lrama 0.4.0 → 0.5.1

data/lib/lrama/states.rb

@@ -2,228 +2,6 @@ require "forwardable"
 require "lrama/report"
 
 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
-
-    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
-
-    # * 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}, #{state}")
-    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
-
   # States is passed to a template file
   #
   # "Efficient Computation of LALR(1) Look-Ahead Sets"
@@ -411,23 +189,13 @@ module Lrama
       @states.flat_map(&:rr_conflicts)
     end
 
-    def initial_attrs
-      h = {}
-
-      attrs.each do |attr|
-        h[attr.id] = false
-      end
-
-      h
-    end
-
     def trace_state
       if @trace_state
         yield STDERR
       end
     end
 
-    def create_state(accessing_symbol, kernels, states_creted)
+    def create_state(accessing_symbol, kernels, states_created)
       # A item can appear in some states,
       # so need to use `kernels` (not `kernels.first`) as a key.
       #
@@ -464,11 +232,11 @@ module Lrama
       #    string_1: string •
       #    string_2: string • '+'
       #
-      return [states_creted[kernels], false] if states_creted[kernels]
+      return [states_created[kernels], false] if states_created[kernels]
 
       state = State.new(@states.count, accessing_symbol, kernels)
       @states << state
-      states_creted[kernels] = state
+      states_created[kernels] = state
 
       return [state, true]
     end
@@ -532,9 +300,9 @@ module Lrama
     def compute_lr0_states
       # State queue
       states = []
-      states_creted = {}
+      states_created = {}
 
-      state, _ = create_state(symbols.first, [Item.new(rule: @grammar.rules.first, position: 0)], states_creted)
+      state, _ = create_state(symbols.first, [Item.new(rule: @grammar.rules.first, position: 0)], states_created)
       enqueue_state(states, state)
 
       while (state = states.shift) do
@@ -550,7 +318,7 @@ module Lrama
         setup_state(state)
 
         state.shifts.each do |shift|
-          new_state, created = create_state(shift.next_sym, shift.next_items, states_creted)
+          new_state, created = create_state(shift.next_sym, shift.next_items, states_created)
           state.set_items_to_state(shift.next_items, new_state)
           enqueue_state(states, new_state) if created
         end

data/lib/lrama/states_reporter.rb

@@ -210,7 +210,7 @@ module Lrama
           io << "\n"
 
 
-          # Reprot reads_relation
+          # Report reads_relation
           io << "  [Reads Relation]\n"
           @states.nterms.each do |nterm|
             a = @states.reads_relation[[state.id, nterm.token_id]]
@@ -224,7 +224,7 @@ module Lrama
           io << "\n"
 
 
-          # Reprot read_sets
+          # Report read_sets
           io << "  [Read sets]\n"
           read_sets = @states.read_sets
           @states.nterms.each do |nterm|
@@ -239,7 +239,7 @@ module Lrama
           io << "\n"
 
 
-          # Reprot includes_relation
+          # Report includes_relation
           io << "  [Includes Relation]\n"
           @states.nterms.each do |nterm|
             a = @states.includes_relation[[state.id, nterm.token_id]]
@@ -267,7 +267,7 @@ module Lrama
           io << "\n"
 
 
-          # Reprot follow_sets
+          # Report follow_sets
           io << "  [Follow sets]\n"
           follow_sets = @states.follow_sets
           @states.nterms.each do |nterm|

data/lib/lrama/version.rb

@@ -1,3 +1,3 @@
 module Lrama
-  VERSION = "0.4.0".freeze
+  VERSION = "0.5.1".freeze
 end

data/lib/lrama.rb

@@ -7,6 +7,7 @@ require "lrama/lexer"
 require "lrama/output"
 require "lrama/parser"
 require "lrama/report"
+require "lrama/state"
 require "lrama/states"
 require "lrama/states_reporter"
 require "lrama/version"

data/sample/calc.output

@@ -0,0 +1,263 @@
+Grammar
+
+    0 $accept: list "end of file"
+
+    1 list: ε
+    2     | list LF
+    3     | list expr LF
+
+    4 expr: NUM
+    5     | expr '+' expr
+    6     | expr '-' expr
+    7     | expr '*' expr
+    8     | expr '/' expr
+    9     | '(' expr ')'
+
+
+State 0
+
+    0 $accept: • list "end of file"
+    1 list: ε •
+    2     | • list LF
+    3     | • list expr LF
+
+    $default  reduce using rule 1 (list)
+
+    list  go to state 1
+
+
+State 1
+
+    0 $accept: list • "end of file"
+    2 list: list • LF
+    3     | list • expr LF
+    4 expr: • NUM
+    5     | • expr '+' expr
+    6     | • expr '-' expr
+    7     | • expr '*' expr
+    8     | • expr '/' expr
+    9     | • '(' expr ')'
+
+    "end of file"  shift, and go to state 2
+    LF             shift, and go to state 3
+    NUM            shift, and go to state 4
+    '('            shift, and go to state 5
+
+    expr  go to state 6
+
+
+State 2
+
+    0 $accept: list "end of file" •
+
+    $default  accept
+
+
+State 3
+
+    2 list: list LF •
+
+    $default  reduce using rule 2 (list)
+
+
+State 4
+
+    4 expr: NUM •
+
+    $default  reduce using rule 4 (expr)
+
+
+State 5
+
+    4 expr: • NUM
+    5     | • expr '+' expr
+    6     | • expr '-' expr
+    7     | • expr '*' expr
+    8     | • expr '/' expr
+    9     | • '(' expr ')'
+    9     | '(' • expr ')'
+
+    NUM  shift, and go to state 4
+    '('  shift, and go to state 5
+
+    expr  go to state 7
+
+
+State 6
+
+    3 list: list expr • LF
+    5 expr: expr • '+' expr
+    6     | expr • '-' expr
+    7     | expr • '*' expr
+    8     | expr • '/' expr
+
+    LF   shift, and go to state 8
+    '+'  shift, and go to state 9
+    '-'  shift, and go to state 10
+    '*'  shift, and go to state 11
+    '/'  shift, and go to state 12
+
+
+State 7
+
+    5 expr: expr • '+' expr
+    6     | expr • '-' expr
+    7     | expr • '*' expr
+    8     | expr • '/' expr
+    9     | '(' expr • ')'
+
+    '+'  shift, and go to state 9
+    '-'  shift, and go to state 10
+    '*'  shift, and go to state 11
+    '/'  shift, and go to state 12
+    ')'  shift, and go to state 13
+
+
+State 8
+
+    3 list: list expr LF •
+
+    $default  reduce using rule 3 (list)
+
+
+State 9
+
+    4 expr: • NUM
+    5     | • expr '+' expr
+    5     | expr '+' • expr
+    6     | • expr '-' expr
+    7     | • expr '*' expr
+    8     | • expr '/' expr
+    9     | • '(' expr ')'
+
+    NUM  shift, and go to state 4
+    '('  shift, and go to state 5
+
+    expr  go to state 14
+
+
+State 10
+
+    4 expr: • NUM
+    5     | • expr '+' expr
+    6     | • expr '-' expr
+    6     | expr '-' • expr
+    7     | • expr '*' expr
+    8     | • expr '/' expr
+    9     | • '(' expr ')'
+
+    NUM  shift, and go to state 4
+    '('  shift, and go to state 5
+
+    expr  go to state 15
+
+
+State 11
+
+    4 expr: • NUM
+    5     | • expr '+' expr
+    6     | • expr '-' expr
+    7     | • expr '*' expr
+    7     | expr '*' • expr
+    8     | • expr '/' expr
+    9     | • '(' expr ')'
+
+    NUM  shift, and go to state 4
+    '('  shift, and go to state 5
+
+    expr  go to state 16
+
+
+State 12
+
+    4 expr: • NUM
+    5     | • expr '+' expr
+    6     | • expr '-' expr
+    7     | • expr '*' expr
+    8     | • expr '/' expr
+    8     | expr '/' • expr
+    9     | • '(' expr ')'
+
+    NUM  shift, and go to state 4
+    '('  shift, and go to state 5
+
+    expr  go to state 17
+
+
+State 13
+
+    9 expr: '(' expr ')' •
+
+    $default  reduce using rule 9 (expr)
+
+
+State 14
+
+    5 expr: expr • '+' expr
+    5     | expr '+' expr •  [LF, '+', '-', ')']
+    6     | expr • '-' expr
+    7     | expr • '*' expr
+    8     | expr • '/' expr
+
+    '*'  shift, and go to state 11
+    '/'  shift, and go to state 12
+
+    $default  reduce using rule 5 (expr)
+
+    Conflict between rule 5 and token '+' resolved as reduce (%left '+').
+    Conflict between rule 5 and token '-' resolved as reduce (%left '-').
+    Conflict between rule 5 and token '*' resolved as shift ('+' < '*').
+    Conflict between rule 5 and token '/' resolved as shift ('+' < '/').
+
+
+State 15
+
+    5 expr: expr • '+' expr
+    6     | expr • '-' expr
+    6     | expr '-' expr •  [LF, '+', '-', ')']
+    7     | expr • '*' expr
+    8     | expr • '/' expr
+
+    '*'  shift, and go to state 11
+    '/'  shift, and go to state 12
+
+    $default  reduce using rule 6 (expr)
+
+    Conflict between rule 6 and token '+' resolved as reduce (%left '+').
+    Conflict between rule 6 and token '-' resolved as reduce (%left '-').
+    Conflict between rule 6 and token '*' resolved as shift ('-' < '*').
+    Conflict between rule 6 and token '/' resolved as shift ('-' < '/').
+
+
+State 16
+
+    5 expr: expr • '+' expr
+    6     | expr • '-' expr
+    7     | expr • '*' expr
+    7     | expr '*' expr •  [LF, '+', '-', '*', '/', ')']
+    8     | expr • '/' expr
+
+    $default  reduce using rule 7 (expr)
+
+    Conflict between rule 7 and token '+' resolved as reduce ('+' < '*').
+    Conflict between rule 7 and token '-' resolved as reduce ('-' < '*').
+    Conflict between rule 7 and token '*' resolved as reduce (%left '*').
+    Conflict between rule 7 and token '/' resolved as reduce (%left '/').
+
+
+State 17
+
+    5 expr: expr • '+' expr
+    6     | expr • '-' expr
+    7     | expr • '*' expr
+    8     | expr • '/' expr
+    8     | expr '/' expr •  [LF, '+', '-', '*', '/', ')']
+
+    $default  reduce using rule 8 (expr)
+
+    Conflict between rule 8 and token '+' resolved as reduce ('+' < '/').
+    Conflict between rule 8 and token '-' resolved as reduce ('-' < '/').
+    Conflict between rule 8 and token '*' resolved as reduce (%left '*').
+    Conflict between rule 8 and token '/' resolved as reduce (%left '/').
+
+

data/sample/calc.y

@@ -0,0 +1,98 @@
+/*
+ * How to build and run:
+ *
+ * $ lrama -d calc.y -o calc.c && gcc -Wall calc.c -o calc && ./calc
+ * 1
+ * => 1
+ * 1+2*3
+ * => 7
+ * (1+2)*3
+ * => 9
+ *
+ */
+
+%{
+#include <stdio.h>
+#include <stdlib.h>
+#include <ctype.h>
+
+#include "calc.h"
+
+static int yylex(YYSTYPE *val, YYLTYPE *loc);
+static int yyerror(YYLTYPE *loc, const char *str);
+%}
+
+%union {
+    int val;
+}
+%token LF
+%token <val> NUM
+%type <val> expr
+%left '+' '-'
+%left '*' '/'
+
+%%
+
+list : /* empty */
+     | list LF
+     | list expr LF { printf("=> %d\n", $2); }
+     ;
+expr : NUM
+     | expr '+' expr { $$ = $1 + $3; }
+     | expr '-' expr { $$ = $1 - $3; }
+     | expr '*' expr { $$ = $1 * $3; }
+     | expr '/' expr { $$ = $1 / $3; }
+     | '(' expr ')'  { $$ = $2; }
+     ;
+
+%%
+
+static int yylex(YYSTYPE *yylval, YYLTYPE *loc) {
+    int c = getchar();
+    int val;
+
+    switch (c) {
+    case ' ': case '\t':
+        return yylex(yylval, loc);
+
+    case '0': case '1': case '2': case '3': case '4':
+    case '5': case '6': case '7': case '8': case '9':
+        val = c - '0';
+        while (1) {
+            c = getchar();
+            if (isdigit(c)) {
+                val = val * 10 + (c - '0');
+            }
+            else {
+                ungetc(c, stdin);
+                break;
+            }
+        }
+        yylval->val = val;
+        return NUM;
+
+    case '\n':
+        return LF;
+
+    case '+': case '-': case '*': case '/': case '(': case ')':
+        return c;
+
+    case EOF:
+        exit(0);
+
+    default:
+        fprintf(stderr, "unknown character: %c\n", c);
+        exit(1);
+    }
+}
+
+static int yyerror(YYLTYPE *loc, const char *str) {
+    fprintf(stderr, "parse error: %s\n", str);
+    return 0;
+}
+
+int main() {
+    printf("Enter the formula:\n");
+    yyparse();
+    return 0;
+}

data/sig/lrama/bitmap.rbs

@@ -0,0 +1,7 @@
+module Lrama
+  module Bitmap
+    def self.from_array: (Array[int] ary) -> Integer
+
+    def self.to_array: (Integer int) -> Array[Integer]
+  end
+end

data/template/bison/yacc.c

@@ -1,5 +1,5 @@
 <%# b4_generated_by -%>
-/* A Bison parser, made by GNU Bison 3.8.2.  */
+/* A Bison parser, made by Lrama <%= Lrama::VERSION %>.  */
 
 <%# b4_copyright -%>
 /* Bison implementation for Yacc-like parsers in C