Document p_lex and p_token_info_t in user guide - fix #37

This issue manifested when multiple ptypes are present with different sizes.

.github/workflows/run-tests.yml

@@ -0,0 +1,38 @@
+name: Run Propane Tests
+
+on:
+  push:
+    branches:
+    - master
+  pull_request:
+
+jobs:
+  test:
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ubuntu-latest, macos-latest]
+        ruby-version: ['3.4']
+
+    steps:
+    - name: Install dependencies (Linux)
+      if: runner.os == 'Linux'
+      run: sudo apt-get update && sudo apt-get install -y gcc gdc ldc
+
+    - name: Install dependencies (macOS)
+      if: runner.os == 'macOS'
+      run: brew install gcc ldc
+
+    - name: Checkout repository
+      uses: actions/checkout@v4
+
+    - name: Set up Ruby
+      uses: ruby/setup-ruby@v1
+      with:
+        ruby-version: ${{ matrix.ruby-version }}
+
+    - name: Install dependencies
+      run: bundle install
+
+    - name: Run tests
+      run: rake all

CHANGELOG.md

@@ -1,3 +1,28 @@
+## v2.3.0
+
+### New Features
+
+- Add \D, \S, \w, \W special character classes
+
+### Improvements
+
+- Include line numbers for pattern errors
+- Improve performance in a few places
+- Parallelize parser table generation on Linux hosts
+- Add github workflow to run unit tests
+
+### Fixes
+
+- Fix a couple clang warnings for C backend
+- Fix C backend not fully initializing pvalues when multiple ptypes are used with different sizes.
+- Fix some user guide examples
+
+## v2.2.1
+
+### Fixes
+
+- Fix GC issue for D backend when AST is enabled (#36)
+
 ## v2.2.0
 
 ### Improvements

Gemfile.lock

@@ -1,37 +1,39 @@
 GEM
   remote: https://rubygems.org/
   specs:
-    base64 (0.2.0)
+    base64 (0.3.0)
     date (3.4.1)
-    diff-lcs (1.6.1)
+    diff-lcs (1.6.2)
     docile (1.4.1)
-    psych (5.2.3)
+    erb (5.0.2)
+    psych (5.2.6)
       date
       stringio
-    rake (13.2.1)
-    rdoc (6.13.1)
+    rake (13.3.0)
+    rdoc (6.14.2)
+      erb
       psych (>= 4.0.0)
     redcarpet (3.6.1)
-    rspec (3.13.0)
+    rspec (3.13.1)
       rspec-core (~> 3.13.0)
       rspec-expectations (~> 3.13.0)
       rspec-mocks (~> 3.13.0)
-    rspec-core (3.13.3)
+    rspec-core (3.13.5)
       rspec-support (~> 3.13.0)
-    rspec-expectations (3.13.3)
+    rspec-expectations (3.13.5)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.13.0)
-    rspec-mocks (3.13.2)
+    rspec-mocks (3.13.5)
       diff-lcs (>= 1.2.0, < 2.0)
       rspec-support (~> 3.13.0)
-    rspec-support (3.13.2)
+    rspec-support (3.13.4)
     simplecov (0.22.0)
       docile (~> 1.1)
       simplecov-html (~> 0.11)
       simplecov_json_formatter (~> 0.1)
-    simplecov-html (0.13.1)
+    simplecov-html (0.13.2)
     simplecov_json_formatter (0.1.4)
-    stringio (3.1.6)
+    stringio (3.1.7)
     syntax (1.2.2)
 
 PLATFORMS

assets/parser.c.erb

@@ -55,7 +55,8 @@ const char * <%= @grammar.prefix %>token_names[] = {
 void <%= @grammar.prefix %>context_init(<%= @grammar.prefix %>context_t * context, uint8_t const * input, size_t input_length)
 {
     /* New default-initialized context structure. */
-    <%= @grammar.prefix %>context_t newcontext = {0};
+    <%= @grammar.prefix %>context_t newcontext;
+    memset(&newcontext, 0, sizeof(newcontext));
 
     /* Lexer initialization. */
     newcontext.input = input;
@@ -344,8 +345,10 @@ static lexer_state_id_t check_lexer_transition(uint32_t current_state, uint32_t
 static size_t find_longest_match(<%= @grammar.prefix %>context_t * context,
     lexer_match_info_t * out_match_info, size_t * out_unexpected_input_length)
 {
-    lexer_match_info_t longest_match = {0};
-    lexer_match_info_t attempt_match = {0};
+    lexer_match_info_t longest_match;
+    memset(&longest_match, 0, sizeof(longest_match));
+    lexer_match_info_t attempt_match;
+    memset(&attempt_match, 0, sizeof(attempt_match));
     *out_match_info = longest_match;
     uint32_t current_state = lexer_mode_table[context->mode].state_table_offset;
     for (;;)
@@ -359,6 +362,7 @@ static size_t find_longest_match(<%= @grammar.prefix %>context_t * context,
         switch (result)
         {
         case P_SUCCESS:
+            {
             lexer_state_id_t transition_state = check_lexer_transition(current_state, code_point);
             if (transition_state != INVALID_LEXER_STATE_ID)
             {
@@ -390,6 +394,7 @@ static size_t find_longest_match(<%= @grammar.prefix %>context_t * context,
                 *out_unexpected_input_length = attempt_match.length + code_point_length;
                 return P_UNEXPECTED_INPUT;
             }
+            }
             break;
 
         case P_EOF:
@@ -447,7 +452,8 @@ static size_t find_longest_match(<%= @grammar.prefix %>context_t * context,
  */
 static size_t attempt_lex_token(<%= @grammar.prefix %>context_t * context, <%= @grammar.prefix %>token_info_t * out_token_info)
 {
-    <%= @grammar.prefix %>token_info_t token_info = {0};
+    <%= @grammar.prefix %>token_info_t token_info;
+    memset(&token_info, 0, sizeof(token_info));
     token_info.position = context->text_position;
     token_info.token = INVALID_TOKEN_ID;
     lexer_match_info_t match_info;
@@ -456,6 +462,7 @@ static size_t attempt_lex_token(<%= @grammar.prefix %>context_t * context, <%= @
     switch (result)
     {
     case P_SUCCESS:
+        {
         <%= @grammar.prefix %>token_t token_to_accept = match_info.accepting_state->token;
         if (match_info.accepting_state->code_id != INVALID_USER_CODE_ID)
         {
@@ -507,6 +514,7 @@ static size_t attempt_lex_token(<%= @grammar.prefix %>context_t * context, <%= @
             token_info.end_position.col = token_info.position.col + match_info.end_delta_position.col;
         }
         *out_token_info = token_info;
+        }
         return P_SUCCESS;
 
     case P_EOF:
@@ -999,7 +1007,8 @@ size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * context)
                 state_values_stack_index(&statevalues, -1)->ast_node = reduced_parser_node;
 <% else %>
                 state_values_stack_index(&statevalues, -1)->pvalue = reduced_parser_value;
-                <%= @grammar.prefix %>value_t new_parse_result = {0};
+                <%= @grammar.prefix %>value_t new_parse_result;
+                memset(&new_parse_result, 0, sizeof(new_parse_result));
                 reduced_parser_value = new_parse_result;
 <% end %>
                 reduced_rule_set = INVALID_ID;
@@ -1061,7 +1070,8 @@ size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * context)
                 reduced_parser_node = NULL;
             }
 <% else %>
-            <%= @grammar.prefix %>value_t reduced_parser_value2 = {0};
+            <%= @grammar.prefix %>value_t reduced_parser_value2;
+            memset(&reduced_parser_value2, 0, sizeof(reduced_parser_value2));
             if (parser_user_code(&reduced_parser_value2, parser_reduce_table[reduce_index].rule, &statevalues, parser_reduce_table[reduce_index].n_states, context) == P_USER_TERMINATED)
             {
                 return P_USER_TERMINATED;

doc/user_guide.md

@@ -67,10 +67,10 @@ import std.math;
 ptype ulong;
 
 # A few basic arithmetic operators.
-token plus /\\+/;
-token times /\\*/;
-token power /\\*\\*/;
-token integer /\\d+/ <<
+token plus /\+/;
+token times /\*/;
+token power /\*\*/;
+token integer /\d+/ <<
   ulong v;
   foreach (c; match)
   {
@@ -79,10 +79,10 @@ token integer /\\d+/ <<
   }
   $$ = v;
 >>
-token lparen /\\(/;
-token rparen /\\)/;
+token lparen /\(/;
+token rparen /\)/;
 # Drop whitespace.
-drop /\\s+/;
+drop /\s+/;
 
 Start -> E1 << $$ = $1; >>
 E1 -> E2 << $$ = $1; >>
@@ -155,7 +155,7 @@ Example:
 ```
 ptype ulong;
 
-token integer /\\d+/ <<
+token integer /\d+/ <<
   ulong v;
   foreach (c; match)
   {
@@ -228,9 +228,9 @@ token two /2/;
 token comma /,/ <<
   $$ = 42;
 >>
-token lparen /\\(/;
-token rparen /\\)/;
-drop /\\s+/;
+token lparen /\(/;
+token rparen /\)/;
+drop /\s+/;
 
 Start -> Items;
 
@@ -432,10 +432,10 @@ A regular expression begins and ends with a `/` character.
 Example:
 
 ```
-/#.*$/
+/#.*/
 ```
 
-Regular expressions can include many special characters:
+Regular expressions can include many special characters/sequences:
 
   * The `.` character matches any input character other than a newline.
   * The `*` character matches any number of the previous regex element.
@@ -447,14 +447,17 @@ Regular expressions can include many special characters:
   * The `\` character escapes the following character and changes its meaning:
     * The `\a` sequence matches an ASCII bell character (0x07).
     * The `\b` sequence matches an ASCII backspace character (0x08).
-    * The `\d` sequence matches any character `0` through `9`.
+    * The `\d` sequence is shorthand for the `[0-9]` character class.
+    * The `\D` sequence matches every code point not matched by `\d`.
     * The `\f` sequence matches an ASCII form feed character (0x0C).
     * The `\n` sequence matches an ASCII new line character (0x0A).
     * The `\r` sequence matches an ASCII carriage return character (0x0D).
-    * The `\s` sequence matches a space, horizontal tab `\t`, carriage return
-    `\r`, a form feed `\f`, or a vertical tab `\v` character.
+    * The `\s` sequence is shorthand for the `[ \t\r\n\f\v]` character class.
+    * The `\S` sequence matches every code point not matched by `\s`.
     * The `\t` sequence matches an ASCII tab character (0x09).
     * The `\v` sequence matches an ASCII vertical tab character (0x0B).
+    * The `\w` sequence is shorthand for the `[a-zA-Z0-9_]` character class.
+    * The `\W` sequence matches every code point not matched by `\w`.
     * Any other character matches itself.
   * The `|` character creates an alternate match.
 
@@ -552,10 +555,10 @@ default, identonly: token ident /[a-z]+/ <<
   $mode(default);
   return $token(ident);
 >>
-token dot /\\./ <<
+token dot /\./ <<
   $mode(identonly);
 >>
-default, identonly: drop /\\s+/;
+default, identonly: drop /\s+/;
 ```
 
 ##> Specifying parser value types - the `ptype` statement
@@ -748,7 +751,7 @@ Some example uses of this functionality could be to:
 
   * Detect integer overflow when lexing an integer literal constant.
   * Detect and report an error as soon as possible during parsing before continuing to parse any more of the input.
-  * Determine whether parsing should stop and instead be performed using a different parser version.
+  * Determine whether parsing should stop and instead be retried using a different parser version.
 
 To terminate parsing from a lexer or parser user code block, use the
 `$terminate(code)` function, passing an integer expression argument.
@@ -784,7 +787,7 @@ Propane generates the following result code constants:
 * `P_EOF`: The lexer reached the end of the input string.
 * `P_USER_TERMINATED`: A parser user code block has requested to terminate the parser.
 
-Result codes are returned by the functions `p_decode_code_point()`, `p_lex()`, and `p_parse()`.
+Result codes are returned by the API functions `p_decode_code_point()`, `p_lex()`, and `p_parse()`.
 
 ##> Types
 
@@ -804,7 +807,7 @@ A pointer to this instance is passed to the generated functions.
 
 ### `p_position_t`
 
-The `p_position_t` structure contains two fields `row` and `col`.
+The `p_position_t` structure contains two fields: `row` and `col`.
 These fields contain the 1-based row and column describing a parser position.
 
 For D targets, the `p_position_t` structure can be checked for validity by
@@ -814,6 +817,16 @@ For C targets, the `p_position_t` structure can be checked for validity by
 calling `p_position_valid(pos)` where `pos` is a `p_position_t` structure
 instance.
 
+### `p_token_info_t`
+
+The `p_token_info_t` structure contains the following fields:
+
+* `position` (`p_position_t`) holds the text position of the first code point in the token.
+* `end_position` (`p_position_t`) holds the text position of the last code point in the token.
+* `length` (`size_t`) holds the number of input bytes used by the token.
+* `token` (`p_token_t`) holds the token ID of the lexed token
+* `pvalue` (`p_value_t`) holds the parser value associated with the token.
+
 ### AST Node Types
 
 If AST generation mode is enabled, a structure type for each rule will be
@@ -924,6 +937,44 @@ p_context_t context;
 p_context_init(&context, input);
 ```
 
+### `p_lex`
+
+The `p_lex()` function is the main entry point to the lexer.
+It is normally called automatically by the generated parser to retrieve the
+next input token for the parser and does not need to be called by the user.
+However, the user may initialize a context and call `p_lex()` to use the
+generated lexer in a standalone mode.
+
+Example:
+
+```
+p_context_t context;
+p_context_init(&context, input, input_length);
+p_token_info_t token_info;
+size_t result = p_lex(&context, &token_info);
+switch (result)
+{
+case P_DECODE_ERROR:
+    /* UTF-8 decode error */
+    break;
+case P_UNEXPECTED_INPUT:
+    /* Input text does not match any lexer pattern. */
+    break;
+case P_USER_TERMINATED:
+    /* Lexer user code block requested to terminate the lexer. */
+    break;
+case P_SUCCESS:
+    /*
+     * token_info.position holds the text position of the first code point in the token.
+     * token_info.end_position holds the text position of the last code point in the token.
+     * token_info.length holds the number of input bytes used by the token.
+     * token_info.token holds the token ID of the lexed token
+     * token_info.pvalue holds the parser value associated with the token.
+     */
+    break;
+}
+```
+
 ### `p_parse`
 
 The `p_parse()` function is the main entry point to the parser.

extra/vim/syntax/propane.vim

@@ -22,7 +22,7 @@ syn match propaneFieldOperator ":" contained
 syn match propaneOperator "?"
 syn keyword propaneKeyword ast ast_prefix ast_suffix drop module prefix ptype start token tokenid
 
-syn region propaneRegex start="/" end="/" skip="\\/"
+syn region propaneRegex start="/" end="/" skip="\v\\\\|\\/"
 
 hi def link propaneComment Comment
 hi def link propaneKeyword Keyword

lib/propane/parser.rb

@@ -64,27 +64,16 @@ class Propane
       @shift_table = []
       @reduce_table = []
       @item_sets.each do |item_set|
-        shift_entries = item_set.next_symbols.map do |next_symbol|
-          state_id =
-            if next_symbol.name == "$EOF"
-              0
-            else
-              item_set.next_item_set[next_symbol].id
-            end
-          {
-            symbol: next_symbol,
-            state_id: state_id,
-          }
-        end
         unless item_set.reduce_rules.empty?
-          shift_entries.each do |shift_entry|
+          item_set.shift_entries.each do |shift_entry|
             token = shift_entry[:symbol]
-            if get_lookahead_reduce_actions_for_item_set(item_set).include?(token)
-              rule = item_set.reduce_actions[token]
+            if item_set.reduce_actions
+              if rule = item_set.reduce_actions[token]
                 @warnings << "Shift/Reduce conflict (state #{item_set.id}) between token #{token.name} and rule #{rule.name} (defined on line #{rule.line_number})"
               end
             end
           end
+        end
         reduce_entries =
           if rule = item_set.reduce_rule
             [{token_id: @grammar.invalid_token_id, rule_id: rule.id, rule: rule,
@@ -101,11 +90,11 @@ class Propane
           end
         @state_table << {
           shift_index: @shift_table.size,
-          n_shifts: shift_entries.size,
+          n_shifts: item_set.shift_entries.size,
           reduce_index: @reduce_table.size,
           n_reduces: reduce_entries.size,
         }
-        @shift_table += shift_entries
+        @shift_table += item_set.shift_entries
         @reduce_table += reduce_entries
       end
     end
@@ -125,8 +114,110 @@ class Propane
     # @return [void]
     def build_reduce_actions!
       @item_sets.each do |item_set|
+        build_shift_entries(item_set)
         build_reduce_actions_for_item_set(item_set)
       end
+      item_sets_to_process = @item_sets.select do |item_set|
+        # We need lookahead reduce actions if:
+        # 1) There is more than one possible rule to reduce. In this case the
+        #    lookahead token can help choose which rule to reduce.
+        # 2) There is at least one shift action and one reduce action for
+        #    this item set. In this case the lookahead reduce actions are
+        #    needed to test for a Shift/Reduce conflict.
+        item_set.reduce_rules.size > 1 ||
+          (item_set.reduce_rules.size > 0 && item_set.shift_entries.size > 0)
+      end
+      if RbConfig::CONFIG["host_os"] =~ /linux/
+        item_sets_by_id = {}
+        item_sets_to_process.each do |item_set|
+          item_sets_by_id[item_set.object_id] = item_set
+        end
+        tokens_by_id = {}
+        @grammar.tokens.each do |token|
+          tokens_by_id[token.object_id] = token
+        end
+        rules_by_id = {}
+        @grammar.rules.each do |rule|
+          rules_by_id[rule.object_id] = rule
+        end
+        n_threads = Util.determine_n_threads
+        semaphore = Mutex.new
+        queue = Queue.new
+        threads = {}
+        n_threads.times do
+          piper, pipew = IO.pipe
+          thread = Thread.new do
+            loop do
+              item_set = nil
+              semaphore.synchronize do
+                item_set = item_sets_to_process.slice!(0)
+              end
+              break if item_set.nil?
+              fork do
+                piper.close
+                build_lookahead_reduce_actions_for_item_set(item_set, pipew)
+              end
+            end
+            queue.push(Thread.current)
+          end
+          threads[thread] = [piper, pipew]
+        end
+        until threads.empty?
+          thread = queue.pop
+          piper, pipew = threads[thread]
+          pipew.close
+          thread_txt = piper.read
+          thread_txt.each_line do |line|
+            if line.start_with?("RA,")
+              parts = line.split(",")
+              item_set_id, token_id, rule_id = parts[1..3].map(&:to_i)
+              item_set = item_sets_by_id[item_set_id]
+              unless item_set
+                raise "Internal error: could not find item set from thread"
+              end
+              token = tokens_by_id[token_id]
+              unless item_set
+                raise "Internal error: could not find token from thread"
+              end
+              rule = rules_by_id[rule_id]
+              unless item_set
+                raise "Internal error: could not find rule from thread"
+              end
+              item_set.reduce_actions ||= {}
+              item_set.reduce_actions[token] = rule
+            elsif line.start_with?("Error: ")
+              @errors << line.chomp
+            else
+              raise "Internal error: unhandled thread line #{line}"
+            end
+          end
+          thread.join
+          threads.delete(thread)
+        end
+      else
+        # Fall back to single threaded algorithm.
+        item_sets_to_process.each do |item_set|
+          item_set.reduce_actions = build_lookahead_reduce_actions_for_item_set(item_set)
+        end
+      end
+    end
+
+    # Build the shift entries for a single item set.
+    #
+    # @return [void]
+    def build_shift_entries(item_set)
+      item_set.shift_entries = item_set.next_symbols.map do |next_symbol|
+        state_id =
+          if next_symbol.name == "$EOF"
+            0
+          else
+            item_set.next_item_set[next_symbol].id
+          end
+        {
+          symbol: next_symbol,
+          state_id: state_id,
+        }
+      end
     end
 
     # Build the reduce actions for a single item set (parser state).
@@ -145,33 +236,18 @@ class Propane
       if item_set.reduce_rules.size == 1
         item_set.reduce_rule = item_set.reduce_rules.first
       end
-
-      if item_set.reduce_rules.size > 1
-        # Force item_set.reduce_actions to be built to store the lookahead
-        # tokens for the possible reduce rules if there is more than one.
-        get_lookahead_reduce_actions_for_item_set(item_set)
-      end
-    end
-
-    # Get the reduce actions for a single item set (parser state).
-    #
-    # @param item_set [ItemSet]
-    #   ItemSet (parser state)
-    #
-    # @return [Hash]
-    #   Mapping of lookahead Tokens to the Rules to reduce.
-    def get_lookahead_reduce_actions_for_item_set(item_set)
-      item_set.reduce_actions ||= build_lookahead_reduce_actions_for_item_set(item_set)
     end
 
     # Build the reduce actions for a single item set (parser state).
     #
     # @param item_set [ItemSet]
     #   ItemSet (parser state)
+    # @param fh [File]
+    #   Output file handle for multiprocessing mode.
     #
     # @return [Hash]
     #   Mapping of lookahead Tokens to the Rules to reduce.
-    def build_lookahead_reduce_actions_for_item_set(item_set)
+    def build_lookahead_reduce_actions_for_item_set(item_set, fh = nil)
       # We will be looking for all possible tokens that can follow instances of
       # these rules. Rather than looking through the entire grammar for the
       # possible following tokens, we will only look in the item sets leading
@@ -182,9 +258,12 @@ class Propane
         lookahead_tokens_for_rule = build_lookahead_tokens_to_reduce(reduce_rule, item_sets)
         lookahead_tokens_for_rule.each do |lookahead_token|
           if existing_reduce_rule = reduce_actions[lookahead_token]
-            @errors << "Error: reduce/reduce conflict (state #{item_set.id}) between rule #{existing_reduce_rule.name}##{existing_reduce_rule.id} (defined on line #{existing_reduce_rule.line_number}) and rule #{reduce_rule.name}##{reduce_rule.id} (defined on line #{reduce_rule.line_number}) for lookahead token #{lookahead_token}"
+            error = "Error: reduce/reduce conflict (state #{item_set.id}) between rule #{existing_reduce_rule.name}##{existing_reduce_rule.id} (defined on line #{existing_reduce_rule.line_number}) and rule #{reduce_rule.name}##{reduce_rule.id} (defined on line #{reduce_rule.line_number}) for lookahead token #{lookahead_token}"
+            @errors << error
+            fh.puts(error) if fh
           end
           reduce_actions[lookahead_token] = reduce_rule
+          fh.puts "RA,#{item_set.object_id},#{lookahead_token.object_id},#{reduce_rule.object_id}" if fh
         end
         reduce_actions
       end
@@ -231,6 +310,7 @@ class Propane
                   rule_set = item.rule.rule_set
                   unless checked_rule_sets.include?(rule_set)
                     rule_sets_to_check_after << rule_set
+                    checked_rule_sets << rule_set
                   end
                   break
                 when Token

lib/propane/parser/item.rb

@@ -22,6 +22,7 @@ class Propane
       def initialize(rule, position)
         @rule = rule
         @position = position
+        @_hash = [@rule, @position].hash
       end
 
       # Hash function.
@@ -29,7 +30,7 @@ class Propane
       # @return [Integer]
       #   Hash code.
       def hash
-        [@rule, @position].hash
+        @_hash
       end
 
       # Compare Item objects.

lib/propane/parser/item_set.rb

@@ -34,6 +34,10 @@ class Propane
       #   more than one rule that could be reduced.
       attr_accessor :reduce_actions
 
+      # @return [Array<Hash>]
+      #   Shift table entries.
+      attr_accessor :shift_entries
+
       # Build an ItemSet.
       #
       # @param items [Array<Item>]

lib/propane/pattern.rb

@@ -55,7 +55,7 @@ class Propane
       @line_number = options[:line_number]
       @modes = options[:modes]
       @ptypename = options[:ptypename]
-      regex = Regex.new(@pattern)
+      regex = Regex.new(@pattern, @line_number)
       regex.nfa.end_state.accepts = self
       @nfa = regex.nfa
     end

lib/propane/regex.rb

@@ -4,12 +4,13 @@ class Propane
     attr_reader :unit
     attr_reader :nfa
 
-    def initialize(pattern)
+    def initialize(pattern, line_number)
       @pattern = pattern.dup
+      @line_number = line_number
       @unit = parse_alternates
       @nfa = @unit.to_nfa
       if @pattern != ""
-        raise Error.new(%[Unexpected "#{@pattern}" in pattern])
+        raise Error.new(%[Line #{@line_number}: unexpected "#{@pattern}" in pattern])
       end
     end
 
@@ -41,7 +42,7 @@ class Propane
             mu = MultiplicityUnit.new(last_unit, min_count, max_count)
             au.replace_last!(mu)
           else
-            raise Error.new("#{c} follows nothing")
+            raise Error.new("Line #{@line_number}: #{c} follows nothing")
           end
         when "|"
           au.new_alternate!
@@ -59,7 +60,7 @@ class Propane
     def parse_group
       au = parse_alternates
       if @pattern[0] != ")"
-        raise Error.new("Unterminated group in pattern")
+        raise Error.new("Line #{@line_number}: unterminated group in pattern")
       end
       @pattern.slice!(0)
       au
@@ -70,7 +71,7 @@ class Propane
       index = 0
       loop do
         if @pattern == ""
-          raise Error.new("Unterminated character class")
+          raise Error.new("Line #{@line_number}: unterminated character class")
         end
         c = @pattern.slice!(0)
         if c == "]"
@@ -84,13 +85,13 @@ class Propane
         elsif c == "-" && @pattern[0] != "]"
           begin_cu = ccu.last_unit
           unless begin_cu.is_a?(CharacterRangeUnit) && begin_cu.code_point_range.size == 1
-            raise Error.new("Character range must be between single characters")
+            raise Error.new("Line #{@line_number}: character range must be between single characters")
           end
           if @pattern[0] == "\\"
             @pattern.slice!(0)
             end_cu = parse_backslash
             unless end_cu.is_a?(CharacterRangeUnit) && end_cu.code_point_range.size == 1
-              raise Error.new("Character range must be between single characters")
+              raise Error.new("Line #{@line_number}: character range must be between single characters")
             end
             max_code_point = end_cu.code_point
           else
@@ -116,7 +117,7 @@ class Propane
         elsif max_count.to_s != ""
           max_count = max_count.to_i
           if max_count < min_count
-            raise Error.new("Maximum repetition count cannot be less than minimum repetition count")
+            raise Error.new("Line #{@line_number}: maximum repetition count cannot be less than minimum repetition count")
           end
         else
           max_count = nil
@@ -124,28 +125,33 @@ class Propane
         @pattern = pattern
         [min_count, max_count]
       else
-        raise Error.new("Unexpected match count at #{@pattern}")
+        raise Error.new("Line #{@line_number}: unexpected match count following {")
       end
     end
 
     def parse_backslash
       if @pattern == ""
-        raise Error.new("Error: unfollowed \\")
+        raise Error.new("Line #{@line_number}: error: unfollowed \\")
       else
         c = @pattern.slice!(0)
         case c
         when "a"
-          CharacterRangeUnit.new("\a", "\a")
+          CharacterRangeUnit.new("\a")
         when "b"
-          CharacterRangeUnit.new("\b", "\b")
+          CharacterRangeUnit.new("\b")
         when "d"
           CharacterRangeUnit.new("0", "9")
+        when "D"
+          ccu = CharacterClassUnit.new
+          ccu << CharacterRangeUnit.new("0", "9")
+          ccu.negate = true
+          ccu
         when "f"
-          CharacterRangeUnit.new("\f", "\f")
+          CharacterRangeUnit.new("\f")
         when "n"
-          CharacterRangeUnit.new("\n", "\n")
+          CharacterRangeUnit.new("\n")
         when "r"
-          CharacterRangeUnit.new("\r", "\r")
+          CharacterRangeUnit.new("\r")
         when "s"
           ccu = CharacterClassUnit.new
           ccu << CharacterRangeUnit.new(" ")
@@ -155,10 +161,35 @@ class Propane
           ccu << CharacterRangeUnit.new("\f")
           ccu << CharacterRangeUnit.new("\v")
           ccu
+        when "S"
+          ccu = CharacterClassUnit.new
+          ccu << CharacterRangeUnit.new(" ")
+          ccu << CharacterRangeUnit.new("\t")
+          ccu << CharacterRangeUnit.new("\r")
+          ccu << CharacterRangeUnit.new("\n")
+          ccu << CharacterRangeUnit.new("\f")
+          ccu << CharacterRangeUnit.new("\v")
+          ccu.negate = true
+          ccu
         when "t"
-          CharacterRangeUnit.new("\t", "\t")
+          CharacterRangeUnit.new("\t")
         when "v"
-          CharacterRangeUnit.new("\v", "\v")
+          CharacterRangeUnit.new("\v")
+        when "w"
+          ccu = CharacterClassUnit.new
+          ccu << CharacterRangeUnit.new("_")
+          ccu << CharacterRangeUnit.new("0", "9")
+          ccu << CharacterRangeUnit.new("a", "z")
+          ccu << CharacterRangeUnit.new("A", "Z")
+          ccu
+        when "W"
+          ccu = CharacterClassUnit.new
+          ccu << CharacterRangeUnit.new("_")
+          ccu << CharacterRangeUnit.new("0", "9")
+          ccu << CharacterRangeUnit.new("a", "z")
+          ccu << CharacterRangeUnit.new("A", "Z")
+          ccu.negate = true
+          ccu
         else
           CharacterRangeUnit.new(c)
         end

lib/propane/regex/unit.rb

@@ -92,17 +92,20 @@ class Propane
         @units = []
         @negate = false
       end
-      def initialize
-        @units = []
-      end
-      def method_missing(*args)
-        @units.__send__(*args)
+      def method_missing(*args, &block)
+        @units.__send__(*args, &block)
       end
       def <<(thing)
         if thing.is_a?(CharacterClassUnit)
+          if thing.negate
+            CodePointRange.invert_ranges(thing.map(&:code_point_range)).each do |cpr|
+              CharacterRangeUnit.new(cpr.first, cpr.last)
+            end
+          else
             thing.each do |ccu_unit|
               @units << ccu_unit
             end
+          end
         else
           @units << thing
         end

lib/propane/util.rb

@@ -10,6 +10,32 @@ class Propane
         "#{s}\n* #{message} *\n#{s}\n"
       end
 
+      # Determine the number of threads to use.
+      #
+      # @return [Integer]
+      #   The number of threads to use.
+      def determine_n_threads
+        # Try to figure out how many threads are available on the host hardware.
+        begin
+          case RbConfig::CONFIG["host_os"]
+          when /linux/
+            return File.read("/proc/cpuinfo").scan(/^processor\s*:/).size
+          when /mswin|mingw|msys/
+            if `wmic cpu get NumberOfLogicalProcessors -value` =~ /NumberOfLogicalProcessors=(\d+)/
+              return $1.to_i
+            end
+          when /darwin/
+            if `sysctl -n hw.ncpu` =~ /(\d+)/
+              return $1.to_i
+            end
+          end
+        rescue
+        end
+
+        # If we can't figure it out, default to 4.
+        4
+      end
+
     end
 
   end

lib/propane/version.rb

@@ -1,3 +1,3 @@
 class Propane
-  VERSION = "2.2.1"
+  VERSION = "2.3.0"
 end

spec/propane/lexer/dfa_spec.rb

@@ -126,6 +126,74 @@ EOF
     ]
     expect(run(<<EOF, ";")).to eq expected
 token semicolon /;/;
+EOF
+  end
+
+  it "matches a negated character class" do
+    expected = [
+      ["pattern", "/abc/"],
+    ]
+    expect(run(<<EOF, "/abc/")).to eq expected
+token pattern /\\/[^\\s]*\\//;
+EOF
+  end
+
+  it "matches special character classes " do
+    expected = [
+      ["a", "abc123_FOO"],
+    ]
+    expect(run(<<EOF, "abc123_FOO")).to eq expected
+token a /\\w+/;
+EOF
+    expected = [
+      ["b", "FROG*%$#"],
+    ]
+    expect(run(<<EOF, "FROG*%$#")).to eq expected
+token b /FROG\\D{1,4}/;
+EOF
+    expected = [
+      ["c", "$883366"],
+    ]
+    expect(run(<<EOF, "$883366")).to eq expected
+token c /$\\d+/;
+EOF
+    expected = [
+      ["d", "^&$@"],
+    ]
+    expect(run(<<EOF, "^&$@")).to eq expected
+token d /^\\W+/;
+EOF
+    expected = [
+      ["a", "abc123_FOO"],
+      [nil, " "],
+      ["b", "FROG*%$#"],
+      [nil, " "],
+      ["c", "$883366"],
+      [nil, " "],
+      ["d", "^&$@"],
+    ]
+    expect(run(<<EOF, "abc123_FOO FROG*%$# $883366 ^&$@")).to eq expected
+token a /\\w+/;
+token b /FROG\\D{1,4}/;
+token c /$\\d+/;
+token d /^\\W+/;
+drop /\\s+/;
+EOF
+  end
+
+  it "matches a negated character class with a nested inner negated character class" do
+    expected = [
+      ["t", "$&*"],
+    ]
+    expect(run(<<EOF, "$&*")).to eq expected
+token t /[^%\\W]+/;
+EOF
+  end
+
+  it "\\s matches a newline" do
+    expected = [["s", "\n"]]
+    expect(run(<<EOF, "\n")).to eq expected
+token s /\\s/;
 EOF
   end
 end

spec/propane/regex_spec.rb

@@ -2,14 +2,14 @@ class Propane
   RSpec.describe Regex do
 
     it "parses an empty expression" do
-      regex = Regex.new("")
+      regex = Regex.new("", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0].size).to eq 0
     end
 
     it "parses a single character unit expression" do
-      regex = Regex.new("a")
+      regex = Regex.new("a", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -19,7 +19,7 @@ class Propane
     end
 
     it "parses a group with a single character unit expression" do
-      regex = Regex.new("(a)")
+      regex = Regex.new("(a)", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -33,7 +33,7 @@ class Propane
     end
 
     it "parses a *" do
-      regex = Regex.new("a*")
+      regex = Regex.new("a*", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -47,7 +47,7 @@ class Propane
     end
 
     it "parses a +" do
-      regex = Regex.new("a+")
+      regex = Regex.new("a+", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -61,7 +61,7 @@ class Propane
     end
 
     it "parses a ?" do
-      regex = Regex.new("a?")
+      regex = Regex.new("a?", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -75,7 +75,7 @@ class Propane
     end
 
     it "parses a multiplicity count" do
-      regex = Regex.new("a{5}")
+      regex = Regex.new("a{5}", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -89,7 +89,7 @@ class Propane
     end
 
     it "parses a minimum-only multiplicity count" do
-      regex = Regex.new("a{5,}")
+      regex = Regex.new("a{5,}", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -103,7 +103,7 @@ class Propane
     end
 
     it "parses a minimum and maximum multiplicity count" do
-      regex = Regex.new("a{5,8}")
+      regex = Regex.new("a{5,8}", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -118,7 +118,7 @@ class Propane
     end
 
     it "parses an escaped *" do
-      regex = Regex.new("a\\*")
+      regex = Regex.new("a\\*", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -131,7 +131,7 @@ class Propane
     end
 
     it "parses an escaped +" do
-      regex = Regex.new("a\\+")
+      regex = Regex.new("a\\+", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -144,7 +144,7 @@ class Propane
     end
 
     it "parses an escaped \\" do
-      regex = Regex.new("\\\\d")
+      regex = Regex.new("\\\\d", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -157,7 +157,7 @@ class Propane
     end
 
     it "parses a character class" do
-      regex = Regex.new("[a-z_]")
+      regex = Regex.new("[a-z_]", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -175,7 +175,7 @@ class Propane
     end
 
     it "parses a negated character class" do
-      regex = Regex.new("[^xyz]")
+      regex = Regex.new("[^xyz]", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -189,8 +189,25 @@ class Propane
       expect(ccu[0].first).to eq "x".ord
     end
 
+    it "parses a negated character class with inner character classes" do
+      regex = Regex.new("[^x\\sz]", 1)
+      expect(regex.unit).to be_a Regex::AlternatesUnit
+      expect(regex.unit.alternates.size).to eq 1
+      expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
+      seq_unit = regex.unit.alternates[0]
+      expect(seq_unit.size).to eq 1
+      expect(seq_unit[0]).to be_a Regex::CharacterClassUnit
+      ccu = seq_unit[0]
+      expect(ccu.negate).to be_truthy
+      expect(ccu.size).to eq 8
+      expect(ccu[0]).to be_a Regex::CharacterRangeUnit
+      expect(ccu[0].first).to eq "x".ord
+      expect(ccu[1].first).to eq " ".ord
+      expect(ccu[7].first).to eq "z".ord
+    end
+
     it "parses - as a plain character at beginning of a character class" do
-      regex = Regex.new("[-9]")
+      regex = Regex.new("[-9]", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -204,7 +221,7 @@ class Propane
     end
 
     it "parses - as a plain character at end of a character class" do
-      regex = Regex.new("[0-]")
+      regex = Regex.new("[0-]", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -220,7 +237,7 @@ class Propane
     end
 
     it "parses - as a plain character at beginning of a negated character class" do
-      regex = Regex.new("[^-9]")
+      regex = Regex.new("[^-9]", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -235,7 +252,7 @@ class Propane
     end
 
     it "parses . as a plain character in a character class" do
-      regex = Regex.new("[.]")
+      regex = Regex.new("[.]", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -250,7 +267,7 @@ class Propane
     end
 
     it "parses - as a plain character when escaped in middle of character class" do
-      regex = Regex.new("[0\\-9]")
+      regex = Regex.new("[0\\-9]", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -269,7 +286,7 @@ class Propane
     end
 
     it "parses alternates" do
-      regex = Regex.new("ab|c")
+      regex = Regex.new("ab|c", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 2
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -279,7 +296,7 @@ class Propane
     end
 
     it "parses a ." do
-      regex = Regex.new("a.b")
+      regex = Regex.new("a.b", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 1
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit
@@ -290,7 +307,7 @@ class Propane
     end
 
     it "parses something complex" do
-      regex = Regex.new("(a|)*|[^^]|\\|v|[x-y]+")
+      regex = Regex.new("(a|)*|[^^]|\\|v|[x-y]+", 1)
       expect(regex.unit).to be_a Regex::AlternatesUnit
       expect(regex.unit.alternates.size).to eq 4
       expect(regex.unit.alternates[0]).to be_a Regex::SequenceUnit

spec/propane_spec.rb

@@ -249,6 +249,18 @@ EOF
     expect(results.status).to eq 0
   end
 
+  it "shows error line number for unmatched left curly brace" do
+    write_grammar <<EOF
+# Line 1
+# Line 2
+token a /a{/;
+Start -> a;
+EOF
+    results = run_propane(capture: true)
+    expect(results.stderr).to match /Line 3: unexpected match count following \{/
+    expect(results.status).to_not eq 0
+  end
+
   %w[d c].each do |language|
 
     context "#{language.upcase} language" do