Parse a PDF

This guide shows you how to load existing PDFs and extract information from them.

Load a PDF

Pass a Uint8Array containing the PDF bytes to PDF.load():

import { PDF } from "@libpdf/core";

const pdf = await PDF.load(bytes);

From a File (Node.js / Bun)

import { readFile } from "fs/promises";
import { PDF } from "@libpdf/core";

const bytes = await readFile("document.pdf");
const pdf = await PDF.load(bytes);

From a URL (Browser)

import { PDF } from "@libpdf/core";

const response = await fetch("/document.pdf");
const bytes = new Uint8Array(await response.arrayBuffer());
const pdf = await PDF.load(bytes);

Encrypted PDFs

If the PDF is password-protected, provide the password:

const pdf = await PDF.load(bytes, {
  credentials: "secret"
});

Both user and owner passwords work. If you provide the owner password, you get full access to the document regardless of permission restrictions.

Inspect the Document

Page Count

console.log(pdf.getPageCount()); // e.g., 5

Metadata

console.log(pdf.getTitle());    // "Annual Report 2024"
console.log(pdf.getAuthor());   // "John Doe"
console.log(pdf.getSubject());  // "Financial Results"
console.log(pdf.getKeywords()); // ["finance", "quarterly"]
console.log(pdf.getCreator());  // "Microsoft Word"
console.log(pdf.getProducer()); // "@libpdf/core"

Page Dimensions

const page = await pdf.getPage(0); // Zero-indexed
const { width, height } = page;

console.log(`${width} x ${height} points`);
// e.g., "612 x 792 points" for US Letter

Common page sizes in points:

Size	Width	Height
Letter	612	792
A4	595	842
Legal	612	1008

Rotation

const rotation = page.rotation; // 0, 90, 180, or 270

Extract Text

Simple Extraction

Get all text from a page as a single string:

const page = await pdf.getPage(0);
const pageText = await page.extractText();

console.log(pageText.text);

Text with Positions

Get text lines with their coordinates for more advanced processing:

const pageText = await page.extractText();

for (const line of pageText.lines) {
  console.log(`"${line.text}" at (${line.bbox.x}, ${line.bbox.y})`);
}

Each line includes:

text - the text content
bbox - bounding box with x, y, width, height
spans - individual text spans with font information

Search Text

Find text matching a pattern:

// String search
const results = await page.findText("invoice");

// Regex search
const invoiceNumbers = await page.findText(/INV-\d{6}/g);

for (const match of results) {
  console.log(`Found "${match.text}" at`, match.bbox);
}

Iterate Over Pages

const pages = await pdf.getPages();

for (const page of pages) {
  const pageText = await page.extractText();
  console.log(`Page ${page.index + 1}: ${pageText.text.slice(0, 100)}...`);
}

Check for Forms

const form = await pdf.getForm();

if (form && !form.isEmpty) {
  const fields = form.getFields();
  console.log(`Found ${fields.length} form fields`);
}

See the Forms Guide for working with form fields.

Check Encryption Status

if (pdf.isEncrypted) {
  console.log("Document is encrypted");
  console.log("Permissions:", pdf.getPermissions());
}

Error Handling

import { PDF, SecurityError } from "@libpdf/core";

try {
  const pdf = await PDF.load(bytes);
} catch (error) {
  if (error instanceof SecurityError) {
    console.log("Password required or incorrect");
  } else if (error instanceof Error) {
    console.log("Invalid or corrupted PDF:", error.message);
  } else {
    throw error;
  }
}

Next Steps

Create a PDF

Generate PDFs from scratch

Text Extraction

Advanced text extraction techniques

Forms

Read and fill form fields

Parse a PDF

Create a PDF

Text Extraction

Forms

On this page