arugula 0.2.1 → 0.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 450b105a13d2bee3427f3ab008d1a0172e0bbacf
-  data.tar.gz: 091ebc89a888684923c2c05175c7370e43844a6f
+  metadata.gz: f59b134599b696a71d41c30d0bfc84fbab54e2c0
+  data.tar.gz: d3d49f304685d6f458284a712780a3274bc43b97
 SHA512:
-  metadata.gz: afb71d903455aa9d1508078919cccc34fd7cb1fb5ad7e15d3e39ab8170f9d31cfd12d21633098088b57e0003e99d267dd097d5acf8aca10901afac03c1266e9e
-  data.tar.gz: 27130b99003c17307830266fc1f2b29045d109c9bfaec2fc505f03e4978c3347e35e4945caebe4fe7437b16d7608c3c3510f910a0afc0443450fa26b598cfc5f
+  metadata.gz: 6cf01040b54906eaf6c9b4b4e6eff08843ad11fd6c4eb7da26865e068fb627ae9b9defd4382295120874e58bf68c48753f7c8dc80c2c2faf4bff7e1da542f883
+  data.tar.gz: 83b14f9ecf7a4cf2a8feef267d5599c999575db458bd35f3e08302414d4c16580d5234aa92c1d6e1edf09b544f49fc226fb0fc0b4626f650eaaf419b92a349a6

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    arugula (0.2.1)
+    arugula (0.3.0)
 
 GEM
   remote: https://rubygems.org/

data/lib/arugula.rb

@@ -2,16 +2,29 @@
 class Arugula
   require 'arugula/version'
 
+  attr_reader :captures
+
+  autoload :MatchData, 'arugula/match_data'
   autoload :Parser, 'arugula/parser'
 
   def initialize(pattern)
-    @root = Parser.new(pattern).parse!
+    @root, @captures = Parser.new(pattern).parse!
   end
 
   def match?(str, index = 0)
+    match_data = match(str, index)
+    match_data && match_data.start_index
+  end
+
+  def match(str, index = 0)
+    match_data = MatchData.new(self, str)
     loop do
-      match, = @root.match(str, index)
-      return index if match
+      match, end_index = @root.match(str, index, match_data)
+      if match
+        match_data.start_index = index
+        match_data.end_index = end_index
+        return match_data.freeze
+      end
       index += 1
       return if index > str.size
     end

data/lib/arugula/match_data.rb

@@ -0,0 +1,38 @@
+class Arugula
+  class MatchData
+    def initialize(regexp, string)
+      # require "awesome_print"
+      # ap regexp, raw: true
+      @regexp = regexp
+      @string = string.dup.freeze
+      @captures = Hash[regexp.captures.map { |c| [c.name, nil] }]
+    end
+
+    def add_capture(name, start_index, end_index)
+      @captures[name] = start_index...end_index
+    end
+
+    attr_accessor :start_index
+    attr_accessor :end_index
+
+    def to_s
+      @string[start_index...end_index]
+    end
+
+    def inspect
+      captures_part = @captures.map do |name, range|
+        " #{name}:#{@string[range].dump}"
+      end.join
+      "#<MatchData #{to_s.dump}#{captures_part}>"
+    end
+
+    def to_a
+      @captures.map { |_name, range| @string[range] }.unshift(to_s)
+    end
+
+    def freeze
+      @captures.freeze
+      super
+    end
+  end
+end

data/lib/arugula/parser.rb

@@ -6,6 +6,7 @@ class Arugula
     def initialize(str)
       @pattern = str.dup
       @states = [AndPart.new]
+      @captures = []
     end
 
     def state
@@ -14,13 +15,13 @@ class Arugula
 
     def parse!
       consume until pattern.empty?
-      @states.first
+      [@states.first, @captures]
     end
 
     Part.all.each do |part|
       type = part.type
       define_method(:"#{type}_type?") do
-        state && state.class.type == type
+        return true if state && state.class.type == type
       end
     end
 
@@ -56,6 +57,15 @@ class Arugula
         end
       elsif characterclass_type?
         push_part(:literal, tok)
+      elsif tok == '('
+        push_capture
+      elsif tok == ')'
+        pop_part until capture_type?
+        pop_part
+      elsif tok == '|'
+        pop_part until state == @states.first || or_type? || capture_type?
+        wrap_state(:or) unless or_type?
+        push_part(:and)
       elsif tok == '.'
         push_part(:dot)
       elsif tok == '*'
@@ -85,5 +95,11 @@ class Arugula
       @states.pop until @states.last.respond_to?(:parts)
       @states.pop
     end
+
+    def push_capture
+      push_part(:capture, @captures.size.succ.to_s)
+      @captures << state
+      push_part(:and)
+    end
   end
 end

data/lib/arugula/parts.rb

@@ -24,41 +24,44 @@ class Arugula
       literal.gsub('\\', '\\\\')
     end
 
-    def match(str, index)
+    def match(str, index, _match_data)
       length = literal.size
       matches = str[index, length] == literal
       [matches, index + (matches ? length : 0)]
     end
   end
 
-  class AndPart < Part
+  module MatchAll
     attr_accessor :parts
     def initialize
       @parts = []
     end
 
-    def to_s
-      parts.join
-    end
-
-    def match(str, index)
+    def match(str, index, match_data)
       parts.each do |part|
-        match, index = part.match(str, index)
+        match, index = part.match(str, index, match_data)
         return false, index unless match
       end
       [true, index]
     end
   end
 
+  class AndPart < Part
+    include MatchAll
+    def to_s
+      parts.join
+    end
+  end
+
   module MatchAny
     attr_accessor :parts
     def initialize
       @parts = []
     end
 
-    def match(str, index)
+    def match(str, index, match_data)
       parts.each do |part|
-        match, match_index = part.match(str, index)
+        match, match_index = part.match(str, index, match_data)
         return true, match_index if match
       end
       [false, index]
@@ -67,6 +70,11 @@ class Arugula
 
   class OrPart < Part
     include MatchAny
+    def initialize(*parts)
+      super()
+      @parts += parts
+    end
+
     def to_s
       parts.join '|'
     end
@@ -88,7 +96,7 @@ class Arugula
       "#{@range.begin}-#{@range.end}"
     end
 
-    def match(str, index)
+    def match(str, index, _match_data)
       matches = @range.member?(str[index])
       [matches, index + (matches ? 1 : 0)]
     end
@@ -114,8 +122,9 @@ class Arugula
       @metachar = metachar.to_sym
     end
 
-    def match(str, index)
-      [MATCHERS[@metachar][str, index], index + OFFSETS[@metachar]]
+    def match(str, index, _match_data)
+      matches = MATCHERS[@metachar][str, index]
+      [matches, index + (matches ? OFFSETS[@metachar] : 0)]
     end
 
     def to_s
@@ -123,15 +132,24 @@ class Arugula
     end
   end
 
-  class MatchPart < Part
-    attr_accessor :parts
-    def initialize
-      @parts = AndPart.new
+  class CapturePart < Part
+    include MatchAll
+    attr_reader :name
+
+    def initialize(name)
+      @name = name
+      super()
     end
 
     def to_s
       "(#{parts.join})"
     end
+
+    def match(str, index, match_data)
+      matches, end_index = super
+      match_data.add_capture(@name, index, end_index) if matches
+      [matches, end_index]
+    end
   end
 
   class EOLPart < Part
@@ -139,9 +157,9 @@ class Arugula
       '$'
     end
 
-    def match(str, index)
+    def match(str, index, _match_data)
       matches = str[index] == "\n" || index == str.size
-      return true, index + 1 if matches
+      return true, index if matches
       [false, index]
     end
   end
@@ -150,7 +168,7 @@ class Arugula
       '^'
     end
 
-    def match(str, index)
+    def match(str, index, _match_data)
       matches = (index == 0) || (str[index - 1] == "\n")
       [matches, index]
     end
@@ -169,9 +187,9 @@ class Arugula
       "#{wrapped}*"
     end
 
-    def match(str, index)
+    def match(str, index, match_data)
       loop do
-        matches, index = wrapped.match(str, index)
+        matches, index = wrapped.match(str, index, match_data)
         return true, index unless matches
       end
     end
@@ -183,10 +201,10 @@ class Arugula
       "#{wrapped}+"
     end
 
-    def match(str, index)
+    def match(str, index, match_data)
       has_matched = false
       loop do
-        matches, index = wrapped.match(str, index)
+        matches, index = wrapped.match(str, index, match_data)
         has_matched = true if matches
         return has_matched, index unless matches
       end
@@ -198,7 +216,7 @@ class Arugula
       '.'
     end
 
-    def match(str, index)
+    def match(str, index, _match_data)
       matches = index < str.size
       [matches, index + (matches ? 1 : 0)]
     end

data/lib/arugula/version.rb

@@ -1,4 +1,4 @@
 # frozen_string_literal: true
 class Arugula
-  VERSION = '0.2.1'.freeze
+  VERSION = '0.3.0'.freeze
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: arugula
 version: !ruby/object:Gem::Version
-  version: 0.2.1
+  version: 0.3.0
 platform: ruby
 authors:
 - Samuel Giddins
@@ -14,42 +14,42 @@ dependencies:
   name: bundler
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.11'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.11'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '10.5'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '10.5'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '3.4'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '3.4'
 description: 
@@ -59,10 +59,10 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- .gitignore
-- .rspec
-- .rubocop.yml
-- .travis.yml
+- ".gitignore"
+- ".rspec"
+- ".rubocop.yml"
+- ".travis.yml"
 - CODE_OF_CONDUCT.md
 - Gemfile
 - Gemfile.lock
@@ -76,6 +76,7 @@ files:
 - bin/rubocop
 - bin/setup
 - lib/arugula.rb
+- lib/arugula/match_data.rb
 - lib/arugula/parser.rb
 - lib/arugula/parts.rb
 - lib/arugula/version.rb
@@ -89,12 +90,12 @@ require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []