Class: HeadMusic::Notation::LilyPond::ParsePreflight

Inherits:
Object
  • Object
show all
Defined in:
lib/head_music/notation/lily_pond/parse_preflight.rb

Overview

Boundary checks the Parser runs before it reads a document, so a caller never receives a reference to a partially built flow. Each check raises on the first problem it finds and is otherwise a no-op.

Constant Summary collapse

OPENERS = 
{open_brace: :close_brace, open_parallel: :close_parallel, open_chord: :close_chord}.freeze
LEXEMES = 
{
  open_brace: "{", close_brace: "}",
  open_parallel: "<<", close_parallel: ">>",
  open_chord: "<", close_chord: ">"
}.freeze

Class Method Summary collapse

Class Method Details

.ensure_balanced_delimiters(tokens) ⇒ Object

Braces, << >>, and chord brackets must nest; an unmatched opener is reported at its own line, since that is where the author must look.



24
25
26
27
28
29
30
31
32
33
34
35
36
37
 
# File 'lib/head_music/notation/lily_pond/parse_preflight.rb', line 24

def self.ensure_balanced_delimiters(tokens)
  open = []
  tokens.each do |token|
    if OPENERS.key?(token.type)
      open << token
    elsif OPENERS.value?(token.type)
      opener = open.pop
      unless opener && OPENERS[opener.type] == token.type
        raise unbalanced_error(token, "Unexpected")
      end
    end
  end
  raise unbalanced_error(open.last, "Unclosed") if open.any?
end

.ensure_input_present(lily_pond_string) ⇒ Object

Checked on the raw bytes: String#strip raises on invalid UTF-8, and that case belongs to the Lexer's encoding guard.

Raises:



16
17
18
19
20
 
# File 'lib/head_music/notation/lily_pond/parse_preflight.rb', line 16

def self.ensure_input_present(lily_pond_string)
  return unless lily_pond_string.to_s.b.strip.empty?

  raise ParseError, "LilyPond input is blank"
end

.unbalanced_error(token, adjective) ⇒ Object (private) 



39
40
41
42
43
44
45
 
# File 'lib/head_music/notation/lily_pond/parse_preflight.rb', line 39

def self.unbalanced_error(token, adjective)
  lexeme = LEXEMES.fetch(token.type)
  ParseError.new(
    %(#{adjective} "#{lexeme}"),
    line_number: token.line, column: token.column, snippet: lexeme
  )
end