skeem 0.0.2 → 0.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: a4521e994b28e81e3fa1f8852bf096c2e375def5
-  data.tar.gz: 86e7e0738987b88b9043740632b510e5bee6a49f
+  metadata.gz: 670bb82d2fecac647938c03d92b8c910f9385a49
+  data.tar.gz: 3ac772b051b36ded30aefe62c6745239c9a1941f
 SHA512:
-  metadata.gz: 0197b9df68cba199a243b53d22f329728575ec2c8021b813f6f3eb4be4e39d530cfc53fa221be7994a636da06fdc1ad17977985d7d4e0c54dd3f4dda1c8b6cae
-  data.tar.gz: 493fb5dd6786dda700abf79ea211d96b832772540986b36eca887ddb2abba4816954080fb5508fdc55dc9ab4b375b5b129f031f507efcf9caf081e05b28ba10f
+  metadata.gz: 6b8488389ba2a11fa1664bb6f3e2a40c752a16045061f516360cce2744e9e05e2746d823f8854fa40f23fe24183105384863f19156a7b425a86d95071d2ad9c4
+  data.tar.gz: 1c6029799408c83027072cdfca72f9d86dd6bbf5519af149fbb3736e4224f223f553c1463e56fd0461b0e977ae91a2be14ab2a8f4355e9a1ae5449094167a08b

data/CHANGELOG.md

@@ -1,3 +1,12 @@
+## [0.0.3] - 2018-08-25
+### Added
+- File `grammar.rb` with minimalist grammar.
+- Initial `Parser` class commit
+
+### Changed
+- Class`Tokenizer` recognizes `define` keyword
+- Spec file `Tokenizer_spec.rb` expanded with more tests.
+
 ## [0.0.2] - 2018-08-25
 ### Changed
 - Class`Tokenizer` improved, does recognize delimiters, booleans, integers, real numbers, strings, and identifiers.

data/README.md

@@ -1,4 +1,8 @@
 # Skeem
+[![Linux Build Status](https://travis-ci.org/famished-tiger/Skeem.svg?branch=master)](https://travis-ci.org/famished-tiger/Skeem)
+[![Gem Version](https://badge.fury.io/rb/skeem.svg)](https://badge.fury.io/rb/skeem)
+[![License](https://img.shields.io/badge/license-MIT-brightgreen.svg?style=flat)](https://github.com/famished-tiger/Skeem/blob/master/LICENSE.txt)
+
 __Skeem__ will be an interpreter of a subset of the Scheme programming language.
 
 
@@ -20,8 +24,12 @@ Or install it yourself as:
 
 ## Usage
 
-The __Skeem__ project has just started.
-At this stage, it is too early to use the gem.  
+The __Skeem__ project has just started.  
+At this stage, the gem consists of a tokenizer.  
+At least these steps must be done in order to have a first interpreter:
+- TODO: define subset of Scheme grammar
+- TODO: buuild parser of Scheme subset
+- TODO: Implement Skeem interpreter (and REPL)
 
 Roadmap:
 - Implement an equivalent of [lis.py](http://www.norvig.com/lispy.html)
@@ -33,7 +41,7 @@ Roadmap:
 TODO: Write usage instructions here
 
 Good to know:
-Online book: [The Scheme Programming Language (4th Ed.)](https://www.scheme.com/tspl4/)
+Online book: [The Scheme Programming Language (4th Ed.)](https://www.scheme.com/tspl4/). Remark: covers an older version of Scheme.
   
 ## Development
 

data/lib/skeem.rb

@@ -1,5 +1,8 @@
-require 'skeem/version'
+# File: skeem.rb
+# This file acts as a jumping-off point for loading dependencies expected
+# for a Skeem client.
 
-module Skeem
-  # Your code goes here...
-end
+require_relative './skeem/version'
+require_relative './skeem/parser'
+
+# End of file

data/lib/skeem/grammar.rb

@@ -0,0 +1,42 @@
+# Grammar for Skeem (a subset of Scheme language)
+require 'rley' # Load the gem
+
+module Skeem
+  ########################################
+  # Define a grammar for Skeem
+  # Official Small Scheme grammar is available at:
+  # https://bitbucket.org/cowan/r7rs/src/draft-10/rnrs/r7rs.pdf
+  # Names of grammar elements are based on the R7RS documentation
+  builder = Rley::Syntax::GrammarBuilder.new do
+    # Delimitersn, separators...
+    # add_terminals('APOSTROPHE', 'BACKQUOTE')    
+    add_terminals('LPAREN', 'RPAREN')
+    # add_terminals('PERIOD')
+
+    # Literal values...
+    add_terminals('BOOLEAN', 'INTEGER', 'REAL')
+    add_terminals('STRING_LIT', 'IDENTIFIER')
+
+    # Keywords...
+    # add_terminals('BEGIN', 'DEFINE')
+    add_terminals('DEFINE')
+    
+    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' => 'definition'
+    rule 'definition' => 'LPAREN DEFINE IDENTIFIER expression RPAREN'
+    rule 'expression' =>  'IDENTIFIER'
+    rule 'expression' =>  'literal'
+    rule 'literal' => 'self-evaluating'
+    rule 'self-evaluating' => 'BOOLEAN'
+    rule 'self-evaluating' => 'number'
+    rule 'self-evaluating' => 'STRING_LIT'
+    rule 'number' => 'INTEGER'
+    rule 'number' => 'REAL'
+  end
+
+  # And now build the grammar and make it accessible via a global constant
+  # [Rley::Syntax::Grammar]
+  Grammar = builder.grammar
+end # module

data/lib/skeem/parser.rb

@@ -0,0 +1,36 @@
+require_relative 'tokenizer'
+require_relative 'grammar'
+
+module Skeem
+  class Parser
+    attr_reader(:engine)
+    
+    def initialize()
+      # Create a Rley facade object
+      @engine = Rley::Engine.new { |cfg| cfg.diagnose = true }
+
+      # Step 1. Load Skeem grammar
+      @engine.use_grammar(Skeem::Grammar)
+    end
+      
+    # Parse the given Skeem expression into a parse tree.
+    # @param source [String] Skeem expression to parse
+    # @return [ParseTree] A regexp object equivalent to the Skeem expression.
+    # @example Defining a function that computes the area of a circle
+    #   source = "(define circle-area (lambda (r) (* pi (* r r))))"
+    #   regex = Skeem::parse(source)
+    def parse(source)
+      lexer = Skeem::Tokenizer.new(source)
+      result = engine.parse(lexer.tokens)
+
+      unless result.success?
+        # Stop if the parse failed...
+        line1 = "Parsing failed\n"
+        line2 = "Reason: #{result.failure_reason.message}"
+        raise StandardError, line1 + line2
+      end
+
+      return result
+    end
+  end # class
+end # module

data/lib/skeem/tokenizer.rb

@@ -24,6 +24,12 @@ module Skeem
       '(' => 'LPAREN',
       ')' => 'RPAREN'
     }.freeze
+    
+    # Here are all the SRL keywords (in uppercase)
+    @@keywords = %w[
+      BEGIN
+      DEFINE
+    ].map { |x| [x, x] } .to_h      
 
     class ScanError < StandardError; end
 
@@ -77,7 +83,9 @@ module Skeem
         unquoted = lexeme.gsub(/(^")|("$)/, '')
         token = build_token('STRING_LIT', unquoted)
       elsif (lexeme = scanner.scan(/[a-zA-Z!$%&*\/:<=>?@^_~][a-zA-Z0-9!$%&*+-.\/:<=>?@^_~+-]*/))
-        token = build_token('IDENTIFIER', lexeme)
+        keyw = @@keywords[lexeme.upcase]
+        tok_type = keyw ? keyw : 'IDENTIFIER'
+        token = build_token(tok_type, lexeme)
       elsif (lexeme = scanner.scan(/\|(?:[^|])*\|/)) # Vertical bar delimited
         token = build_token('IDENTIFIER', lexeme)
       elsif (lexeme = scanner.scan(/([\+\-])((?=\s|[|()";])|$)/))

data/lib/skeem/version.rb

@@ -1,3 +1,3 @@
 module Skeem
-  VERSION = '0.0.2'.freeze
+  VERSION = '0.0.3'.freeze
 end

data/spec/skeem/parser_spec.rb

@@ -0,0 +1,26 @@
+require_relative '../spec_helper' # Use the RSpec framework
+require_relative '../../lib/skeem/tokenizer' # Load the class under test
+
+module Skeem
+  describe Parser do
+  
+    context 'Initialization:' do
+      it 'should be initialized without argument' do
+        expect { Parser.new() }.not_to raise_error
+      end
+
+      it 'should have its enginer initialized' do
+        expect(subject.engine).to be_kind_of(Rley::Engine)
+      end        
+    end # context
+    
+    context 'Parsing:' 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
+
+# End of file

data/spec/skeem/tokenizer_spec.rb

@@ -10,7 +10,7 @@ module Skeem
         expect(token.lexeme).to eq(lexeme)
       end
     end
-    
+
     def unquoted(aString)
       aString.gsub(/(^")|("$)/, '')
     end
@@ -141,18 +141,64 @@ containing just one line"
         end
       end
     end # context
-    
+
     context 'Scanning Scheme sample code' do
       it 'should read examples from lis.py page' do
+        # Shallow tokenizer testing
         source = <<-SCHEME
-(if (> (val x) 0) 
-    (fn (+ (aref A i) (* 3 i)) 
+(if (> (val x) 0)
+    (fn (+ (aref A i) (* 3 i))
         (quote (one two)))
       end
     end
 SCHEME
         subject.reinitialize(source)
         expect { subject.tokens }.not_to raise_error
+
+        source = "(define circle-area (lambda (r) (* pi (* r r))))"
+        subject.reinitialize(source)
+        expect { subject.tokens }.not_to raise_error
+
+        source = "(define fact (lambda (n) (if (<= n 1) 1 (* n (fact (- n 1))))))"
+        subject.reinitialize(source)
+        expect { subject.tokens }.not_to raise_error
+
+        source = <<-SCHEME
+define first car)
+define rest cdr)
+define count (lambda (item L) (if L (+ (equal? item (first L)) (count item (rest L))) 0)))
+count 0 (list 0 1 2 3 0 0))
+SCHEME
+        subject.reinitialize(source)
+        expect { subject.tokens }.not_to raise_error
+      end
+
+      it 'should produce a sequence of token objects' do
+        # Deeper tokenizer testing
+        source = "(define circle-area (lambda (r) (* pi (* r r))))"
+        subject.reinitialize(source)
+        predicted = [
+          ['LPAREN', '('],
+          ['DEFINE', 'define'],
+          ['IDENTIFIER', 'circle-area'],
+          ['LPAREN', '('],
+          ['IDENTIFIER', 'lambda'],
+          ['LPAREN', '('],
+          ['IDENTIFIER', 'r'],
+          ['RPAREN', ')'],
+          ['LPAREN', '('],
+          ['IDENTIFIER', '*'],
+          ['IDENTIFIER', 'pi'],
+          ['LPAREN', '('],
+          ['IDENTIFIER', '*'],
+          ['IDENTIFIER', 'r'],
+          ['IDENTIFIER', 'r'],
+          ['RPAREN', ')'],
+          ['RPAREN', ')'],
+          ['RPAREN', ')'],
+          ['RPAREN', ')']
+        ]
+        match_expectations(subject, predicted)
       end
     end # context
   end # describe

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: skeem
 version: !ruby/object:Gem::Version
-  version: 0.0.2
+  version: 0.0.3
 platform: ruby
 authors:
 - Dimitri Geshef
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-08-25 00:00:00.000000000 Z
+date: 2018-08-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rley
@@ -85,10 +85,13 @@ files:
 - Rakefile
 - appveyor.yml
 - lib/skeem.rb
+- lib/skeem/grammar.rb
+- lib/skeem/parser.rb
 - lib/skeem/stoken.rb
 - lib/skeem/tokenizer.rb
 - lib/skeem/version.rb
 - skeem.gemspec
+- spec/skeem/parser_spec.rb
 - spec/skeem/tokenizer_spec.rb
 - spec/skeem_spec.rb
 - spec/spec_helper.rb
@@ -119,5 +122,6 @@ specification_version: 4
 summary: Skeem is an interpreter of a subset of the Scheme programming language. Scheme
   is a descendent of the Lisp language.
 test_files:
+- spec/skeem/parser_spec.rb
 - spec/skeem/tokenizer_spec.rb
 - spec/skeem_spec.rb