scripts/strippdf

#!/usr/bin/zsh

usage() {
  echo "Usage: $0 [-q quality] [--no-transitions] input.pdf"
  echo "  -q quality    Set output quality level (0: screen, 1: ebook (default), 2: prepress)"
  echo "  --no-transitions  Remove transitional pages (keep only the last frame before slide number changes)"
  exit 1
}

# Default values
quality=1
remove_transitions=0

# Parse arguments
while [[ $# -gt 0 ]]; do
  case "$1" in
    -q)
      shift
      if [[ "$1" =~ ^[0-2]$ ]]; then
        quality=$1
      else
        echo "Invalid quality level. Use 0 (screen), 1 (ebook), or 2 (prepress)."
        exit 1
      fi
      ;;
    --no-transitions)
      remove_transitions=1
      ;;
    -*)
      echo "Unknown option: $1"
      usage
      ;;
    *)
      input_file="$1"
      ;;
  esac
  shift
done

# Check if input file is provided
if [[ -z "$input_file" ]]; then
  echo "Error: No input file provided."
  usage
fi

# Set Ghostscript quality settings
case "$quality" in
  0) gs_quality="/screen" ;;
  1) gs_quality="/ebook" ;;
  2) gs_quality="/prepress" ;;
esac

# Create temporary file
temp_file=$(mktemp)

# Uncompress PDF and remove annotations
pdftk "$input_file" output - uncompress | sed '/^\/Annots/d' > "$temp_file.pdf"

if [[ $remove_transitions -eq 1 ]]; then
  # Run dump_data once and capture its output.
  dump=$(pdftk "$temp_file.pdf" dump_data)
  
  # Get the total number of pages.
  total_pages=$(echo "$dump" | awk '/^NumberOfPages/ { print $2 }')
  
  # Read all PageLabelNewIndex values into an array.
  new_indexes=($(echo "$dump" | awk '/^PageLabelNewIndex/ { print $2 }'))
  
  endings=()
  
  # Compute end pages for all groups except the first.
  # For each boundary between slides:
  #   If the difference is >1, the previous slide spans multiple pages:
  #      so use (current new index – 1) as the end page.
  #   If the difference is exactly 1, then the previous slide is a single page;
  #      so keep its PageLabelNewIndex.
  for ((i=1; i < ${#new_indexes[@]}; i++)); do
    prev=${new_indexes[i-1]}
    curr=${new_indexes[i]}
    if (( curr - prev == 1 )); then
      endings+=($prev)
    else
      endings+=($((curr - 1)))
    fi
  done
  
  # For the final slide group:
  last_index=${new_indexes[-1]}
  if (( total_pages - last_index == 0 )); then
    # Single–frame slide.
    endings+=($last_index)
  else
    # Multi–frame slide.
    endings+=($total_pages)
  fi
  
  # Remove duplicates (if any) and sort in ascending order.
  pages_to_keep=($(echo "${endings[@]}" | tr ' ' '\n' | sort -n | uniq))
  
  # (Optional: Uncomment to debug)
  # echo "Pages to keep: ${pages_to_keep[@]}"
  
  if [[ ${#pages_to_keep[@]} -gt 0 ]]; then
    pdftk "$temp_file.pdf" cat ${pages_to_keep[@]} output "$temp_file-clean.pdf"
    if [[ $? -ne 0 ]]; then
      echo "Error: Failed to remove transitional pages."
      rm "$temp_file.pdf"
      exit 1
    fi
    mv "$temp_file-clean.pdf" "$temp_file.pdf"
  else
    echo "Warning: No transitional pages identified; keeping full document."
  fi
fi

# Compress the PDF
gs -sDEVICE=pdfwrite -dCompatibilityLevel=1.4 -dPDFSETTINGS="$gs_quality" -dNOPAUSE -dQUIET -dBATCH -sOutputFile="${input_file%.pdf}-compressed.pdf" "$temp_file.pdf"

# Clean up
rm "$temp_file.pdf"