lrama 0.4.0 → 0.5.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: ef1a46b0de83840d8353cf0e832898b7567a6a546b5ab8402cee59be5178d3af
-  data.tar.gz: 81e546a5f907ed2a1a2ca4e1028f5be05df8251bd59fb8c7d8a3a7565e0f9dce
+  metadata.gz: 20a0927bedeb14281abb8a5c59daf643fdfe556b5899cbe4872878b022804da2
+  data.tar.gz: 731fdd0201266db9724a2ebcc1162427a2cfaa9d0d94543e1298b3f2683eb14f
 SHA512:
-  metadata.gz: d388d548999a8e96d08698d191a44ea5c71893060eef624a1fd7845cfe23f96f27720646897fc12930f948f3dc27771a0e69ec26f57b604711ebe537984cb752
-  data.tar.gz: afe54b54f3852fd65c243bff4a800a2186bc89cdfc92a83bae85eabf10aae7b4989d9daea9904477bff109fd48ae38af0aefaf1e10d245cb74f107397db5ae55
+  metadata.gz: 3a6e7c0ef4e7266dae41dba15e3a5d7609cb7171b82e0dbd5b75595e8e3cf8a4a402c5b710cc0f714c54357ef9fc20fd0935e55b8c529c99cd3c09ea173efb87
+  data.tar.gz: fa867c9be43384c7f282bf22ecd2b0939cd0dca263a51914871dba42f409ddc01fb00c34d8696288c3668177fdc450ff1bc9fb5e90fdf053e2d56a5bf31bbe04

data/.github/dependabot.yml

@@ -0,0 +1,6 @@
+version: 2
+updates:
+  - package-ecosystem: 'github-actions'
+    directory: '/'
+    schedule:
+      interval: 'weekly'

data/.github/workflows/test.yaml

@@ -22,6 +22,20 @@ jobs:
           bundler-cache: true
       - run: bundle install
       - run: bundle exec rspec
+  steep-check:
+    runs-on: ubuntu-20.04
+    strategy:
+      fail-fast: false
+      matrix:
+        ruby: ['head']
+    steps:
+      - uses: actions/checkout@v3
+      - uses: ruby/setup-ruby@v1
+        with:
+          ruby-version: ${{ matrix.ruby }}
+          bundler-cache: true
+      - run: bundle install
+      - run: bundle exec steep check
   test-ruby:
     runs-on: ubuntu-20.04
     strategy:
@@ -40,10 +54,12 @@ jobs:
           bundler-cache: true
       - run: git clone --depth=1 https://github.com/ruby/ruby.git -b ${{ matrix.ruby_branch }} ../ruby
         working-directory:
-      - run: mkdir tool/lrama
+      - run: mkdir -p tool/lrama
         working-directory: ../ruby
       - name: Copy Lrama to ruby/tool
-        run: cp -r exe lib template ../ruby/tool/lrama
+        run: cp -r exe lib ../ruby/tool/lrama
+        # TODO: Consider how to manage changes on ruby/ruby master and ruby/lrama
+        # run: cp -r exe lib template ../ruby/tool/lrama
         working-directory:
       - run: tree tool/lrama
         working-directory: ../ruby
@@ -68,5 +84,5 @@ jobs:
       - run: sudo apt-get --purge remove bison
       - run: ../autogen.sh
       - run: ../configure -C --disable-install-doc
-      - run: make YACC=$(readlink -f $(pwd)/../tool/lrama/exe/lrama)
+      - run: make
       - run: make test-all

data/Gemfile

@@ -6,3 +6,5 @@ gem "rspec"
 gem "pry"
 gem "stackprof"
 gem "rake"
+gem "rbs", require: false
+gem "steep", require: false

data/README.md

@@ -21,9 +21,11 @@ $ gem install lrama
 From source codes,
 
 ```shell
+$ cd "$(lrama root)"
+$ bundle install
 $ bundle exec rake install
-$ lrama --version
-0.1.0
+$ bundle exec lrama --version
+0.5.0
 ```
 
 ## Usage
@@ -33,11 +35,36 @@ $ lrama --version
 $ lrama -d sample/parse.y
 ```
 
+```shell
+# "calc", "calc.c", and "calc.h" are generated
+$ lrama -d sample/calc.y -o calc.c && gcc -Wall calc.c -o calc && ./calc
+Enter the formula:
+1
+=> 1
+1+2*3
+=> 7
+(1+2)*3
+=> 9
+```
+
+## Versions and Branches
+
+### v0_4 (`lrama_0_4` branch)
+
+This branch generates "parse.c" compatible with Bison 3.8.2 for ruby 3.0, 3.1, 3.2. The first version migrated to ruby is ["0.4.0"](https://github.com/ruby/ruby/pull/7798) therefore keep this branch for Bison compatible branch.
+
 ## Build Ruby
 
 1. Install Lrama
 2. Run `make YACC=lrama`
 
+## Release flow
+
+1. Update `Lrama::VERSION`
+2. Release as a gem by `rake release`
+3. Update Lrama in ruby/ruby by `cp -r exe lib ruby/tool/lrama`
+4. Create new release on [GitHub](https://github.com/ruby/lrama/releases)
+
 ## License
 
 See LEGAL.md file.

data/Steepfile

@@ -0,0 +1,7 @@
+# D = Steep::Diagnostic
+#
+target :lib do
+  signature "sig"
+
+  check "lib/lrama/bitmap.rb"
+end

data/lib/lrama/command.rb

@@ -57,14 +57,18 @@ module Lrama
       end
 
       if !grammar_file
-        puts "File should be specified\n"
-        exit 1
+        abort "File should be specified\n"
       end
 
       Report::Duration.enable if trace_opts[:time]
 
       warning = Lrama::Warning.new
-      y = File.read(grammar_file)
+      if grammar_file == '-'
+        grammar_file = argv.shift or abort "File name for STDIN should be specified\n"
+        y = STDIN.read
+      else
+        y = File.read(grammar_file)
+      end
       grammar = Lrama::Parser.new(y).parse
       states = Lrama::States.new(grammar, warning, trace_state: (trace_opts[:automaton] || trace_opts[:closure]))
       states.compute

data/lib/lrama/grammar.rb

@@ -166,7 +166,7 @@ module Lrama
         when ref.type == :at # @n
           raise "@#{ref.number} can not be used in %printer."
         else
-          raise "Unexpected. #{code}, #{ref}"
+          raise "Unexpected. #{self}, #{ref}"
         end
 
         t_code[first_column..last_column] = str
@@ -205,7 +205,7 @@ module Lrama
           i = -ref.position_in_rhs + ref.number
           str = "(yylsp[#{i}])"
         else
-          raise "Unexpected. #{code}, #{ref}"
+          raise "Unexpected. #{self}, #{ref}"
         end
 
         t_code[first_column..last_column] = str
@@ -235,7 +235,7 @@ module Lrama
         when ref.type == :at # @n
           raise "@#{ref.number} can not be used in initial_action."
         else
-          raise "Unexpected. #{code}, #{ref}"
+          raise "Unexpected. #{self}, #{ref}"
         end
 
         t_code[first_column..last_column] = str
@@ -716,7 +716,7 @@ module Lrama
         # If id is Token::Char, it uses ASCII code
         if sym.term? && sym.token_id.nil?
           if sym.id.type == Token::Char
-            # Igonre ' on the both sides
+            # Ignore ' on the both sides
             case sym.id.s_value[1..-2]
             when "\\b"
               sym.token_id = 8
@@ -844,7 +844,7 @@ module Lrama
 
       return if invalid.empty?
 
-      raise "Symbol number is dupulicated. #{invalid}"
+      raise "Symbol number is duplicated. #{invalid}"
     end
   end
 end

data/lib/lrama/lexer.rb

@@ -206,6 +206,8 @@ module Lrama
         when ss.scan(/\/\*/)
           # TODO: Need to keep comment?
           line = lex_comment(ss, line, lines, "")
+        when ss.scan(/\/\//)
+          line = lex_line_comment(ss, line, "")
         when ss.scan(/'(.)'/)
           tokens << create_token(Token::Char, ss[0], line, ss.pos - column)
         when ss.scan(/'\\(.)'/) # '\\', '\t'
@@ -218,7 +220,7 @@ module Lrama
           l = line - lines.first[1]
           split = ss.string.split("\n")
           col = ss.pos - split[0...l].join("\n").length
-          raise "Parse error (unknow token): #{split[l]} \"#{ss.string[ss.pos]}\" (#{line}: #{col})"
+          raise "Parse error (unknown token): #{split[l]} \"#{ss.string[ss.pos]}\" (#{line}: #{col})"
         end
       end
     end
@@ -276,6 +278,9 @@ module Lrama
         when ss.scan(/\/\*/)
           str << ss[0]
           line = lex_comment(ss, line, lines, str)
+        when ss.scan(/\/\//)
+          str << ss[0]
+          line = lex_line_comment(ss, line, str)
         else
           # noop, just consume char
           str << ss.getch
@@ -314,8 +319,6 @@ module Lrama
       raise "Parse error (quote mismatch): #{ss.string.split("\n")[l]} \"#{ss.string[ss.pos]}\" (#{line}: #{ss.pos})"
     end
 
-    # TODO: Need to handle // style comment
-    #
     # /*  */ style comment
     def lex_comment(ss, line, lines, str)
       while !ss.eos? do
@@ -337,6 +340,23 @@ module Lrama
       raise "Parse error (comment mismatch): #{ss.string.split("\n")[l]} \"#{ss.string[ss.pos]}\" (#{line}: #{ss.pos})"
     end
 
+    # // style comment
+    def lex_line_comment(ss, line, str)
+      while !ss.eos? do
+        case
+        when ss.scan(/\n/)
+          return line + 1
+        else
+          str << ss.getch
+          next
+        end
+
+        str << ss[0]
+      end
+
+      line # Reach to end of input
+    end
+
     def lex_grammar_rules_tokens
       lex_common(@grammar_rules, @grammar_rules_tokens)
     end

data/lib/lrama/output.rb

@@ -53,9 +53,7 @@ module Lrama
           if @header_out
             @header_out << tmp
           else
-            File.open(@header_file_path, "w+") do |f|
-              f << tmp
-            end
+            File.write(@header_file_path, tmp)
           end
         end
       end

data/lib/lrama/parser/token_scanner.rb

@@ -0,0 +1,55 @@
+module Lrama
+  class Parser
+    class TokenScanner
+      def initialize(tokens)
+        @tokens = tokens
+        @index = 0
+      end
+
+      def current_token
+        @tokens[@index]
+      end
+
+      def current_type
+        current_token && current_token.type
+      end
+
+      def next
+        token = current_token
+        @index += 1
+        return token
+      end
+
+      def consume(*token_types)
+        if token_types.include?(current_type)
+          token = current_token
+          self.next
+          return token
+        end
+
+        return nil
+      end
+
+      def consume!(*token_types)
+        consume(*token_types) || (raise "#{token_types} is expected but #{current_type}. #{current_token}")
+      end
+
+      def consume_multi(*token_types)
+        a = []
+
+        while token_types.include?(current_type)
+          a << current_token
+          self.next
+        end
+
+        raise "No token is consumed. #{token_types}" if a.empty?
+
+        return a
+      end
+
+      def eots?
+        current_token.nil?
+      end
+    end
+  end
+end

data/lib/lrama/parser.rb

@@ -1,4 +1,5 @@
 require "lrama/report"
+require "lrama/parser/token_scanner"
 
 module Lrama
   # Parser for parse.y, generates a grammar
@@ -7,58 +8,6 @@ module Lrama
 
     T = Lrama::Lexer::Token
 
-    class TokenScanner
-      def initialize(tokens)
-        @tokens = tokens
-        @index = 0
-      end
-
-      def current_token
-        @tokens[@index]
-      end
-
-      def current_type
-        current_token && current_token.type
-      end
-
-      def next
-        token = current_token
-        @index += 1
-        return token
-      end
-
-      def consume(*token_types)
-        if token_types.include?(current_type)
-          token = current_token
-          self.next
-          return token
-        end
-
-        return nil
-      end
-
-      def consume!(*token_types)
-        consume(*token_types) || (raise "#{token_types} is expected but #{current_type}. #{current_token}")
-      end
-
-      def consume_multi(*token_types)
-        a = []
-
-        while token_types.include?(current_type)
-          a << current_token
-          self.next
-        end
-
-        raise "No token is consumed. #{token_types}" if a.empty?
-
-        return a
-      end
-
-      def eots?
-        current_token.nil?
-      end
-    end
-
     def initialize(text)
       @text = text
     end

data/lib/lrama/state/reduce.rb

@@ -0,0 +1,35 @@
+module Lrama
+  class State
+    class Reduce
+      # https://www.gnu.org/software/bison/manual/html_node/Default-Reductions.html
+      attr_reader :item, :look_ahead, :not_selected_symbols
+      attr_accessor :default_reduction
+
+      def initialize(item)
+        @item = item
+        @look_ahead = nil
+        @not_selected_symbols = []
+      end
+
+      def rule
+        @item.rule
+      end
+
+      def look_ahead=(look_ahead)
+        @look_ahead = look_ahead.freeze
+      end
+
+      def add_not_selected_symbol(sym)
+        @not_selected_symbols << sym
+      end
+
+      def selected_look_ahead
+        if @look_ahead
+          @look_ahead - @not_selected_symbols
+        else
+          []
+        end
+      end
+    end
+  end
+end

data/lib/lrama/state/shift.rb

@@ -0,0 +1,13 @@
+module Lrama
+  class State
+    class Shift
+      attr_reader :next_sym, :next_items
+      attr_accessor :not_selected
+
+      def initialize(next_sym, next_items)
+        @next_sym = next_sym
+        @next_items = next_items
+      end
+    end
+  end
+end

data/lib/lrama/state.rb

@@ -0,0 +1,184 @@
+require "lrama/state/reduce"
+require "lrama/state/shift"
+
+module Lrama
+  class State
+    # * symbol: A symbol under discussion
+    # * reduce: A reduce under discussion
+    # * which: For which a conflict is resolved. :shift, :reduce or :error (for nonassociative)
+    ResolvedConflict = Struct.new(:symbol, :reduce, :which, :same_prec, keyword_init: true) do
+      def report_message
+        s = symbol.display_name
+        r = reduce.rule.precedence_sym.display_name
+        case
+        when which == :shift && same_prec
+          msg = "resolved as #{which} (%right #{s})"
+        when which == :shift
+          msg = "resolved as #{which} (#{r} < #{s})"
+        when which == :reduce && same_prec
+          msg = "resolved as #{which} (%left #{s})"
+        when which == :reduce
+          msg = "resolved as #{which} (#{s} < #{r})"
+        when which == :error
+          msg = "resolved as an #{which} (%nonassoc #{s})"
+        else
+          raise "Unknown direction. #{self}"
+        end
+
+        "Conflict between rule #{reduce.rule.id} and token #{s} #{msg}."
+      end
+    end
+
+    Conflict = Struct.new(:symbols, :reduce, :type, keyword_init: true)
+
+    attr_reader :id, :accessing_symbol, :kernels, :conflicts, :resolved_conflicts,
+                :default_reduction_rule, :closure, :items
+    attr_accessor :shifts, :reduces
+
+    def initialize(id, accessing_symbol, kernels)
+      @id = id
+      @accessing_symbol = accessing_symbol
+      @kernels = kernels.freeze
+      @items = @kernels
+      # Manage relationships between items to state
+      # to resolve next state
+      @items_to_state = {}
+      @conflicts = []
+      @resolved_conflicts = []
+      @default_reduction_rule = nil
+    end
+
+    def closure=(closure)
+      @closure = closure
+      @items = @kernels + @closure
+    end
+
+    def non_default_reduces
+      reduces.select do |reduce|
+        reduce.rule != @default_reduction_rule
+      end
+    end
+
+    def compute_shifts_reduces
+      _shifts = {}
+      reduces = []
+      items.each do |item|
+        # TODO: Consider what should be pushed
+        if item.end_of_rule?
+          reduces << Reduce.new(item)
+        else
+          key = item.next_sym
+          _shifts[key] ||= []
+          _shifts[key] << item.new_by_next_position
+        end
+      end
+
+      # It seems Bison 3.8.2 iterates transitions order by symbol number
+      shifts = _shifts.sort_by do |next_sym, new_items|
+        next_sym.number
+      end.map do |next_sym, new_items|
+        Shift.new(next_sym, new_items.flatten)
+      end
+      self.shifts = shifts.freeze
+      self.reduces = reduces.freeze
+    end
+
+    def set_items_to_state(items, next_state)
+      @items_to_state[items] = next_state
+    end
+
+    #
+    def set_look_ahead(rule, look_ahead)
+      reduce = reduces.find do |r|
+        r.rule == rule
+      end
+
+      reduce.look_ahead = look_ahead
+    end
+
+    # Returns array of [nterm, next_state]
+    def nterm_transitions
+      return @nterm_transitions if @nterm_transitions
+
+      @nterm_transitions = []
+
+      shifts.each do |shift|
+        next if shift.next_sym.term?
+
+        @nterm_transitions << [shift, @items_to_state[shift.next_items]]
+      end
+
+      @nterm_transitions
+    end
+
+    # Returns array of [term, next_state]
+    def term_transitions
+      return @term_transitions if @term_transitions
+
+      @term_transitions = []
+
+      shifts.each do |shift|
+        next if shift.next_sym.nterm?
+
+        @term_transitions << [shift, @items_to_state[shift.next_items]]
+      end
+
+      @term_transitions
+    end
+
+    def selected_term_transitions
+      term_transitions.select do |shift, next_state|
+        !shift.not_selected
+      end
+    end
+
+    # Move to next state by sym
+    def transition(sym)
+      result = nil
+
+      if sym.term?
+        term_transitions.each do |shift, next_state|
+          term = shift.next_sym
+          result = next_state if term == sym
+        end
+      else
+        nterm_transitions.each do |shift, next_state|
+          nterm = shift.next_sym
+          result = next_state if nterm == sym
+        end
+      end
+
+      raise "Can not transit by #{sym} #{self}" if result.nil?
+
+      result
+    end
+
+    def find_reduce_by_item!(item)
+      reduces.find do |r|
+        r.item == item
+      end || (raise "reduce is not found. #{item}")
+    end
+
+    def default_reduction_rule=(default_reduction_rule)
+      @default_reduction_rule = default_reduction_rule
+
+      reduces.each do |r|
+        if r.rule == default_reduction_rule
+          r.default_reduction = true
+        end
+      end
+    end
+
+    def sr_conflicts
+      @conflicts.select do |conflict|
+        conflict.type == :shift_reduce
+      end
+    end
+
+    def rr_conflicts
+      @conflicts.select do |conflict|
+        conflict.type == :reduce_reduce
+      end
+    end
+  end
+end