.6)
-                if match:
-                    for key in data_dict:
-                        if key.lower() == match[0]:
-                            run.text = data_dict[key]
-                            break
-    return doc
-
-def process_files(pdf_file, word_template):
-    # Extract data from PDF
-    raw_text = extract_text_from_pdf(pdf_file)
-    data_dict = parse_data_blocks(raw_text)
-
-    # Replace red text in Word
-    final_doc = replace_red_text_with_data(word_template, data_dict)
-
-    # Save and return output
-    output_path = "filled_output.docx"
-    final_doc.save(output_path)
-    return output_path
-
-demo = gr.Interface(
+            if is_red_font(run):
+                old_text = run.text.strip()
+                new_text = find_best_match(old_text, replacements)
+                if new_text:
+                    run.text = new_text
+                    run.font.color.rgb = RGBColor(0, 0, 0)  # Set to black
+
+    for table in doc.tables:
+        for row in table.rows:
+            for cell in row.cells:
+                for para in cell.paragraphs:
+                    for run in para.runs:
+                        if is_red_font(run):
+                            old_text = run.text.strip()
+                            new_text = find_best_match(old_text, replacements)
+                            if new_text:
+                                run.text = new_text
+                                run.font.color.rgb = RGBColor(0, 0, 0)  # Set to black
+
+    temp_dir = tempfile.mkdtemp()
+    updated_path = f"{temp_dir}/updated.docx"
+    doc.save(updated_path)
+    return updated_path
+
+
+def process_files(pdf_file, word_file):
+    pdf_path = pdf_file.name
+    word_path = word_file.name
+
+    pdf_text = extract_text_pdf_raw(pdf_path)
+    word_data = extract_red_text_with_labels(word_path)
+
+    # Flatten red text entries
+    red_values = []
+    for values in word_data.values():
+        red_values.extend(values)
+    red_values = list(set(red_values))  # dedupe
+
+    # Match red values to PDF
+    replacements = []
+    for val in red_values:
+        match = find_best_match(val, pdf_text)
+        if match:
+            replacements.append(match)
+
+    # Replace in Word
+    updated_doc_path = replace_red_text_in_doc(word_path, replacements)
+
+    return updated_doc_path
+
+
+gr.Interface(
     fn=process_files,
     inputs=[
-        gr.File(label="Upload PDF Report", file_types=[".pdf"]),
-        gr.File(label="Upload Word Template (.docx)", file_types=[".docx"])
+        gr.File(label="Upload PDF File", type="file"),
+        gr.File(label="Upload Word File", type="file")
     ],
-    outputs=gr.File(label="Download Updated Word (.docx)"),
-    title="Audit Report Auto-Filler",
-    description="Replaces outdated red text in Word using updated values from a PDF report."
-)
-
-if __name__ == "__main__":
-    demo.launch()
+    outputs=gr.File(label="Download Updated Word File"),
+    title="Red Text Replacer",
+    description="Upload a PDF and Word document. Red-colored text in the Word doc will be replaced by matching content from the PDF."
+).launch()