treetop 1.4.5 → 1.4.7

data/lib/treetop/compiler/node_classes/repetition.rb

@@ -16,28 +16,37 @@ module Treetop
           builder.else_ do
             builder.break
           end
+          if max && !max.empty?
+            builder.if_ "#{accumulator_var}.size == #{max.text_value}" do
+              builder.break
+            end
+          end
         end
       end
-      
+
       def inline_module_name
         parent_expression.inline_module_name
       end
-      
+
       def assign_and_extend_result
         assign_result "instantiate_node(#{node_class_name},input, #{start_index_var}...index, #{accumulator_var})"
         extend_result_with_inline_module
       end
     end
 
-  
+
     class ZeroOrMore < Repetition
       def compile(address, builder, parent_expression)
         super
         assign_and_extend_result
         end_comment(parent_expression)
       end
+
+      def max
+        nil
+      end
     end
-  
+
     class OneOrMore < Repetition
       def compile(address, builder, parent_expression)
         super
@@ -50,6 +59,31 @@ module Treetop
         end
         end_comment(parent_expression)
       end
+
+      def max
+        nil
+      end
+    end
+
+    class OccurrenceRange < Repetition
+      def compile(address, builder, parent_expression)
+        super
+
+        if min.empty? || min.text_value.to_i == 0
+          assign_and_extend_result
+        else
+          # We got some, but fewer than we wanted. There'll be a failure reported already
+          builder.if__ "#{accumulator_var}.size < #{min.text_value}" do
+            reset_index
+            assign_failure start_index_var
+          end
+          builder.else_ do
+            assign_and_extend_result
+          end
+        end
+        end_comment(parent_expression)
+      end
     end
+
   end
-end
+end

data/lib/treetop/version.rb

@@ -2,7 +2,7 @@ module Treetop #:nodoc:
   module VERSION #:nodoc:
     MAJOR = 1
     MINOR = 4
-    TINY  = 5
+    TINY  = 7
 
     STRING = [MAJOR, MINOR, TINY].join('.')
   end

data/spec/compiler/and_predicate_spec.rb

@@ -0,0 +1,36 @@
+require 'spec_helper'
+
+module AndPredicateSpec
+  describe "An &-predicated terminal symbol" do
+    testing_expression '&"foo"'
+
+    it "successfully parses input matching the terminal symbol, returning an epsilon syntax node" do
+      parse('foo', :consume_all_input => false) do |result|
+        result.should_not be_nil
+        result.interval.should == (0...0)
+      end
+    end
+  end
+
+  describe "A sequence of a terminal and an and another &-predicated terminal" do
+    testing_expression '"foo" &"bar"'
+
+    it "matches input matching both terminals, but only consumes the first" do
+      parse('foobar', :consume_all_input => false) do |result|
+        result.should_not be_nil
+        result.text_value.should == 'foo'
+      end
+    end
+  
+    it "fails to parse input matching only the first terminal, with a terminal failure recorded at index 3" do
+      parse('foo') do |result|
+        result.should be_nil
+        terminal_failures = parser.terminal_failures
+        terminal_failures.size.should == 1
+        failure = terminal_failures[0]
+        failure.index.should == 3
+        failure.expected_string.should == 'bar'
+      end
+    end
+  end
+end

data/spec/compiler/anything_symbol_spec.rb

@@ -0,0 +1,44 @@
+require 'spec_helper'
+
+module AnythingSymbolSpec
+  class Foo < Treetop::Runtime::SyntaxNode
+  end
+
+  describe "an anything symbol followed by a node class declaration and a block" do
+    testing_expression '. <AnythingSymbolSpec::Foo> { def a_method; end }'
+  
+    it "matches any single character in a big range, returning an instance of the declared node class that responds to methods defined in the inline module" do
+      (33..127).each do |digit|
+        parse(digit.chr) do |result|
+          result.should_not be_nil
+          result.should be_an_instance_of(Foo)
+          result.should respond_to(:a_method)
+          result.interval.should == (0...1)
+        end
+      end
+    end
+  
+    it "fails to parse epsilon" do
+      parse('').should be_nil
+    end
+  end
+    
+  module ModFoo
+  end
+
+  describe "an anything symbol followed by a module declaration and a block" do
+    testing_expression '. <AnythingSymbolSpec::ModFoo> { def a_method; end }'
+  
+    it "matches any single character in a big range, returning an instance of SyntaxNode extended by the declared module that responds to methods defined in the inline module" do
+      (33..127).each do |digit|
+        parse(digit.chr) do |result|
+          result.should_not be_nil
+          result.should be_an_instance_of(Treetop::Runtime::SyntaxNode)
+          result.should be_a_kind_of(ModFoo)
+          result.should respond_to(:a_method)
+          result.interval.should == (0...1)
+        end
+      end
+    end
+  end
+end

data/spec/compiler/character_class_spec.rb

@@ -0,0 +1,276 @@
+require 'spec_helper'
+
+module CharacterClassSpec
+  class Foo < Treetop::Runtime::SyntaxNode
+  end
+
+  describe "a character class followed by a node class declaration and a block" do
+
+    testing_expression "[A-Z] <CharacterClassSpec::Foo> { def a_method; end }"
+
+    it "matches single characters within that range, returning instances of the declared node class that respond to the method defined in the inline module" do
+      result = parse('A')
+      result.should be_an_instance_of(Foo)
+      result.should respond_to(:a_method)
+      result = parse('N')
+      result.should be_an_instance_of(Foo)
+      result.should respond_to(:a_method)
+      result = parse('Z')
+      result.should be_an_instance_of(Foo)
+      result.should respond_to(:a_method)
+    end
+
+    it "does not match single characters outside of that range" do
+      parse('8').should be_nil
+      parse('a').should be_nil
+    end
+
+    it "matches a single character within that range at index 1" do
+      parse(' A', :index => 1).should_not be_nil
+    end
+
+    it "fails to match a single character out of that range at index 1" do
+      parse(' 1', :index => 1).should be_nil
+    end
+  end
+
+  module ModFoo
+  end
+
+  describe "a character class followed by a node module declaration and a block" do
+
+    testing_expression "[A-Z] <CharacterClassSpec::ModFoo> { def a_method; end }"
+
+    it "matches single characters within that range, returning instances of SyntaxNode extended by the specified module" do
+      result = parse('A')
+      result.should be_an_instance_of(Treetop::Runtime::SyntaxNode)
+      result.should be_a_kind_of(ModFoo)
+      result.should respond_to(:a_method)
+      result = parse('N')
+      result.should be_an_instance_of(Treetop::Runtime::SyntaxNode)
+      result.should be_a_kind_of(ModFoo)
+      result.should respond_to(:a_method)
+      result = parse('Z')
+      result.should be_an_instance_of(Treetop::Runtime::SyntaxNode)
+      result.should be_a_kind_of(ModFoo)
+      result.should respond_to(:a_method)
+    end
+
+    it "does not match single characters outside of that range" do
+      parse('8').should be_nil
+      parse('a').should be_nil
+    end
+
+    it "matches a single character within that range at index 1" do
+      parse(' A', :index => 1).should_not be_nil
+    end
+
+    it "fails to match a single character out of that range at index 1" do
+      parse(' 1', :index => 1).should be_nil
+    end
+  end
+  
+  describe "a character class followed by a node class declaration and a block" do
+
+    testing_expression "[A-Z] <CharacterClassSpec::Foo>"
+    
+    it "actively generates nodes for the character when it is the primary node" do
+      result = parse('A')
+      result.should be_a(Treetop::Runtime::SyntaxNode)
+      result.elements.should be_nil
+    end
+    
+  end
+
+  describe "A character class containing quotes" do
+    testing_expression "[\"']"
+
+    it "matches a quote" do
+      parse("'").should_not be_nil
+    end
+
+    it "matches a double-quote" do
+      parse('"').should_not be_nil
+    end
+  end
+
+  describe "A character class containing a special character" do
+    testing_expression "[\t]"
+    it "matches that character only" do
+      parse("\t").should_not be_nil
+      parse('t').should be_nil
+    end
+  end
+
+  describe "A character class containing an escaped backslash" do
+    slash = "\\"  # Make it explicit that there are *two* backslashes here
+    testing_expression "[#{slash}#{slash}]"
+    it "matches a backslash only" do
+      parse("\\").should_not be_nil
+      parse('t').should be_nil
+    end
+  end
+
+  describe "A character class containing a hex escape" do
+    slash = "\\"
+    testing_expression "[#{slash}x41]"
+    it "matches that character only" do
+      parse('A').should_not be_nil
+      parse('\\').should be_nil
+      parse('x').should be_nil
+      parse('4').should be_nil
+      parse('1').should be_nil
+    end
+  end
+
+  describe "A character class containing an octal escape" do
+    slash = "\\"
+    testing_expression "[#{slash}101]"
+    it "matches that character only" do
+      parse('A').should_not be_nil
+      parse('\\').should be_nil
+      parse('1').should be_nil
+      parse('0').should be_nil
+    end
+  end
+
+  describe "A character class containing a \\c control-char escape" do
+    slash = "\\"
+    testing_expression "[#{slash}cC]"
+    it "matches that character only" do
+      parse("\003").should_not be_nil
+      parse('\\').should be_nil
+      parse('c').should be_nil
+      parse('C').should be_nil
+    end
+  end
+
+  describe "A character class containing a \\C- control-char escape" do
+    slash = "\\"
+    testing_expression "[#{slash}C-C]"
+    it "matches that character only" do
+      parse("\003").should_not be_nil
+      parse('\\').should be_nil
+      parse('C').should be_nil
+      parse('-').should be_nil
+    end
+  end
+
+  if RUBY_VERSION =~ /\A1\.8\./
+    describe "A character class containing a \\M- meta-char escape" do
+      slash = "\\"
+      testing_expression "[#{slash}M- ]"
+      it "matches that character only" do
+        parse("\240").should_not be_nil
+        parse('\\').should be_nil
+        parse('M').should be_nil
+        parse('-').should be_nil
+        parse(' ').should be_nil
+      end
+    end
+  end
+
+  describe "A character class containing an escaped non-special character" do
+    slash = "\\"
+    testing_expression "[#{slash}y]"
+    it "matches that character only" do
+      parse("y").should_not be_nil
+      parse('\\').should be_nil
+    end
+  end
+
+  describe "A character class containing an \#{...} insertion" do
+    testing_expression "[\#{raise 'error'}]"
+    it "doesn't evaluate the insertion" do
+      x = true
+      lambda{
+	x = parse("y")
+      }.should_not raise_error
+      x.should be_nil
+      parse('#').should_not be_nil
+      parse("'").should_not be_nil
+      parse("0").should be_nil
+    end
+  end
+  
+  describe "a character class" do
+    testing_expression "[A-Z]"
+    it "actively generates a node for the character because it is the primary node" do
+      result = parse('A')
+      result.should be_a(Treetop::Runtime::SyntaxNode)
+      result.elements.should be_nil
+    end
+  end
+  
+  describe "a character class mixed with other expressions" do
+    testing_expression '[A-Z] "a"'
+    it "lazily instantiates a node for the character" do
+      result = parse('Aa')
+      result.instance_variable_get("@elements").should include(true)
+      result.elements.should_not include(true)
+      result.elements.size.should == 2
+    end
+  end
+  
+  describe "a character class with a node class declaration mixed with other expressions" do
+    testing_expression '([A-Z] <CharacterClassSpec::Foo>) "a"'
+    it "actively generates a node for the character because it has a node class declared" do
+      result = parse('Aa')
+      result.instance_variable_get("@elements").should_not include(true)
+      result.elements.should_not include(true)
+      result.elements.size.should == 2
+    end
+  end
+  
+  describe "a character class with a node module declaration mixed with other expressions" do
+    testing_expression '([A-Z] <CharacterClassSpec::ModFoo>) "a"'
+    it "actively generates a node for the character because it has a node module declared" do
+      result = parse('Aa')
+      result.instance_variable_get("@elements").should_not include(true)
+      result.elements.should_not include(true)
+      result.elements.size.should == 2
+    end
+  end
+  
+  describe "a character class with an inline block mixed with other expressions" do
+    testing_expression '([A-Z] { def a_method; end }) "a"'
+    it "actively generates a node for the character because it has an inline block" do
+      result = parse('Aa')
+      result.instance_variable_get("@elements").should_not include(true)
+      result.elements.should_not include(true)
+      result.elements.size.should == 2
+    end
+  end
+  
+  describe "a character class with a label mixed with other expressions" do
+    testing_expression 'upper:([A-Z]) "b"'
+    it "returns the correct element for the labeled expression" do
+      result = parse('Ab')
+      result.upper.text_value.should == "A"
+      result.elements.size.should == 2
+    end
+  end
+  
+  describe "a character class repetition mixed with other expressions" do
+    testing_expression '[A-Z]+ "a"'
+    it "lazily instantiates a node for the character" do
+      result = parse('ABCa')
+      result.elements[0].instance_variable_get("@elements").should include(true)
+      result.elements[0].elements.should_not include(true)
+      result.elements[0].elements.size.should == 3
+      result.elements.size.should == 2
+      result.elements.inspect.should == %Q{[SyntaxNode offset=0, "ABC":\n  SyntaxNode offset=0, "A"\n  SyntaxNode offset=1, "B"\n  SyntaxNode offset=2, "C", SyntaxNode offset=3, "a"]}
+    end
+  end
+  
+  describe "a character class that gets cached because of a choice" do
+    testing_expression "[A-Z] 'a' / [A-Z]"
+    
+    it "generates a node for the lazily-instantiated character when it is the primary node" do
+      result = parse('A')
+      result.should be_a(Treetop::Runtime::SyntaxNode)
+      result.elements.should be_nil
+    end
+  end
+
+end

data/spec/compiler/choice_spec.rb

@@ -0,0 +1,80 @@
+require 'spec_helper'
+
+module ChoiceSpec
+  describe "A choice between terminal symbols" do
+    testing_expression '"foo" { def foo_method; end } / "bar" { def bar_method; end } / "baz" { def baz_method; end }'
+
+    it "successfully parses input matching any of the alternatives, returning a node that responds to methods defined in its respective inline module" do
+      result = parse('foo')
+      result.should_not be_nil
+      result.should respond_to(:foo_method)
+    
+      result = parse('bar')
+      result.should_not be_nil
+      result.should respond_to(:bar_method)
+    
+      result = parse('baz')
+      result.should_not be_nil
+      result.should respond_to(:baz_method)
+    end
+  
+    it "upon parsing a string matching the second alternative, records the failure of the first terminal" do
+      result = parse('bar')
+      terminal_failures = parser.terminal_failures
+      terminal_failures.size.should == 1
+      failure = terminal_failures[0]
+      failure.expected_string.should == 'foo'
+      failure.index.should == 0
+    end
+  
+    it "upon parsing a string matching the third alternative, records the failure of the first two terminals" do
+      result = parse('baz')
+      
+      terminal_failures = parser.terminal_failures
+      
+      terminal_failures.size.should == 2
+
+      failure_1 = terminal_failures[0]
+      failure_1.expected_string == 'foo'
+      failure_1.index.should == 0
+    
+      failure_2 = terminal_failures[1]
+      failure_2.expected_string == 'bar'
+      failure_2.index.should == 0
+    end
+  end
+
+  describe "A choice between sequences" do
+    testing_expression "'foo' 'bar' 'baz'\n/\n'bing' 'bang' 'boom'"
+
+    it "successfully parses input matching any of the alternatives" do
+      parse('foobarbaz').should_not be_nil
+      parse('bingbangboom').should_not be_nil
+    end
+  end
+
+  describe "A choice between terminals followed by a block" do  
+    testing_expression "('a'/ 'b' / 'c') { def a_method; end }"
+
+    it "extends a match of any of its subexpressions with a module created from the block" do
+      ['a', 'b', 'c'].each do |letter|
+        parse(letter).should respond_to(:a_method)
+      end
+    end
+  end
+
+  module TestModule
+    def a_method
+    end
+  end
+
+  describe "a choice followed by a declared module" do  
+    testing_expression "('a'/ 'b' / 'c') <ChoiceSpec::TestModule>"
+
+    it "extends a match of any of its subexpressions with a module created from the block" do
+      ['a', 'b', 'c'].each do |letter|
+        parse(letter).should respond_to(:a_method)
+      end
+    end
+  end
+end