Page-by-page PDF diff tool

diffpdf.sh

#!/bin/bash

# usage: diffpdf.sh file_1.pdf file_2.pdf

# requirements:
# - ImageMagick
# - Poppler's pdftoppm and pdfinfo tools (works with 0.18.4 and 0.41.0,
#                                         fails with 0.42.0)
#   (could be replaced with Ghostscript if speed is
#    not important - see commented commands below)

DIFFDIR="pdfdiff"                        # directory to place diff images in
MAXPROCS=$(getconf _NPROCESSORS_ONLN)    # number of parallel processes

pdf_file1=$1
pdf_file2=$2

function diff_page {
    # based on http://stackoverflow.com/a/33673440/438249
    pdf_file1=$1
    pdf_file2=$2
    page_number=$3
    page_index=$(($page_number - 1))

    # 2+x faster
    (cat $pdf_file1 | pdftoppm -f $page_number -singlefile -gray - | convert - miff:- ; \
     cat $pdf_file2 | pdftoppm -f $page_number -singlefile -gray - | convert - miff:- ) | \
    convert - \( -clone 0-1 -compose darken -composite \) \
            -channel RGB -combine $DIFFDIR/$page_number.jpg

    # 2x faster (breaks when using TIFF format instead of JPEG, and PNG is slow)
#    (pdftocairo -f $page_number -singlefile -jpeg $pdf_file1 -gray - | convert - miff:- ; \
#     pdftocairo -f $page_number -singlefile -jpeg $pdf_file2 -gray - | convert - miff:- ) | \
#    convert - \( -clone 0-1 -compose darken -composite \) \
#            -channel RGB -combine $DIFFDIR/$page_number.jpg

    # 1x (using Ghostscript for PDF to bitmap conversion)
#    convert -respect-parenthesis \
#            \( $pdf_file1[$page_index] -flatten -colorspace gray \) \
#            \( $pdf_file2[$page_index] -flatten -colorspace gray \) \
#            \( -clone 0-1 -compose darken -composite \) \
#            -channel RGB -combine $DIFFDIR/$page_number.jpg

#    compare $pdf_file1[$page_index] $pdf_file2[$page_index] \
#            -highlight-color blue $DIFFDIR/$page_number.png

    if (($? > 0)); then
        echo "Problem running pdftoppm or convert!"
        exit 1
    fi
    grayscale=$(convert $DIFFDIR/$page_number.jpg -colorspace HSL -channel g -separate +channel -format "%[fx:mean]" info:)
    if [ "$grayscale" != "0" ]; then
        echo "page $page_number ($grayscale)"
        return 1
    fi
    return 0
}

function num_pages {
    pdf_file=$1

    pdfinfo $pdf_file | grep "Pages:" | awk '{print $2}'
}

function minimum {
    echo $(( $1 < $2 ? $1 : $2 ))
}

# guard against accidental deletion of files in the root directory
if [ -z "$DIFFDIR" ]; then
    echo "DIFFDIR needs to be set!"
    exit 1
fi

pdf1_num_pages=$(num_pages $pdf_file1)
pdf2_num_pages=$(num_pages $pdf_file2)

min_pages=$(minimum $pdf1_num_pages $pdf2_num_pages)

if [ "$pdf1_num_pages" -ne "$pdf2_num_pages" ]; then
    echo "PDF files have different lengths ($pdf1_num_pages and $pdf2_num_pages)"
    rc=1
fi

if [ -d "$DIFFDIR" ]; then
    rm -f $DIFFDIR/*
else
    mkdir $DIFFDIR
fi


# get exit status from subshells http://stackoverflow.com/a/29535256/438249

function wait_for_processes {
    local rc=0

    while (( "$#" )); do
        # wait returns the exit status for the process
        if ! wait "$1"; then
            rc=1
        fi
        shift
    done
    return $rc
}

function howmany() {
    echo $#
}

rc=0
pids=""
for page_number in `seq 1 $min_pages`;
do
    diff_page $pdf_file1 $pdf_file2 $page_number &
    pids+=" $!"
    if [ $(howmany $pids) -eq "$MAXPROCS" ]; then
        if ! wait_for_processes $pids; then
            rc=1
        fi
        pids=""
    fi
done

if ! wait_for_processes $pids; then
    rc=1
fi

exit $rc

One other idea would be to convert the jpgs to a pdf and remove the images and folder.
The following would create a PDF diff. The output filename could be specified on the command line after the two files to compare.
convert pdfdiff/*.jpg +compress output.pdf

@frederickjh You can find an improved Python version of this script here: diffpdf.py. The two changes you suggest aren't included, however. I'm using this script for regression testing, and creating a PDF requires significant extra time overhead, so that's why that is not included. I agree with your other suggestion, but I will probably not be implementing that any time soon, I'm afraid. Pull request are welcomed ;-)

Just an minor observation that in many industrial applications Red/Green Colour use was historically
RED = PROPOSED CHANGE / ADDITION / ADDED (AKA redlined)
GREEN = (TO)GO i.e.. REMOVAL/REMOVED
BLUE = (TOBE) MODIFIED / MOVED

@GitHubRulesOK Thanks for the input. I didn't know that! When I first wrote this script, I probably just reasoned: green good, red bad :-)

Thanks for this! Gist doesn't allow pull requests, but if I could submit one, I'd fix the two places where pdfdiff was hard-coded rather than the environment variable DIFFDIR (and correct the typos in comments, of "against" being misspelled "agains", and of "file" being misspelled on the usage line).

@subrook Thanks, I made those changes. Though I highly recommend you use diffpdf.py instead which has many improvements.

Thanks for the updates. I needed to customize it for my use case, and I already know bash and not python, so this version made more sense for me.

Thanks for the script. Here are some changes to create a diff pdf and optimize the temp files:

who uploads an image of their code.. thanks pal

@WillCohenInfotrack
Its a good example of the task save the source as a PDF save the above SVG suggestions as a PDF and run a PDF differ to see how it performs :-) 🤣