srl_ruby 0.2.6 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 475a892be14d441ff82714324f489707839eae57
-  data.tar.gz: 52ad8fffe6fb8abe17e99e690dc182bb0b3a48b0
+  metadata.gz: 17b59ca17f397a24c01b84bd9d1a4a67743e2f73
+  data.tar.gz: 8b6b397b9d5e95bf1f5a6f6b5cdd866d7a8e966d
 SHA512:
-  metadata.gz: 24b028b5becb5f2ac57c6fbf1aaa5eb45d7b3e3bde1e8ea2892b56f7d67fc643bf8d28cff0e431e7a3f97c996b0c44321bfd52eb17d818daeb51a334cb2d056b
-  data.tar.gz: b21b2fefc35eac997bd06a0a4a7bc586d80eba2074cfb365c7a22a4e6390e48a45960df3132b09abc2e53569c5c6ad4f925510191808212ba6a8b991dfc259a8
+  metadata.gz: 5b7e2d0a99199a431eb2c381f566171161e7c7aaa6098211d9740e3f4bbc86568aafa60d6f41a5e334a35c81cddffc2260e11c1a6a8392e39f8095f7cba05571
+  data.tar.gz: 6d7cd5fe49e6277558f2cfaa2601ee515cb14556fc5d8c4b831d853f4f77f52301758166618b0cf3e7f5ae6cf9cab957052ad7cb0f5f53a323726c4bb307824d

data/CHANGELOG.md

@@ -1,3 +1,13 @@
+## [0.3.0] - 2018-04-04
+Version bump: SrlRuby passes the complete official SRL test suite! 
+### Changed
+- File `acceptance/srl_test_suite_spec.rb`. 15 test files from official test suite are passing.
+
+### Fixed
+- Class `SrlRuby::ASTBuilder` Fixing the capture...until semantic.
+  If last captured subexpression is a repetition, then it is made lazy (instead of greedy).
+
+
 ## [0.2.6] - 2018-04-03
 SrlRuby passes 13 tests out of 15 from standard SRL test suite. 
 ### Changed

data/lib/regex/non_capturing_group.rb

@@ -7,10 +7,12 @@ module Regex # This module is used as a namespace
   # of sub-expressions
   class NonCapturingGroup < MonadicExpression
     # Constructor.
-    # [aChildExpression]  A sub-expression to match. When successful
+    # [aChild]  A sub-expression to match. When successful
     # the matching text is assigned to the capture variable.
-    def initialize(aChildExpression)
-      super(aChildExpression)
+    def initialize(aChild)
+      # If necessary get rid of nested non-capturing groups
+      effective_child = aChild.kind_of?(self.class) ? aChild.child : aChild
+      super(effective_child)
     end
 
     protected

data/lib/srl_ruby/ast_builder.rb

@@ -367,9 +367,31 @@ module SrlRuby
       return Regex::CapturingGroup.new(theChildren[1])
     end
 
+    # If the rightmost (sub)expression is a repetition, then make it lazy
+    def make_last_repetition_lazy(anExpr)
+      sub_expr = anExpr
+      loop do
+        if sub_expr.is_a?(Regex::Repetition)
+            # Make repetition lazy
+            cardinality = sub_expr.multiplicity
+            cardinality.instance_variable_set(:@policy, :lazy)
+            break
+        elsif sub_expr.kind_of?(Regex::PolyadicExpression)
+          sub_expr = sub_expr.children.last
+        elsif sub_expr.kind_of?(Regex::MonadicExpression)
+          sub_expr = sub_expr.child
+        elsif sub_expr.kind_of?(Regex::AtomicExpression)
+          break
+        end
+      end
+    end
+
     # rule('capturing_group' => %w[CAPTURE assertable UNTIL assertable]).as
     #   'capture_until'
     def reduce_capture_until(_production, _range, _tokens, theChildren)
+      # Until semantic requires that the last pattern in capture to be lazy
+      make_last_repetition_lazy(theChildren[1])
+
       group = Regex::CapturingGroup.new(theChildren[1])
       return Regex::Concatenation.new(group, theChildren[3])
     end
@@ -384,6 +406,9 @@ module SrlRuby
     # rule('capturing_group' => %w[CAPTURE assertable AS var_name
     #   UNTIL assertable]).as 'named_capture_until'
     def reduce_named_capture_until(_production, _range, _tokens, theChildren)
+      # Until semantic requires that the last pattern in capture to be lazy
+      make_last_repetition_lazy(theChildren[1])
+
       name = theChildren[3].token.lexeme.dup
       group = Regex::CapturingGroup.new(theChildren[1], name)
       return Regex::Concatenation.new(group, theChildren[5])

data/lib/srl_ruby/version.rb

@@ -1,3 +1,3 @@
 module SrlRuby
-  VERSION = '0.2.6'.freeze
+  VERSION = '0.3.0'.freeze
 end

data/spec/acceptance/srl_test_suite_spec.rb

@@ -10,7 +10,7 @@ require_relative '../../lib/srl_ruby'
 # | word.rule        | '(word)'           |
 
 
-RSpec.describe Acceptance do
+RSpec.describe SrlRuby do
   def rule_path
     __FILE__.sub(/spec\/.+$/, 'srl_test/Test-Rules/')
   end
@@ -19,8 +19,12 @@ RSpec.describe Acceptance do
     return Acceptance::RuleFileParser.load_file(rule_path + aFilename)
   end
 
+#   
+#   CaptureTest = Struct.new(:test_string, :expectations)
+   
   def test_rule_file(aRuleFileRepr)
     regex = SrlRuby::parse(aRuleFileRepr.srl.value)
+    puts regex.source
     expect(regex).to be_kind_of(Regexp)
 
     aRuleFileRepr.match_tests.each do |test|
@@ -29,16 +33,39 @@ RSpec.describe Acceptance do
     aRuleFileRepr.no_match_tests.each do |test|
       expect(regex.match(test.test_string.value)).to be_nil
     end
-    aRuleFileRepr.capture_tests.each do |test|
-      matching = regex.match(test.test_string.value)
-      expect(matching).not_to be_nil
-      test.expectations do |exp|
-        var = exp.var_name.value.to_s
-        captured = exp.captured_text.value
-        name_index = matching.names.index(var)
-        expect(name_index).not_to be_nil
-        expect(matching.captures[name_index]).to eq(captured)
+    aRuleFileRepr.capture_tests.each do |cp_test|
+      test_string = cp_test.test_string.value
+      expect(test_string).to match(regex)
+      if regex.names.empty?
+        indices = cp_test.expectations.map { |exp| exp.result_index.value.to_s }
+        actual_names = indices.uniq.sort
+      else
+        actual_names = regex.names
       end
+      
+      # CaptureExpectation = Struct.new(:result_index, :var_name, :captured_text)
+      # Compare actual vs. expected capture names
+      cp_test.expectations.each do |expec|
+        expected_name = expec.var_name.value.to_s
+        unless actual_names.empty?
+          expect(actual_names).to be_include(expected_name)
+        end
+      end
+      
+      scan_results = test_string.scan(regex)
+      actual_captures = scan_results.map do |capture_tuples|
+        actual_names.zip(capture_tuples).to_h
+      end
+      
+      # Compare actual vs. expected captured texts
+      cp_test.expectations.each do |expec|
+        index =  expec.result_index.value
+        var_name = expec.var_name.value.to_s
+        expected_capture = expec.captured_text.value
+        names2val = actual_captures[index]
+        actual = names2val[var_name].nil? ? "" : names2val[var_name]
+        expect(actual).to eq(expected_capture )
+      end      
     end
   end
 
@@ -76,6 +103,11 @@ RSpec.describe Acceptance do
     rule_file_repr = load_file('none_of.rule')
     test_rule_file(rule_file_repr)
   end
+  
+  it 'should support negative character class' do
+    rule_file_repr = load_file('sample_capture.rule')
+    test_rule_file(rule_file_repr)
+  end
 
   it 'should match a tab' do
     rule_file_repr = load_file('tab.rule')
@@ -87,6 +119,11 @@ RSpec.describe Acceptance do
     test_rule_file(rule_file_repr)
   end
 
+  it 'should match mail address' do
+    rule_file_repr = load_file('website_example_email_capture.rule')
+    test_rule_file(rule_file_repr)
+  end  
+ 
   it 'should support lookahead' do
     rule_file_repr = load_file('website_example_lookahead.rule')
     test_rule_file(rule_file_repr)
@@ -106,4 +143,4 @@ RSpec.describe Acceptance do
     rule_file_repr = load_file('word.rule')
     test_rule_file(rule_file_repr)
   end  
-end
+end # describe

data/spec/srl_ruby_spec.rb

@@ -304,7 +304,7 @@ END_SRL
 
     it 'should parse simple anonymous until capturing group' do
       regexp = SrlRuby.parse('capture anything once or more until literally "!"')
-      expect(regexp.source).to eq('(.+)!')
+      expect(regexp.source).to eq('(.+?)!')
     end
 
     it 'should parse unquoted named capturing group' do
@@ -335,7 +335,7 @@ END_SRL
     it 'should parse complex named until capturing group' do
       source = 'capture (anything once or more) as "foo" until literally "m"'
       regexp = SrlRuby.parse(source)
-      expect(regexp.source).to eq('(?<foo>.+)m')
+      expect(regexp.source).to eq('(?<foo>.+?)m')
     end
   end # context
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: srl_ruby
 version: !ruby/object:Gem::Version
-  version: 0.2.6
+  version: 0.3.0
 platform: ruby
 authors:
 - Dimitri Geshef
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2018-04-03 00:00:00.000000000 Z
+date: 2018-04-04 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rley