PDF embedded font sizes

pdf_font_sizes.sh

#!/usr/bin/env bash
#
# PDF Font Sizer - Analyze embedded font sizes in PDFs
#
# Assisted by DeepSeek & ChatGPT
#
# Requirements: pdffonts (poppler-utils) and mutool (mupdf-tools)
#
# Usage: ./pdf_font_sizes.sh document.pdf

set -euo pipefail

pdf=${1:?Usage: $0 PDF}

echo "Font sizes in: $pdf"
echo "=========================================="

pdffonts "$pdf" |
awk 'NR > 2 && NF >= 7 { print $1, $7 }' |
while read -r font obj; do
    # Remove the six-character PDF subset prefix
    name=${font#??????+}
    
    # Get the FontDescriptor object number
    desc=$(mutool show "$pdf" "$obj" 0 |
           grep -m1 'FontDescriptor' |
           grep -oE '[0-9]+ [0-9]+ R' |
           awk '{print $1}')
    
    [[ -n "$desc" ]] || continue
    
    # Get the embedded font stream object number
    data_obj=$(mutool show "$pdf" "$desc" 0 |
               grep -m1 -E 'FontFile[23]?' |
               grep -oE '[0-9]+ [0-9]+ R' |
               awk '{print $1}')
    
    [[ -n "$data_obj" ]] || continue
    
    size=$(mutool show "$pdf" "$data_obj" 0 | wc -c)
    size_kb=$(echo "scale=1; $size/1024" | bc)
    
    printf '%-30s %8d bytes (%6s KB)\n' "$name" "$size" "$size_kb"
done

echo "=========================================="