Comparison of two pdf files

Question

I need to compare the contents of two almost similar files and highlight the dissimilar portions in the corresponding pdf file. Am using pdfbox. Please help me atleast with the logic.

Solution 1

If you prefer a tool with a GUI, you could try this one: diffpdf. It's by Mark Summerfield, and since it's written with Qt, it should be available (or should be buildable) on all platforms where Qt runs on.

Here's a screenshot:

Solution 2

You can do the same thing with a shell script on Linux. The script wraps 3 components:

1. ImageMagick's compare command
2. the pdftk utility
3. Ghostscript

It's rather easy to translate this into a .bat Batch file for DOS/Windows...

Here are the building blocks:

pdftk

Use this command to split multipage PDF files into multiple singlepage PDFs:

pdftk  first.pdf  burst  output  somewhere/firstpdf_page_%03d.pdf
pdftk  2nd.pdf    burst  output  somewhere/2ndpdf_page_%03d.pdf

compare

Use this command to create a "diff" PDF page for each of the pages:

compare \
       -verbose \
       -debug coder -log "%u %m:%l %e" \
        somewhere/firstpdf_page_001.pdf \
        somewhere/2ndpdf_page_001.pdf \
       -compose src \
        somewhereelse/diff_page_001.pdf

Note, that compare is part of ImageMagick. But for PDF processing it needs Ghostscript as a 'delegate', because it cannot do so natively itself.

Once more, pdftk

Now you can again concatenate your "diff" PDF pages with pdftk:

pdftk \
      somewhereelse/diff_page_*.pdf \
      cat \
      output somewhereelse/diff_allpages.pdf

Ghostscript

Ghostscript automatically inserts meta data (such as the current date+time) into its PDF output. Therefore this is not working well for MD5hash-based file comparisons.

If you want to automatically discover all cases which consist of purely white pages (that means: there are no visible differences in your input pages), you could also convert to a meta-data free bitmap format using the bmp256 output device. You can do that for the original PDFs (first.pdf and 2nd.pdf), or for the diff-PDF pages:

 gs \
   -o diff_page_001.bmp \
   -r72 \
   -g595x842 \
   -sDEVICE=bmp256 \
    diff_page_001.pdf

 md5sum diff_page_001.bmp

Just create an all-white BMP page with its MD5sum (for reference) like this:

 gs \
   -o reference-white-page.bmp \
   -r72 \
   -g595x842 \
   -sDEVICE=bmp256 \
   -c "showpage quit"

 md5sum reference-white-page.bmp

Solution 3

I had this very problem myself and the quickest way that I've found is to use PHP and its bindings for ImageMagick (Imagick).

<?php
$im1 = new \Imagick("file1.pdf");
$im2 = new \Imagick("file2.pdf");

$result = $im1->compareImages($im2, \Imagick::METRIC_MEANSQUAREERROR);

if($result[1] > 0.0){
    // Files are DIFFERENT
}
else{
    // Files are IDENTICAL
}

$im1->destroy();
$im2->destroy();

Of course, you need to install the ImageMagick bindings first:

sudo apt-get install php5-imagick # Ubuntu/Debian

Comments

[scc]

I got an error when tried to download that file "The transferred file contained a virus and was therefore blocked. URL: testautomationguru.com/download/304 Media Type: application/java-vm Virus Name: McAfeeGW: BehavesLike.Java.Suspicious.xm"

[Brecht Machiels]

Here's a script to visually diff two PDFs page-by-page using ImageMagick and Poppler tools (for speed): gist.github.com/brechtm/891de9f72516c1b2cbc1. It outputs one JPG for each page of the PDFs in a pdfdiff directory and additionally prints the numbers of the pages which differ between the two PDFs.

[Nachiappan R]

I try to run the jar file from above mentioned site,but I am getting error like "no main manifest attribute, in taguru-pdf-util.jar", could you please help me on this

[caw]

Is there any chance you could use this on the CLI, skip the GUI and redirect the output directly to a file?

[Kurt Pfeifle]

@caw: (1) Did you see my other answer? -- (2) AFAIK, newer versions of DiffPDF can redirect output to a CSV file. I don't know if this completely skips the GUI, though. -- (3) There is a "purely-CLI" version of DiffPDF available, called DiffPDFc, to be found at www.qtrac.eu -- however, it is for Windows only.

[caw]

I haven't, but tried ImageMagick, pdftk and Ghostscript before. Not in that combination, but separately. Since the results of diffpdf are so good, in fact excellent, I had hoped that all this functionality which is already there could just be used to redirect into a PDF on the CLI. What a pity! Thanks for the information on the other versions of that tool as well. Unfortunately, newer versions are not open-source anymore and Windows-only is not perfect, either.

[snapshot]

I need to install also ghostscript

[Luciano Fantuzzi]

This is the right solution.