How to redact a PDF permanently (and why highlighting isn't enough)

pip install pymupdf

import fitz  # PyMuPDF
 
def redact_text_in_pdf(input_path: str, output_path: str, search_term: str) -> int:
    doc = fitz.open(input_path)
    total_redactions = 0
 
    for page in doc:
        # Find all instances of the search term
        instances = page.search_for(search_term)
        for rect in instances:
            # Add a redaction annotation with a black fill
            page.add_redact_annot(rect, fill=(0, 0, 0))
            total_redactions += 1
 
        # Apply redactions — this permanently removes the underlying text
        page.apply_redactions()
 
    doc.save(output_path, garbage=4, deflate=True)
    doc.close()
    return total_redactions
 
count = redact_text_in_pdf("contract.pdf", "contract_redacted.pdf", "John Smith")
print(f"Redacted {count} instance(s)")

import fitz
import re
 
PATTERNS = {
    "ssn": r"\b\d{3}-\d{2}-\d{4}\b",
    "email": r"\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b",
    "phone": r"\b(\+\d{1,3}[\s.-])?\(?\d{3}\)?[\s.-]\d{3}[\s.-]\d{4}\b",
}
 
def redact_patterns(input_path: str, output_path: str) -> dict:
    doc = fitz.open(input_path)
    counts = {k: 0 for k in PATTERNS}
 
    for page in doc:
        # Extract text with position data
        words = page.get_text("words")  # returns (x0, y0, x1, y1, word, ...)
        full_text = page.get_text("text")
 
        for pattern_name, pattern in PATTERNS.items():
            for match in re.finditer(pattern, full_text):
                # Search for the matched string on the page
                instances = page.search_for(match.group())
                for rect in instances:
                    page.add_redact_annot(rect, fill=(0, 0, 0))
                    counts[pattern_name] += 1
 
        page.apply_redactions()
 
    doc.save(output_path, garbage=4, deflate=True)
    doc.close()
    return counts
 
result = redact_patterns("document.pdf", "document_redacted.pdf")
print(result)
# {'ssn': 3, 'email': 7, 'phone': 2}

import fitz
 
def scrub_metadata(input_path: str, output_path: str) -> None:
    doc = fitz.open(input_path)
    doc.set_metadata({
        "author": "",
        "creator": "",
        "producer": "",
        "subject": "",
        "title": "",
        "keywords": "",
        "creationDate": "",
        "modDate": "",
    })
    doc.save(output_path, garbage=4, deflate=True, clean=True)
    doc.close()

import { PDFDocument, rgb, degrees } from "pdf-lib";
import fs from "fs";
 
interface RedactionRect {
  page: number; // 0-indexed
  x: number;
  y: number;
  width: number;
  height: number;
}
 
async function redactPdf(
  inputPath: string,
  outputPath: string,
  redactions: RedactionRect[]
): Promise<void> {
  const pdfBytes = fs.readFileSync(inputPath);
  const pdfDoc = await PDFDocument.load(pdfBytes);
  const pages = pdfDoc.getPages();
 
  for (const rect of redactions) {
    const page = pages[rect.page];
    const { height: pageHeight } = page.getSize();
 
    // PDF coordinate system: origin is bottom-left
    // Convert from top-left coordinates to bottom-left
    page.drawRectangle({
      x: rect.x,
      y: pageHeight - rect.y - rect.height,
      width: rect.width,
      height: rect.height,
      color: rgb(0, 0, 0),
      opacity: 1,
    });
  }
 
  const outputBytes = await pdfDoc.save();
  fs.writeFileSync(outputPath, outputBytes);
}
 
// Example: redact a 200x20 rectangle on page 0, 50px from top, 100px from left
await redactPdf("input.pdf", "output.pdf", [
  { page: 0, x: 100, y: 50, width: 200, height: 20 },
]);

// Instead of embedding a full credit card number in the PDF
const invoiceData = {
  customer_name: "Jane Doe",
  card_number: "4111111111111111", // ❌ don't embed this
};
 
// Mask it before passing to the PDF template
const invoiceData = {
  customer_name: "Jane Doe",
  card_last_four: "1111", // ✅ only embed what's needed
};
 
const response = await fetch("https://pdf4.dev/api/v1/render", {
  method: "POST",
  headers: {
    Authorization: "Bearer p4_live_xxx",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    template_id: "invoice",
    data: invoiceData,
  }),
});