Remove multiple text watermarks from a PDF file. Requires xxd and qpdf to work correctly.

remove-pdf-watermark.sh

#!/bin/bash

# Remove multiple text watermarks from a PDF file.  Requires xxd and qpdf to work correctly.
#
# Usage:
#
#   remove-pdf-watermark.sh "Your Input File.pdf" "Your Output File.pdf" [WATERMARK1] [WATERMARK2] [WATERMARK3] [...]
#
# For Example:
#
#   remove-pdf-watermark.sh "Your Input File.pdf" "Your Output File.pdf" "Watermark 1" "Watermark 2"
#

# This is a more general (lesser dependencies, more functionality, but slower) version of
# https://gist.github.com/elfsternberg/a96883018d783cbbad7b454ecd0a7ffe


INPUT_FILENAME=$1
OUTPUT_FILENAME=$2

echo "Processing: $INPUT_FILENAME"

UNCOMPRESSED=`mktemp -t 'uncompressed'`
UNCOMPRESSED_HEX=`mktemp -t 'uncompressed-hex'`
UNMARKED_PRE=`mktemp -t 'unmarked-pre'`

qpdf --stream-data=uncompress --decode-level=all "$INPUT_FILENAME" $UNCOMPRESSED
echo " - Decompressing to: $UNCOMPRESSED"
xxd -ps -c 0 $UNCOMPRESSED > $UNCOMPRESSED_HEX
echo " - Hex dumping to: $UNCOMPRESSED_HEX"

rm $UNCOMPRESSED

for ARG in "$@"; do
    COUNT=$((COUNT+1))
    if [[ $COUNT -gt 2 ]]
    then
        WATERMARK=$ARG
        echo " - Processing Watermark: \"$WATERMARK\""
        WATERMARKLEN=${#WATERMARK}
        WATERMARK_HEX=$(echo -n $WATERMARK | xxd -p -c 0)
        BLANKS=$(printf %${WATERMARKLEN}s)
        BLANKS_HEX=$(echo -n "$BLANKS" | xxd -p -c 0)

        NUM_OCCURENCES_=$(grep -o $WATERMARK_HEX $UNCOMPRESSED_HEX | wc -l)
        NUM_OCCURENCES=$((0+$NUM_OCCURENCES_))
        echo "    * Number of occurences: $NUM_OCCURENCES"

        if [[ $NUM_OCCURENCES -gt 0 ]]
        then
            echo "    * Replacing with $WATERMARKLEN blanks..."
            sed -i -e "s/$WATERMARK_HEX/$BLANKS_HEX/g" $UNCOMPRESSED_HEX
            echo "    * Replacement done."
            NUM_OCCURENCES_=$(grep -o $WATERMARK_HEX $UNCOMPRESSED_HEX | wc -l)
            NUM_OCCURENCES=$((0+$NUM_OCCURENCES_))
            echo "    * Final number of occurences: $NUM_OCCURENCES"
        fi
    fi
done

xxd -r -p -c 0 $UNCOMPRESSED_HEX $UNMARKED_PRE

rm $UNCOMPRESSED_HEX

echo " - Reverting from hex dump to: $UNMARKED_PRE"

qpdf --stream-data=compress $UNMARKED_PRE "$OUTPUT_FILENAME"

rm $UNMARKED_PRE

echo " - Output written to: $OUTPUT_FILENAME"
echo
echo


# NO WARRANTY
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
# MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
# NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
# LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
# OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
# WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

It will be interesting to see whether/how qpdf json v2 might help you with this kind of task once qpdf 11 is out. qpdf json v2 is in main now, though the interface isn't frozen until I release. If you want to take a look, you can check out current main. Documentation is at https://qpdf.readthedocs.io/en/latest/json.html

I think that sounds really really cool. Having structured data to work with allows so much more. Also, thanks for qpdf. It has saved me so much time.

Here's one json example which works for some pdfs

qpdf --json --json-stream-data=file in.pdf out.json
for i in `grep -Hi confidential out.json-* | cut -d':' -f1`; do
  cp /dev/null $i;
done
qpdf --json-input out.json out.pdf

@dhewg I think this is the first example I've seen of user-contributed qpdf json. :-)

And such a small yet powerful one! ;) Some watermarks are just too much in-your-face...
Thanks for the json format, this makes it easy to manipulate with standard tools!

Fails with:

mktemp: too few X's in template ‘uncompressed’
mktemp: too few X's in template ‘uncompressed-hex’
mktemp: too few X's in template ‘unmarked-pre’

qpdf: an output file name is required; use - for standard output

For help:
  qpdf --help=usage       usage information
  qpdf --help=topic       help on a topic
  qpdf --help=--option    help on an option
  qpdf --help             general help and a topic list

 - Decompressing to: 
./pdf_remove_watermark.sh: line 29: $UNCOMPRESSED_HEX: ambiguous redirect
 - Hex dumping to: 
rm: missing operand
Try 'rm --help' for more information.

Running on:

qpdf v11.9.1-1
coreutils v9.4-3.1

Fails with:

mktemp: too few X's in template ‘uncompressed’
mktemp: too few X's in template ‘uncompressed-hex’
mktemp: too few X's in template ‘unmarked-pre’

qpdf: an output file name is required; use - for standard output

For help:
  qpdf --help=usage       usage information
  qpdf --help=topic       help on a topic
  qpdf --help=--option    help on an option
  qpdf --help             general help and a topic list

 - Decompressing to: 
./pdf_remove_watermark.sh: line 29: $UNCOMPRESSED_HEX: ambiguous redirect
 - Hex dumping to: 
rm: missing operand
Try 'rm --help' for more information.

Running on:

qpdf v11.9.1-1
coreutils v9.4-3.1

Me too. I changed the mktemp's to this:

UNCOMPRESSED=`mktemp 'uncompressed-XXXXXXXXXX.pdf'`
UNCOMPRESSED_HEX=`mktemp 'uncompressed-hex-XXXXXXXXXX.pdf'`
UNMARKED_PRE=`mktemp 'unmarked-pre-XXXXXXXXXX.pdf'`