fabulator-grammar 0.0.1 → 0.0.3

data/regex.racc

@@ -1,183 +0,0 @@
-class Fabulator::Grammar::Parser
-
-  start rules
-
-rule
-  rules: anchored_rule { result = Fabulator::Grammar::Expr::Rules.new; result.add_alternative(val[0]) }
-       | rules PIPE anchored_rule { result = val[0]; result.add_alternative(val[2]) }
-
-  anchored_rule: rule { result = val[0] }
-       | CARET rule { result = val[1]; result.anchor_start }
-       | rule DOLLAR { result = val[0]; result.anchor_end }
-       | CARET rule DOLLAR { result = val[1]; result.anchor_start; result.anchor_end }
-
-  rule: { result = Fabulator::Grammar::Expr::Rule.new; }
-      | rule sequence { result = val[0]; result.add_sequence(val[1]); }
-
-  sequence: sub_sequence sequence_qualifiers { result = Fabulator::Grammar::Expr::Sequence.new(val[0], val[1]) }
-          | sub_sequence { result = Fabulator::Grammar::Expr::Sequence.new(val[0]) }
-
-  sub_sequence: LT qname GT { result = Fabulator::Grammar::Expr::RuleRef.new(val[1]) }
-              | text { result = Fabulator::Grammar::Expr::Text.new(val[0]) }
-              | DOT { result = Fabulator::Grammar::Expr::Any.new }
-              | LP rules RP { result = val[1] }
-              | LB text RB { result = Fabulator::Grammar::Expr::CharSet.new(val[1]) }
-              | LB CARET text RB { result = Fabulator::Grammar::Expr::CharSet.new(val[2]); result.inverted }
-
-  text: qname { result = val[0] }
-      | TEXT { result = val[0] }
-      | INTEGER { result = val[0] }
-
-  qname: NCNAME { result = val[0] }
-       | NCNAME COLON NCNAME { result = val[0] + ':' + val[2] }
-
-  sequence_qualifiers: STAR { result = [ :zero_or_more ] }
-      | STAR QUESTION { result = [ :zero_or_more, :min ] }
-      | PLUS { result = [ :one_or_more ] }
-      | PLUS QUESTION { result = [ :one_or_more, :min ] }
-      | QUESTION { result = [ :zero_or_one ] }
-      | QUESTION QUESTION { result = [ :zero_or_one, :min ] }
-      | LC INTEGER RC { result = [ :exact, val[1].to_i ] }
-      | LC INTEGER COMMA INTEGER RC { result = [ :range, val[1].to_i, val[3].to_i ] }
-      | LC INTEGER COMMA RC { result = [ :range, val[1], '' ] }
-      | LC INTEGER COMMA RC QUESTION { result = [ :min, :range, val[1], '' ] }
-      | LC INTEGER COMMA INTEGER RC QUESTION { result = [ :min, :range, val[1].to_i, val[3].to_i ] }
-      
-
----- inner
-  require 'fabulator/grammar'
-
-  def parse(t, ctx)
-    @source = t
-    @curpos = 0 
-    @context = ctx
-    @line = 0   
-
-    @yydebug = true
-      
-    @last_token = nil
-  
-    do_parse
-  end
-    
-  def on_error(*args)
-    raise Fabulator::Grammar::ParserError.new("unable to parse '#{args[1]}' near line #{@line + 1}, column #{@col}")
-  end
-
-  @@regex = {
-    :ncname => %r{(?:[a-zA-Z_][-a-zA-Z0-9_.]*)}
-  }
-
-  @@regex[:qname] = %r{((?:#{@@regex[:ncname]}:)?#{@@regex[:ncname]})}
-
-  def next_token
-    @token = nil
-    white_space = 0
-    new_line = 0
-    @col = 0
-    while @curpos < @source.length && @source[@curpos..@curpos] =~ /\s/ do
-      if @source[@curpos..@curpos] =~ /\n/
-        new_line = new_line + 1
-        @line = @line + 1
-        @col = 0
-      else
-        @col = @col + 1
-      end
-      @curpos = @curpos + 1
-      white_space = white_space + 1   
-    end
-
-    # skip comments delimited by (:  :)
-    # comments can be nested
-    # these are XPath 2.0 comments
-    #
-    if @curpos < @source.length && @source[@curpos..@curpos+1] == '(:'
-      comment_depth = 1
-      @curpos = @curpos + 2
-      @col = @col + 2
-      while comment_depth > 0 && @curpos < @source.length
-        if @source[@curpos..@curpos+1] == '(:'
-          comment_depth = comment_depth + 1
-          @curpos = @curpos + 1
-          @col = @col + 1
-        end
-        if @source[@curpos..@curpos+1] == ':)'
-          comment_depth = comment_depth - 1
-          @curpos = @curpos + 1
-          @col = @col + 1
-        end
-        @curpos = @curpos + 1
-        @col = @col + 1
-      end
-      white_space = white_space + 1
-    end
-
-    while @curpos < @source.length && @source[@curpos..@curpos] =~ /\s/ do
-      if @source[@curpos..@curpos] =~ /\n/
-        new_line = new_line + 1
-        @line = @line + 1
-        @col = 0
-      else
-        @col = @col + 1
-      end
-      @curpos = @curpos + 1
-      white_space = white_space + 1
-    end
-
-    if @curpos >= @source.length
-      @last_token = nil
-      return [ false, false ]
-    end
-
-    case @source[@curpos..@curpos]
-      when '<': @token = [ :LT, '<' ]
-      when '>': @token = [ :GT, '>' ]
-      when '[': @token = [ :LB, '[' ]
-      when ']': @token = [ :RB, ']' ]
-      when '(': @token = [ :LP, '(' ]
-      when ')': @token = [ :RP, ')' ]
-      when '{': @token = [ :LC, '{' ]
-      when '}': @token = [ :RC, '}' ]
-      when ':': @token = [ :COLON, ':' ]
-      when ',': @token = [ :COMMA, ',' ]
-      when '|': @token = [ :PIPE, '|' ]
-      when '*': @token = [ :STAR, '*' ]
-      when '+': @token = [ :PLUS, '+' ]
-      when '.': @token = [ :DOT,  '.' ]
-      when '?': @token = [ :QUESTION, '?' ]
-      when '$': @token = [ :DOLLAR, '$' ]
-      when '^': @token = [ :CARET, '^' ]
-    end
-
-    if @token.nil?
-      # get longest sequence of non-special characters
-      # if it's all digits, report INTEGER
-      # if it's a qname, report QNAME
-      # otherwise, report TEXT
-      @source[@curpos..@source.length-1] =~ /^(((\\.)|[^ \$\^\[\]<>\{\}\(\):,|*+.?])+)*/
-      text = $1
-      bits = text.split(/\\/)
-      text = bits.join('')
-      @curpos += bits.size - 1
-      if text.length > 0
-        if @source[@curpos+text.length .. @curpos+text.length] =~ /[*?+\{]/
-          text = text[0..text.length-2]
-          @token = [ :TEXT, text ]
-        else
-          case text
-            when /^\d+$/: @token = [ :INTEGER, text ]
-            when /^#{@@regex[:ncname]}$/: @token = [ :NCNAME, text ]
-            else @token = [ :TEXT, text ]
-          end
-        end
-      end
-    end
-
-    if @token.nil?
-      puts "Uh oh... we don't know what to do: #{@source[@curpos .. @source.length-1]}"
-    else
-      @curpos += @token[1].length
-    end
-
-    return @token
-  end