compress all pdf files recursively

Question

i want to compress all my pdf files in an given directory and its sub-directories using ghostscript.

I´m stuck using the find command within loops by filename including spaces.

Here some example code I´m aiming for:

pdffiles=$(find /path/to/directory -type f -name *.pdf)
for file in pdffiles; do
  gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS=/screen -dNOPAUSE -dBATCH -dQUIET -sOutputFile=new_$file $file; 
  rm $file;
  mv new_$file $file;
done;

Any idea how i could fix the problem with the spaces? Is there a better way to do this?

Answer 1

A single line script is also an option:

find -type f -name "*.pdf" -exec bash -c 'gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS=/screen -dNOPAUSE -dBATCH -dQUIET -sOutputFile="new.pdf" "{}"; rm "{}"; mv "new.pdf" "{}";' {} \;

Answer 2

I refactored my script based on your great response(s) and it's running quite well :)

Here is the refactored, improved code with logging, parameters and some more stuff.  I'm always open to improvements about my code.

#!/bin/bash
        
## Script to compress PDF Files using Ghostscript incl. subdirs
## Copyright (C) 2016 Maximilian Fries - All Rights Reserved
## Contact: [email protected]
## Last revised 2016-07-29

# Usage
# ./pdf-compress.sh [screen|ebook|prepress|default] [verbose]

# Variables and preparation
{
  count=0
  success=0
  successlog=./success.tmp
  gain=0
  gainlog=./gain.tmp
  pdfs=$(find ./ -type f -name "*.pdf")
  total=$(echo "$pdfs" | wc -l)
  log=./log
  verbose="-dQUIET"
  mode="prepress"
  echo "0" | tee $successlog $gainlog > /dev/null
}

# Are there any PDFs?
if [ "$total" -gt 0 ]; then

    #Parameter Handling & Logging
    {
        echo "-- Debugging for Log START --"
        echo "Number of Parameters: $#"
        echo "Parameters are: $*"
        echo "-- Debugging for Log END   --"
    } >> $log

    # Only compression-mode set
    if [ $# -eq 1 ]; then
        mode="$1"
    fi

    # Also Verbose Level Set
    if [ $# -eq 2 ]; then
        mode="$1"
        verbose=""
    fi

    echo "$pdfs" | while read -r file
    do
        ((count++))
        echo "Processing File #$count of $total Files" | tee -a $log
        echo "Current File: $file "| tee -a $log
        gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS="/$mode" -dNOPAUSE \
                -dBATCH $verbose -sOutputFile="$file-new" "$file" | tee -a $log
    
        sizeold=$(wc -c "$file"     | cut -d' ' -f1)
        sizenew=$(wc -c "$file-new" | cut -d' ' -f1)
        difference=$((sizenew-sizeold))

        # Check if new filesize is smaller
        if [ $difference -lt 0 ]
        then
            rm "$file"
            mv "$file-new" "$file"
            printf "Compression was successful. New File is %'.f Bytes smaller\n" \
                    $((-difference)) | tee -a $log
            ((success++)) 
            echo $success > $successlog
            ((gain-=difference))
            echo $gain > $gainlog
        else
            rm "$file-new"
            echo "Compression was not necessary" | tee -a $log
        fi

    done

    # Print Statistics
    printf "Successfully compressed %'.f of %'.f files\n" $(cat $successlog) $total | tee -a $log
    printf "Saved a total of %'.f Bytes\n" $(cat $gainlog) | tee -a $log

    rm $successlog $gainlog

else
    echo "No PDF File in Directory"
fi

Answer 3

Your loop is better written as

find ... | while read -r file

But then you need to make sure you quote the filename inside the loop. So we end up with

find /path/to/directory -type f -name *.pdf | while read -r file
do
  gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS=/screen -dNOPAUSE -dBATCH -dQUIET -sOutputFile="new_$file" "$file"
  rm "$file"
  mv "new_$file" "$file"
done

(Also note all those ; aren't needed).

Now this loop has potential file ownership/permission issues, but that's another question :-)