Watermark Every Page

Read the page count from Recipe metadata, edit each one-based page, and add semi-transparent centered text.

var pdfDoc = new Recipe("input.pdf", "watermarked.pdf");

for (var page = 1; page <= pdfDoc.metadata.pages; page++) {
  pdfDoc
    .editPage(page)
    .text("WATERMARK", "center", "center", {
      bold: true,
      size: 60,
      color: "#0000FF",
      align: "center center",
      opacity: 0.3,
    })
    .endPage();
}

pdfDoc.endPDF();

Watermark In Place

Omit the output path and Recipe writes back over the source file, so no second document is produced:

var pdfDoc = new Recipe("input.pdf");

for (var page = 1; page <= pdfDoc.metadata.pages; page++) {
  pdfDoc
    .editPage(page)
    .text("WATERMARK", "center", "center", {
      bold: true,
      size: 60,
      color: "#0000FF",
      align: "center center",
      opacity: 0.3,
    })
    .endPage();
}

pdfDoc.endPDF();

The source is overwritten with no backup, and modification appends an incremental update rather than rewriting the file, so a repeatedly watermarked document keeps growing. Watermark once into a new file when you need to keep the original or the smallest possible output.

Watermark A Buffer

When the PDF is already in memory, pass the Buffer as the source and omit the output path. endPDF(callback) then hands the finished document back as a Buffer instead of writing a file:

var pdfDoc = new Recipe(inputBuffer);

for (var page = 1; page <= pdfDoc.metadata.pages; page++) {
  pdfDoc
    .editPage(page)
    .text("WATERMARK", "center", "center", {
      size: 60,
      align: "center center",
      opacity: 0.3,
    })
    .endPage();
}

pdfDoc.endPDF(function (outputBuffer) {
  // outputBuffer holds the watermarked PDF.
});

Passing an output path alongside a Buffer source writes that file and gives the path to the callback instead of the bytes. Page insertion and encryption are not available in buffer mode.