skeem 0.0.4 → 0.0.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: b820c8eb3683e1da04e9c28e41a4de2517894573
-  data.tar.gz: 31dfc0fa971d27e165a73653abb824fdc39d00b6
+  metadata.gz: e25bd8499926a963c98886effec8d51eb78bb441
+  data.tar.gz: fd5e2e65ceffbe1ce202ab1d620645d74a2e03f7
 SHA512:
-  metadata.gz: 333f74805437ad0199755a3bdb6bb8ffe8197c5d988979cf0bf8ec95dfd751fcdc423dcaa38c26adb892b982f62505a8bcb25ce62a0dade1c026ab05de0b3a5d
-  data.tar.gz: cd5852bda6bb381855a2ef9a64e3d6875da172ede1c6f520d5b97c697e5a484096cc93cd9874a95311b866ff4b248841ed6d1fd0d802e30dc0e3be60e38d5d08
+  metadata.gz: 181ac16e0465292fc39a1ac157e313348afcb36d83243922b7539969a3e506dbbc0e2f97d78cf41f7bcb4a728dd43d1f78477ec8ca3a468841298f61ca51c2ca
+  data.tar.gz: 8c6a36fd38ebcc92a4f208bba5be6ac9f11423a1cfb405970f40bbf141f4c70dc1fa74279365d9d6277ef68b15bf6ff40bd2903eab6c865a246b25ab1e121dbc

data/CHANGELOG.md

@@ -1,3 +1,10 @@
+## [0.0.5] - 2018-08-30
+Parser now generates correct parse trees for expressions consisting of a single literal.
+
+### Fixed
+- Class`Tokenizer` The regexp for real numbers was too restricitive: it didn't recognize real numbers without fractional part.
+
+
 ## [0.0.4] - 2018-08-29
 ### Added
 - File `s_expr_nodes.rb` with initial implementation of `SExprTerminalNode` classes.

data/lib/skeem/grammar.rb

@@ -24,7 +24,9 @@ module Skeem
     rule 'program' => 'cmd_or_def_plus'
     rule 'cmd_or_def_plus' => 'cmd_or_def_plus cmd_or_def'
     rule 'cmd_or_def_plus' => 'cmd_or_def'
+    rule 'cmd_or_def' => 'command'
     rule 'cmd_or_def' => 'definition'
+    rule 'command' => 'expression'
     rule 'definition' => 'LPAREN DEFINE IDENTIFIER expression RPAREN'
     rule 'expression' =>  'IDENTIFIER'
     rule 'expression' =>  'literal'

data/lib/skeem/parser.rb

@@ -1,5 +1,6 @@
 require_relative 'tokenizer'
 require_relative 'grammar'
+require_relative 's_expr_builder'
 
 module Skeem
   class Parser
@@ -7,7 +8,10 @@ module Skeem
     
     def initialize()
       # Create a Rley facade object
-      @engine = Rley::Engine.new { |cfg| cfg.diagnose = true }
+      @engine = Rley::Engine.new do |cfg| 
+        cfg.diagnose = true
+        cfg.repr_builder = SExprBuilder
+      end
 
       # Step 1. Load Skeem grammar
       @engine.use_grammar(Skeem::Grammar)
@@ -30,7 +34,7 @@ module Skeem
         raise StandardError, line1 + line2
       end
 
-      return result
+      return engine.to_ptree(result)
     end
   end # class
 end # module

data/lib/skeem/{node_builder.rb → s_expr_builder.rb}

@@ -1,7 +1,7 @@
 require 'stringio'
-require_relative 'regex_repr'
+require_relative 's_expr_nodes'
 
-module SrlRuby
+module Skeem
   # The purpose of a ASTBuilder is to build piece by piece an AST
   # (Abstract Syntax Tree) from a sequence of input tokens and
   # visit events produced by walking over a GFGParsing object.
@@ -9,8 +9,14 @@ module SrlRuby
   # The Builder pattern creates a complex object
   # (say, a parse tree) from simpler objects (terminal and non-terminal
   # nodes) and using a step by step approach.
-  class ASTBuilder < Rley::ParseRep::ASTBaseBuilder
-    Terminal2NodeClass = {}.freeze
+  class SExprBuilder < Rley::ParseRep::ASTBaseBuilder
+    Terminal2NodeClass = {
+      'BOOLEAN' => SExprBooleanNode,
+      'IDENTIFIER' => SExprIdentifierNode,
+      'INTEGER' => SExprIntegerNode,
+      'REAL' => SExprRealNode,
+      'STRING_LIT' => SExprStringNode
+    }.freeze
 
     # Create a new AST builder instance.
     # @param theTokens [Array<Token>] The sequence of input tokens.
@@ -28,20 +34,6 @@ module SrlRuby
     def terminal2node
       Terminal2NodeClass
     end
-
-
-    # Overriding method.
-    # Factory method for creating a node object for the given
-    # input token.
-    # @param _production [Rley::Syntax::Production]
-    # @param _terminal [Rley::Syntax::Terminal] Terminal symbol associated with the token
-    # @param aTokenPosition [Integer] Position of token in the input stream
-    # @param aToken [Rley::Lexical::Token] The input token
-    def new_leaf_node(_production, _terminal, aTokenPosition, aToken)
-      node = Rley::PTree::TerminalNode.new(aToken, aTokenPosition)
-
-      return node
-    end
   end # class
 end # module
 # End of file

data/lib/skeem/s_expr_nodes.rb

@@ -13,7 +13,7 @@ module Skeem
 
     # This method can be overriden
     def init_value(aValue)
-      self.value = aValue.dup
+      self.value = aValue
     end
 
     def symbol()
@@ -39,6 +39,7 @@ module Skeem
   end # class
   
   class SExprNumberNode < SExprTerminalNode
+  end # class  
   
   class SExprRealNode < SExprNumberNode
   end # class
@@ -47,6 +48,17 @@ module Skeem
   end # class
   
   class SExprStringNode < SExprTerminalNode
+    # Override
+    def init_value(aValue)
+      self.value = aValue.dup
+    end  
+  end # class
+  
+  class SExprIdentifierNode < SExprTerminalNode
+    # Override
+    def init_value(aValue)
+      self.value = aValue.dup
+    end  
   end # class
   
 =begin

data/lib/skeem/tokenizer.rb

@@ -77,7 +77,8 @@ module Skeem
         token = build_token('BOOLEAN', lexeme)
       elsif (lexeme = scanner.scan(/[+-]?[0-9]+(?=\s|[|()";]|$)/))
         token = build_token('INTEGER', lexeme) # Decimal radix
-      elsif (lexeme = scanner.scan(/[+-]?[0-9]+\.[0-9]+(?:(?:e|E)[+-]?[0-9]+)?/))
+      elsif (lexeme = scanner.scan(/[+-]?[0-9]+(?:\.[0-9]+)?(?:(?:e|E)[+-]?[0-9]+)?/))
+        # Order dependency: must be tested after INTEGER case
         token = build_token('REAL', lexeme)
       elsif (lexeme = scanner.scan(/"(?:\\"|[^"])*"/)) # Double quotes literal?
         token = build_token('STRING_LIT', lexeme)
@@ -130,8 +131,8 @@ module Skeem
         value = to_real(aLexeme, aFormat)
       when 'STRING_LIT'
         value = to_string(aLexeme, aFormat)
-      when 'SYMBOL'
-        value = to_string(aLexeme, aFormat)        
+      when 'IDENTIFIER'
+        value = to_identifier(aLexeme, aFormat)        
       else
         value = aLexeme
       end
@@ -170,7 +171,7 @@ module Skeem
       return value
     end  
 
-    def to_symbol(aLexeme, aFormat)
+    def to_identifier(aLexeme, aFormat)
       case aFormat
       when :default
         value = aLexeme

data/lib/skeem/version.rb

@@ -1,3 +1,3 @@
 module Skeem
-  VERSION = '0.0.4'.freeze
+  VERSION = '0.0.5'.freeze
 end

data/spec/skeem/parser_spec.rb

@@ -14,11 +14,79 @@ module Skeem
       end        
     end # context
     
-    context 'Parsing:' do
-      it 'should parse definitions' do
-        source = "(define r 10)"
-        expect { subject.parse(source) }.not_to raise_error
+    context 'Parsing literals:' do
+      it 'should parse isolated booleans' do
+        samples = [
+        ['#f', false],
+        ['#false', false],
+        ['#t', true],
+        ['#true', true]
+      ]
+        samples.each do |source, predicted|
+          ptree = subject.parse(source)
+          expect(ptree.root).to be_kind_of(SExprBooleanNode)
+          expect(ptree.root.value).to eq(predicted)
+        end 
       end
+      
+      it 'should parse isolated integers' do
+        samples = [
+        ['0', 0],
+        ['3', 3],
+        ['-3', -3],
+        ['+12345', 12345],
+        ['-12345', -12345]
+      ]
+        samples.each do |source, predicted|
+          ptree = subject.parse(source)
+          expect(ptree.root).to be_kind_of(SExprIntegerNode)
+          expect(ptree.root.value).to eq(predicted)
+        end 
+      end
+      
+      it 'should parse isolated real numbers' do
+        samples = [
+        ['0.0', 0.0],
+        ['3.14', 3.14],
+        ['-3.14', -3.14],
+        ['+123e+45', 123e+45],
+        ['-123e-45', -123e-45]
+      ]
+        samples.each do |source, predicted|
+          ptree = subject.parse(source)
+          expect(ptree.root).to be_kind_of(SExprRealNode)
+          expect(ptree.root.value).to eq(predicted)
+        end 
+      end
+
+      it 'should parse isolated strings' do
+        samples = [
+        ['"Hello world!"', 'Hello world!']
+      ]
+        samples.each do |source, predicted|
+          ptree = subject.parse(source)
+          expect(ptree.root).to be_kind_of(SExprStringNode)
+          expect(ptree.root.value).to eq(predicted)
+        end 
+      end
+
+      it 'should parse isolated identifiers' do
+        samples = [
+        ['the-word-recursion-has-many-meanings', 'the-word-recursion-has-many-meanings']
+      ]
+        samples.each do |source, predicted|
+          ptree = subject.parse(source)
+          expect(ptree.root).to be_kind_of(SExprIdentifierNode)
+          expect(ptree.root.value).to eq(predicted)
+        end 
+      end      
+    end # context
+
+    context 'Parsing forms:' do    
+      # it 'should parse definitions' do
+        # source = '(define r 10)'
+        # expect { subject.parse(source) }.not_to raise_error
+      # end
     end # context
   end # describe
 end # module

data/spec/skeem/tokenizer_spec.rb

@@ -84,7 +84,8 @@ module Skeem
           # couple [raw input, expected]
           ["\t\t3.45e+6", 3.45e+6],
           ['+3.45e+6', +3.45e+6],
-          ['-3.45e+6', -3.45e+6]
+          ['-3.45e+6', -3.45e+6],
+          ['123e+45', 1.23e+47]
         ]
 
         tests.each do |(input, prediction)|
@@ -96,7 +97,6 @@ module Skeem
       end
     end # context
 
-    # TODO
     context 'String recognition:' do
       it 'should tokenize strings' do
         examples = [

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: skeem
 version: !ruby/object:Gem::Version
-  version: 0.0.4
+  version: 0.0.5
 platform: ruby
 authors:
 - Dimitri Geshef
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-08-29 00:00:00.000000000 Z
+date: 2018-08-30 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rley
@@ -86,8 +86,8 @@ files:
 - appveyor.yml
 - lib/skeem.rb
 - lib/skeem/grammar.rb
-- lib/skeem/node_builder.rb
 - lib/skeem/parser.rb
+- lib/skeem/s_expr_builder.rb
 - lib/skeem/s_expr_nodes.rb
 - lib/skeem/stoken.rb
 - lib/skeem/tokenizer.rb