Small bash file for transforming pdf files into markdown file leveraging on markitdown from microsoft and python's venv

transformPdfFileToMd.sh

#!/bin/bash

# Bash script responsible for transforming pdf files to md files for LLMs
VENV_PATH="/Users/salo/Documents/personal/repos/customMarkItDown/python_virtual_env"

# Extract filename with extension (e.g., "file.pdf")
stripFilenameFromInputFilePath() {
  local src="$1"
  echo "${src##*/}"
}

# Extract the directory path (e.g., "/path/to")
stripPathFromInputFilePath() {
  local src="$1"
  if [[ "$src" == */* ]]; then
    echo "${src%/*}"
  else
    echo "."
  fi
}

# Extract filename without extension (e.g., "file" from "file.pdf")
stripExtension() {
  local filename="$1"
  echo "${filename%.*}" 
}

activateVenv() {
  source "$VENV_PATH/bin/activate"
}

callMarkItDown() {
  local input_file="$1"
  local output_file="$2"
  markitdown "$input_file" > "$output_file"
}

deactivateVenv() {
  deactivate
}

transform() {
  local input="$1"
  local custom_output="$2"
  local output_target=""

  if [ -n "$custom_output" ]; then
    output_target="$custom_output"
  else
    local filepath=$(stripPathFromInputFilePath "$input")
    local filename_with_ext=$(stripFilenameFromInputFilePath "$input")
    local filename_no_ext=$(stripExtension "$filename_with_ext")
    output_target="$filepath/$filename_no_ext.md"
  fi

  echo "Converting: $input -> $output_target"

  # The actual work
  callMarkItDown "$input" "$output_target"
}

# --- Main Execution ---

if [ $# -eq 0 ]; then
  echo "Usage:"
  echo "  Single file with custom output: pdf2md <input.pdf> [output.md]"
  echo "  Batch processing:               pdf2md <file1.pdf> <file2.pdf> ..."
  echo "  Wildcard processing:            pdf2md *.pdf"
  exit 1
fi

activateVenv

# Check if there are exactly 2 arguments AND the second one is NOT a PDF.
# *.[pP][dD][fF] catches .pdf, .PDF, .Pdf, etc.
if [ $# -eq 2 ] && [[ "$2" != *.[pP][dD][fF] ]]; then
  
  transform "$1" "$2"

else
  
  # Loop through all arguments passed to the script
  for file in "$@"; do
    # Only process if the file actually exists and is a PDF
    if [ -f "$file" ] && [[ "$file" == *.[pP][dD][fF] ]]; then
      transform "$file" ""
    else
      echo "Skipping: $file (Not a valid PDF)"
    fi
  done

fi

deactivateVenv
echo "Done!"
exit 0