Processing through pdf-lib causes xref/corrupt message

[jackwshepherd]

Hello

Thanks once again for supporting such a fantastic library. I've always had trouble with the below PDF.

https://tinyurl.com/y9cwewbz

It appears that whenever I process it through pdf-lib, it opens in every PDF Reader except Adobe Reader. I think it's quite important it works in Adobe Reader.

When I run qpdf on it, I get the below error:

WARNING: /Users/me/Downloads/corrupt.pdf: file is damaged
WARNING: /Users/me/Downloads/corrupt.pdf (offset 6711353): xref not found
WARNING: /Users/me/Downloads/corrupt.pdf: Attempting to reconstruct cross-reference table

My code is below, in case relevant.

Thanks a lot

// Get the bundle
const bundle = req.project.bundles[req.bundleKey];
// Get list of all entries
const entries = bundle.sections.reduce(
  (acc, section) => [...acc, ...section.entries],
  []
);

// Initiate PDFDocument obejct
let mergedPdf = await PDFDocument.create();
const font = await mergedPdf.embedFont(StandardFonts.HelveticaBold);
const paginationFont = await mergedPdf.embedFont(
  StandardFonts.CourierBold
);

// First, put the cover page on
/*  const cover = await readFileAsync("./tmp/cover.pdf");
const coverPdf = await PDFDocument.load(cover);
const coverPage = await mergedPdf.copyPages(coverPdf, [0]);


coverPage[0].drawText(req.project.name, {
  x: 50,
  y: 75,
  font,
  size: 16,
  color: rgb(0.26, 0.25, 0.24),
});

// Put the name of the bundle
coverPage[0].drawText(req.project.bundles[req.bundleKey].name, {
  x: 50,
  y: 50,
  font,
  size: 12,
  color: rgb(0.41, 0.4, 0.37),
});

mergedPdf.addPage(coverPage[0]);*/

// Get dividers for future usage
//  const divider = await readFileAsync("./tmp/divider.pdf");

// Set page count at 1
let pageNumber = 1;

// Array of PDFDocument objects for all files to be merged
for (key in entries) {
  const entry = entries[key];
  if (!entry.uploaded) return false;
  const filename = `./tmp/${entry.uploaded.replace("/", "")}`;

  // Download file to the tmp folder
  await downloadFile(filename, entry.uploaded);

  // First, add the divider
  /*    const dividerPdf = await PDFDocument.load(divider);
  const dividerPage = await mergedPdf.copyPages(dividerPdf, [0]);
  const dividerNo = parseInt(parseInt(key) + 1).toString();
  dividerPage[0].drawText(dividerNo, {
    x: 560,
    y: 710,
    font,
    size: 14,
    color: rgb(0.22, 0.22, 0.22),
  });
  dividerPage[0].drawText(parseFilename(entry.name, key), {
    x: 50,
    y: 710,
    font,
    size: 14,
    maxWidth: 350,
    color: rgb(0.22, 0.22, 0.22),
  });
  mergedPdf.addPage(dividerPage[0]);*/

  // Now load into a PDFDocument object
  const file = await readFileAsync(filename);
  const entryPdf = await PDFDocument.load(file);

  const copiedPages = await mergedPdf.copyPages(
    entryPdf,
    entryPdf.getPageIndices()
  );

  // Now add each page into the merged PDF
  copiedPages.forEach((page) => {
    const width = page.getWidth();
    const bottomRight = width - 20;
    // Draw a background circle for hte number
    page.drawCircle({
      x: bottomRight,
      y: 20,
      size: 15,
      color: rgb(1, 1, 1),
    });
    // Put the number on the page
    page.drawText(pageNumber.toString(), {
      x:
        pageNumber < 10
          ? bottomRight - 3
          : pageNumber < 99
          ? bottomRight - 7
          : bottomRight - 11,
      y: 16,
      font: paginationFont,
      size: 12,
      color: rgb(0, 0, 0),
    });
    // Increment the page nunber
    pageNumber = pageNumber + 1;

    // And add the page in
    mergedPdf.addPage(page);
  });

  // Now delete them
  await unlinkAsync(filename);
}

// Finalise the pdf file
const finalPdf = await mergedPdf.save();

const uploadPdf = new Buffer.from(finalPdf);

// Now write to the file sync
//await writeFileAsync("./tmp/merged.pdf", finalPdf);

// Set filename for the bundle
const bundleFilenameLong = `${bundle._id.toString()}.pdf`;
const bundleFilenameShort = `${req.project.name} - ${bundle.name}`;

// Now upload to s3
const upload = await new Promise((resolve, reject) => {
  s3.putObject(
    {
      Bucket: keys.awsBucket,
      Key: bundleFilenameLong,
      Body: uploadPdf,
    },
    (err, url) => {
      if (err) reject(err);
      else resolve(url);
    }
  );
});
const url = await getFile(bundleFilenameLong, bundleFilenameShort);

return res.status(200).send(url);

[Hopding]

Hello @jackwshepherd!

I’m unable to view the document via the link you shared. Can you please upload the PDF directly to this thread as an attachment?

[jackwshepherd]

@Hopding - I just emailed this to you. Thank you :)

[Hopding]

I just published 1.6.1-rc1. It contains a fix for this issue (see #458). The problem seems to be that the PDF you are copying from contains a number that is too large to fit into a 64 bit integer. Presumably, this causes Acrobat to throw a fit and not render the document. Let me know how this version works for you!

You can install this new version with npm:

npm install pdf-lib@1.6.1-rc1

It's also available on unpkg:

• https://unpkg.com/pdf-lib@1.6.1-rc1/dist/pdf-lib.js
• https://unpkg.com/pdf-lib@1.6.1-rc1/dist/pdf-lib.min.js

As well as jsDelivr:

• https://cdn.jsdelivr.net/npm/pdf-lib@1.6.1-rc1/dist/pdf-lib.js
• https://cdn.jsdelivr.net/npm/pdf-lib@1.6.1-rc1/dist/pdf-lib.min.js

[Hopding]

@jackwshepherd Version 1.6.1 is now published. It introduces a capNumbers option that resolves this issue. You can use it like so:

const pdfDoc = await PDFDocument.load(bytes, { capNumbers: true });

The full release notes are available here.

You can install this new version with npm:

npm install pdf-lib@1.6.1

It's also available on unpkg:

• https://unpkg.com/pdf-lib@1.6.1/dist/pdf-lib.js
• https://unpkg.com/pdf-lib@1.6.1/dist/pdf-lib.min.js

As well as jsDelivr:

• https://cdn.jsdelivr.net/npm/pdf-lib@1.6.1/dist/pdf-lib.js
• https://cdn.jsdelivr.net/npm/pdf-lib@1.6.1/dist/pdf-lib.min.js

[jackwshepherd]

Thank you once again :)