Fixes common markdown formatting issues to comply with GitHub Flavored Markdown standards

marked2-preprocessor

#!/usr/bin/env ruby
# encoding: UTF-8
# Markdown Formatter Preprocessor for Marked 2
# Fixes common markdown formatting issues to comply with GitHub Flavored Markdown standards
#
# This script reads markdown from STDIN and outputs formatted markdown to STDOUT.
# It ensures proper blank lines before lists, headings, code blocks, and tables.
#
# Usage in Marked 2:
#   Settings > Advanced > Custom Preprocessor
#   Path: /opt/homebrew/bin/marked2-preprocessor
#   Check "Use by default"

Encoding.default_external = Encoding::UTF_8
Encoding.default_internal = Encoding::UTF_8

def format_markdown(content)
  lines = content.split("\n")
  output = []
  in_code_block = false
  in_table = false
  prev_line_blank = true
  code_fence_marker = nil
  
  lines.each_with_index do |line, index|
    stripped = line.strip
    prev_line = index > 0 ? lines[index - 1].strip : ""
    next_line = index < lines.length - 1 ? lines[index + 1].strip : ""
    
    # Track code blocks
    if stripped.start_with?('```') || stripped.start_with?('~~~')
      if !in_code_block
        # Starting a code block
        in_code_block = true
        code_fence_marker = stripped[0..2]
        
        # Ensure blank line before code block (unless at start of file or after blank line)
        if !prev_line_blank && index > 0 && !prev_line.empty?
          output << ""
        end
        
        output << line
        prev_line_blank = false
        next
      elsif stripped.start_with?(code_fence_marker)
        # Ending a code block
        in_code_block = false
        code_fence_marker = nil
        output << line
        prev_line_blank = false
        next
      end
    end
    
    # If inside code block, don't modify
    if in_code_block
      output << line
      prev_line_blank = false
      next
    end
    
    # Detect if line is a table row (starts with | or is a separator line)
    is_table_row = stripped.start_with?('|') || (stripped.match?(/^\|?[-:\s|]+\|?$/) && stripped.include?('-'))
    
    # Track table state
    if is_table_row && !in_table
      # Starting a table - ensure blank line before (unless at start or after blank line)
      if !prev_line_blank && index > 0 && !prev_line.empty?
        output << ""
      end
      in_table = true
    elsif !is_table_row && in_table
      # Exiting table
      in_table = false
    end
    
    # Detect headings (lines starting with #)
    if stripped.match?(/^[#]{1,6}\s+/)
      # Ensure blank line before heading (unless at start of file or after blank line)
      if !prev_line_blank && index > 0 && !prev_line.empty?
        output << ""
      end
      output << line
      prev_line_blank = false
      next
    end
    
    # Detect list items (lines starting with -, *, +, or number followed by .)
    is_list_item = stripped.match?(/^[-*+]\s+/) || stripped.match?(/^\d+\.\s+/)
    
    if is_list_item
      # Check if previous line was also a list item or blank
      prev_is_list = prev_line.match?(/^[-*+]\s+/) || prev_line.match?(/^\d+\.\s+/)
      
      # If starting a new list (not continuing), ensure blank line before
      if !prev_is_list && !prev_line_blank && index > 0 && !prev_line.empty?
        output << ""
      end
      
      output << line
      prev_line_blank = false
      next
    end
    
    # Handle blank lines
    if stripped.empty?
      # Only add blank line if the previous line wasn't already blank
      # This prevents multiple consecutive blank lines
      if !prev_line_blank
        output << line
        prev_line_blank = true
      end
      next
    end
    
    # Regular content line
    output << line
    prev_line_blank = false
  end
  
  # Join lines and ensure file ends with newline
  result = output.join("\n")
  result += "\n" unless result.end_with?("\n")
  result
end

# Read from STDIN
input = STDIN.read

# Format the markdown
output = format_markdown(input)

# Write to STDOUT
STDOUT.write(output)