How to append PDF pages using PyPDF2

Question

Is anybody has experience merging two page of PDF file into one using python lib PyPDF2. When I try page1.mergePage(page2) it results with page2 overlayed page1. How to make it to add page2 to the bottom of the page1?

Solution 1

As I'm searching the web for python pdf merging solution, I noticed that there's a general misconception with merging versus appending.

Most people call the appending action a merge but it's not. What you're describing in your question is really the intended use of mergePage which should be called applyPageOnTopOfAnother but that's a little long. What you are (were) looking for is really appending two files/pages into a new file.

Appending PDF files

Using the PdfFileMerger class and its append method.

Identical to the merge() method, but assumes you want to concatenate all pages onto the end of the file instead of specifying a position.

Here's one way to do it taken from pypdf Merging multiple pdf files into one pdf:

from PyPDF2 import PdfFileMerger, PdfFileReader

# ...

merger = PdfFileMerger()

merger.append(PdfFileReader(file(filename1, 'rb')))
merger.append(PdfFileReader(file(filename2, 'rb')))

merger.write("document-output.pdf")

Appending specific PDF pages

And to append specific pages of different PDF files, use the PdfFileWriter class with the addPage method.

Adds a page to this PDF file. The page is usually acquired from a PdfFileReader instance.

file1 = PdfFileReader(file(filename1, "rb"))
file2 = PdfFileReader(file(filename2, "rb"))

output = PdfFileWriter()

output.addPage(file1.getPage(specificPageIndex))
output.addPage(file2.getPage(specificPageIndex))

outputStream = file("document-output.pdf", "wb")
output.write(outputStream)
outputStream.close()

Merging two pages into one page

Using mergePage

Merges the content streams of two pages into one. Resource references (i.e. fonts) are maintained from both pages. The mediabox/cropbox/etc of this page are not altered. The parameter page’s content stream will be added to the end of this page’s content stream, meaning that it will be drawn after, or “on top” of this page.

file1 = PdfFileReader(file(filename1, "rb"))
file2 = PdfFileReader(file(filename2, "rb"))

output = PdfFileWriter()

page = file1.getPage(specificPageIndex)
page.mergePage(file2.getPage(specificPageIndex))

output.addPage(page)

outputStream = file("document-output.pdf", "wb")
output.write(outputStream)
outputStream.close()

Solution 2

If the 2 PDFs do not exist on your local machine, and instead are normally accessed/download via a URL (i.e. http://foo/bar.pdf & http://bar/foo.pdf), we can fetch both PDFs from remote locations and merge them together in memory in one-fell-swoop.

This eliminates the assumed step of downloading the PDF to begin with, and allows us to generalize beyond the simple case of both PDFs existing on disk. Specifically, it generalizes the solution to any HTTP-accessible PDF.

The example:

    from PyPDF2 import PdfFileMerger, PdfFileReader

    pdf_content_1 = requests.get('http://foo/bar.pdf').content
    pdf_content_2 = requests.get('http://bar/foo.pdf').content

    # Write to in-memory file-like buffers        
    pdf_buffer_1 = StringIO.StringIO().write(pdf_content_1)
    pdf_buffer_2 = StringIO.StringIO().write(pdf_content_2)
    pdf_merged_buffer = StringIO.StringIO()

    merger = PdfFileMerger()
    merger.append(PdfFileReader(pdf_buffer_1))
    merger.append(PdfFileReader(pdf_buffer_2))
    merger.write(pdf_merged_buffer)

    # Option 1:
    # Return the content of the buffer in an HTTP response (Flask example below)
    response = make_response(pdf_merged_buffer.getvalue())
    # Set headers so web-browser knows to render results as PDF
    response.headers['Content-Type'] = 'application/pdf'
    response.headers['Content-Disposition'] = \ 
        'attachment; filename=%s.pdf' % 'Merged PDF'
    return response 
    # Option 2: Write to disk
    with open("merged_pdf.pdf", "w") as fp:
        fp.write(pdf_merged_buffer.getvalue())

Solution 3

Did it this way:

reader = PyPDF2.PdfFileReader(open("input.pdf",'rb'))

NUM_OF_PAGES = reader.getNumPages()

page0 = reader.getPage(0)
h = page0.mediaBox.getHeight()
w = page0.mediaBox.getWidth()

newpdf_page = PyPDF2.pdf.PageObject.createBlankPage(None, w, h*NUM_OF_PAGES)
for i in range(NUM_OF_PAGES):
    next_page = reader.getPage(i)
    newpdf_page.mergeScaledTranslatedPage(next_page, 1, 0, h*(NUM_OF_PAGES-i-1))

writer = PdfFileWriter()
writer.addPage(newpdf_page)

with open('output.pdf', 'wb') as f:
    writer.write(f)

It works when every page has the same height and width. Otherwise, it needs some modifications.

Maybe Emile Bergeron solution is better. Didn't try it.

Solution 4

The pdfrw library can do this. There is a 4up example in the examples directory that places 4 input pages on every output page, and a booklet example that takes 8.5x11 input and creates 11x17 output. Disclaimer -- I am the pdfrw author.

View more solutions