Class: Pubid::Ashrae::Parser

Inherits:

Parslet::Parser

• Object
• Parslet::Parser
• Pubid::Ashrae::Parser

Defined in:

lib/pubid/ashrae/parser.rb

Overview

Parser class for ASHRAE identifiers Single Responsibility: Parsing ASHRAE identifier syntax

Class Method Summary

• .parse(string) ⇒ Object

Class Method Details

.parse(string) ⇒ Object

# File 'lib/pubid/ashrae/parser.rb', line 652

def self.parse(string)
  # Strip leading/trailing whitespace
  cleaned = string.strip

  # Handle fixture format: !original!canonical
  # Extract just the original identifier for parsing
  if cleaned.start_with?("!")
    parts = cleaned.split("!")
    # Format: "" (before first !) + "original" + "canonical"
    cleaned = parts[1] if parts.size >= 3
  end

  # Normalize multiple spaces to single space
  cleaned = cleaned.gsub(/\s+/, " ")

  # Normalize reaffirmation patterns
  cleaned = cleaned.gsub(/\(RA\s+(\d{4})\)/, "(RA \\1)")
  cleaned = cleaned.gsub(/RA\s+(\d{4})/, "RA \\1")

  # Remove trailing periods and commas
  cleaned = cleaned.gsub(/[,.]$/, "")

  # Remove trailing double parentheses (typos in source data)
  cleaned = cleaned.gsub("))", ")")

  # Fix unclosed trailing parenthesis (data truncation issue)
  open_count = cleaned.count("(")
  close_count = cleaned.count(")")
  if open_count > close_count
    cleaned += ")" * (open_count - close_count)
  end

  # Normalize dash patterns (remove space around dash in year/code)
  cleaned = cleaned.gsub(/-\s+/, "-")
  cleaned = cleaned.gsub(/\s+/, " ") # Clean up any double spaces created

  # Preprocess ", and" and " and" patterns in addenda code lists
  # This fixes greedy matching where "and" gets consumed as a code
  # Only replace "and" in code lists (between Addendum/Addenda and to/for/(end)
  # NOT in descriptive text that follows
  if cleaned =~ /\bAddenda\s+/ || cleaned =~ /\bAddendum\s+/
    # Find the boundary where code list ends (to, for, or opening paren)
    boundary_pattern = /\s+(?:to|for|\()/
    boundary_match = cleaned.match(boundary_pattern)

    if boundary_match
      # Only replace "and" in the portion before the boundary
      boundary_pos = cleaned.index(boundary_match[0])
      before_boundary = cleaned[0...boundary_pos]
      after_boundary = cleaned[boundary_pos..]

      # Replace ", and" and " and" in the code list portion only
      before_boundary = before_boundary.gsub(/,\s+and\s+/i, ", ")
      before_boundary = before_boundary.gsub(/\s+and\s+/i, ", ")

      cleaned = before_boundary + after_boundary
    else
      # No boundary found, apply to entire string
      cleaned = cleaned.gsub(/,\s+and\s+/i, ", ")
      cleaned = cleaned.gsub(/\s+and\s+/i, ", ")
    end

    # Normalize "Addendum" to "Addenda" when followed by multiple codes (comma-separated)
    # Pattern: "Addendum a, b" -> "Addenda a, b"
    cleaned = cleaned.gsub(/\bAddendum\s+([a-z]+)\s*,\s*/i,
                           "Addenda \\1, ")
  end

  new.parse(cleaned)
end