update,

goodtastesmartie/task1-deal-broken/ocr/Pdfredaction.py

@@ -0,0 +1,95 @@
+import re
+import argparse
+import fitz
+import pytesseract
+import cv2
+import numpy as np
+from reportlab.lib.pagesizes import letter
+from reportlab.pdfgen import canvas
+
+
+def redact_text(text, validators, ignored):
+    redacted_text = text
+
+    for validator_type, patterns in validators.items():
+        for pattern, replacement in patterns:
+            redacted_text, _ = re.subn(pattern, replacement, redacted_text, flags=re.IGNORECASE)
+            print(f"{validator_type}: {redacted_text}")
+
+    for pattern in ignored:
+        redacted_text, _ = re.subn(pattern, '', redacted_text, flags=re.IGNORECASE)
+        print(f"Ignored: {redacted_text}")
+
+    return redacted_text
+
+
+def process_pdf(input_file, output_file, validators, ignored):
+    doc = fitz.open(input_file)
+
+    for page_index, page in enumerate(doc):
+        pixmap = page.get_pixmap()
+
+        image_np = np.frombuffer(pixmap.samples, dtype=np.uint8).reshape(pixmap.h, pixmap.w, pixmap.n)
+        image_rgb = cv2.cvtColor(image_np, cv2.COLOR_BGR2RGB)
+        extracted_text = pytesseract.image_to_string(image_rgb)
+        redacted_text = redact_text(extracted_text, validators, ignored)
+
+        c = canvas.Canvas("temp.pdf", pagesize=letter)
+        c.setPageSize((pixmap.w, pixmap.h))
+
+        c.setFont("Helvetica", 12)
+        for line in redacted_text.split("\n"):
+            if line.strip():
+                c.setFillColorRGB(0, 0, 0)
+                c.rect(0, pixmap.h - 16, pixmap.w, 16, fill=True)
+                c.setFillColorRGB(1, 1, 1)
+                c.drawString(6, pixmap.h - 11, line)
+
+            c.showPage()
+
+        c.save()
+
+        redacted_page = fitz.open("temp.pdf")
+        page_text = redacted_page[0].get_text()
+
+        if page_text:
+            page.insert_text((0, 0), page_text)
+
+    doc.save(output_file)
+    doc.close()
+    redacted_page.close()
+
+    print("Redacted PDF saved successfully!")
+
+
+# Define the validators and ignored patterns
+validators = {
+    "Hong Kong Identity Card Number": [
+        (r'\b[A-Z]\d{6}\([A-Z]\)', 'XXXXXXXXX'),
+        (r'\b[A-Z]\d{6}\(\d\)', 'XXXXXXXXX')
+    ],
+    "Phone Numbers": [
+        (r'\+\d{1,3}[-\s]?\d{1,4}[-\s]?\d{1,4}[-\s]?\d{1,9}', '+XXX XXXX XXXX'),
+        (r'\d{1,4}[-\s]?\d{1,4}[-\s]?\d{1,9}', 'XXXX XXXX')
+    ],
+    "Credit Card Numbers": [
+        (r'\b[0-9]{4}-[0-9]{4}-[0-9]{4}-[0-9]{4}\b', 'XXXX-XXXX-XXXX-XXXX')
+    ],
+    "Email Addresses": [
+        (r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}\b', 'XXXXX@XXXXX.XXX')
+    ],
+}
+
+ignored = [
+    r'0000[-\s]?0000[-\s]?0000[-\s]?0000',
+    r'1111[-\s]?1111[-\s]?1111[-\s]?1111'
+]
+
+# Parse command-line arguments
+parser = argparse.ArgumentParser(description='PDF Redaction Script')
+parser.add_argument('input_file', help='Path to the input PDF file')
+parser.add_argument('output_file', help='Path to the output PDF file')
+args = parser.parse_args()
+
+# Call the process_pdf function with input and output file paths, validators, and ignored patterns
+process_pdf(args.input_file, args.output_file, validators, ignored)