Use more concise RSpec output

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

.rspec

@@ -1,3 +1,2 @@
---format documentation
 --color
 --require spec_helper

CHANGELOG.md

@@ -1,3 +1,122 @@
+## 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
+
+- Allow multiple lexer modes to be specified for a lexer pattern (#35)
+- Document p_decode_code_point() API function (#34)
+
+## v2.1.1
+
+### Fixes
+
+- Field aliases for AST node fields could alias incorrect field when multiple rule alternatives present for one rule set (#33)
+
+## v2.1.0
+
+### Improvements
+
+- Report rule name and line number for conflicting AST node field positions errors (#32)
+
+## v2.0.0
+
+### Improvements
+
+- Log conflicting rules on reduce/reduce conflict (#31)
+- Use 1-based row and column values for position values (#30)
+
+### Fixes
+
+- Fix named optional rules (#29)
+
+### Upgrading
+
+- Adjust all uses of p_position_t row and col values to expect 1-based instead
+of 0-based values.
+
+## v1.5.1
+
+### Improvements
+
+- Improve performance (#28)
+
+## v1.5.0
+
+### New Features
+
+- Track start and end text positions for tokens and rules in AST node structures (#27)
+- Add warnings for shift/reduce conflicts to log file (#25)
+- Add -w command line switch to treat warnings as errors and output to stderr (#26)
+- Add rule field aliases (#24)
+
+### Improvements
+
+- Show line numbers of rules on conflict (#23)
+
+## v1.4.0
+
+### New Features
+
+- Allow user to specify AST node name prefix or suffix
+- Allow specifying the start rule name
+- Allow rule terms to be marked as optional
+
+### Improvements
+
+- Give a better error message when a referenced ptype has not been declared
+
+## v1.3.0
+
+### New Features
+
+- Add AST generation (#22)
+
+## v1.2.0
+
+### New Features
+
+- Allow one line user code blocks (#21)
+- Add backslash escape codes (#19)
+- Add API to access unexpected token found (#18)
+- Add token_names API (#17)
+- Add D example to user guide for p_context_init() (#16)
+- Allow user termination from lexer code blocks (#15)
+
+### Fixes
+
+- Fix generator hang when state transition cycle is present (#20)
+
+## v1.1.0
+
+### New Features
+
+- Add user parser terminations (#13)
+- Document generated parser API in user guide (#14)
+
 ## v1.0.0
 
 - Initial release

Gemfile

@@ -1,5 +1,6 @@
 source "https://rubygems.org"
 
+gem "base64"
 gem "rake"
 gem "rspec"
 gem "rdoc"

Gemfile.lock

@@ -1,40 +1,48 @@
 GEM
   remote: https://rubygems.org/
   specs:
-    diff-lcs (1.5.0)
-    docile (1.4.0)
-    psych (5.1.0)
+    base64 (0.3.0)
+    date (3.5.1)
+    diff-lcs (1.6.2)
+    docile (1.4.1)
+    erb (6.0.1)
+    psych (5.3.1)
+      date
       stringio
-    rake (13.0.6)
-    rdoc (6.5.0)
+    rake (13.3.1)
+    rdoc (7.1.0)
+      erb
       psych (>= 4.0.0)
-    redcarpet (3.6.0)
-    rspec (3.12.0)
-      rspec-core (~> 3.12.0)
-      rspec-expectations (~> 3.12.0)
-      rspec-mocks (~> 3.12.0)
-    rspec-core (3.12.2)
-      rspec-support (~> 3.12.0)
-    rspec-expectations (3.12.3)
+      tsort
+    redcarpet (3.6.1)
+    rspec (3.13.2)
+      rspec-core (~> 3.13.0)
+      rspec-expectations (~> 3.13.0)
+      rspec-mocks (~> 3.13.0)
+    rspec-core (3.13.6)
+      rspec-support (~> 3.13.0)
+    rspec-expectations (3.13.5)
       diff-lcs (>= 1.2.0, < 2.0)
-      rspec-support (~> 3.12.0)
-    rspec-mocks (3.12.6)
+      rspec-support (~> 3.13.0)
+    rspec-mocks (3.13.7)
       diff-lcs (>= 1.2.0, < 2.0)
-      rspec-support (~> 3.12.0)
-    rspec-support (3.12.1)
+      rspec-support (~> 3.13.0)
+    rspec-support (3.13.6)
     simplecov (0.22.0)
       docile (~> 1.1)
       simplecov-html (~> 0.11)
       simplecov_json_formatter (~> 0.1)
-    simplecov-html (0.12.3)
+    simplecov-html (0.13.2)
     simplecov_json_formatter (0.1.4)
-    stringio (3.0.7)
+    stringio (3.2.0)
     syntax (1.2.2)
+    tsort (0.2.0)
 
 PLATFORMS
   ruby
 
 DEPENDENCIES
+  base64
   rake
   rdoc
   redcarpet

README.md

@@ -6,7 +6,8 @@ Propane is a LALR Parser Generator (LPG) which:
   * generates a built-in lexer to tokenize input
   * supports UTF-8 lexer inputs
   * generates a table-driven shift/reduce parser to parse input in linear time
-  * target C or D language outputs
+  * targets C, C++, or D language outputs
+  * optionally supports automatic full AST generation
   * is MIT-licensed
   * is distributable as a standalone Ruby script
 
@@ -17,6 +18,7 @@ can be copied into and versioned in a project's source tree.
 The only requirement to run Propane is that the system has a Ruby interpreter
 installed.
 The latest release can be downloaded from [https://github.com/holtrop/propane/releases](https://github.com/holtrop/propane/releases).
+
 Simply copy the `propane` executable script into the desired location within
 the project to be built (typically the root of the repository) and mark it
 executable.
@@ -29,9 +31,14 @@ Propane is typically invoked from the command-line as `./propane`.
 
     Usage: ./propane [options] <input-file> <output-file>
     Options:
-      --log LOG   Write log file
-      --version   Show program version and exit
-      -h, --help  Show this usage and exit
+      -h, --help  Show this usage and exit.
+      --log LOG   Write log file. This will show all parser states and their
+                  associated shifts and reduces. It can be helpful when
+                  debugging a grammar.
+      --version   Show program version and exit.
+      -w          Treat warnings as errors. This option will treat shift/reduce
+                  conflicts as fatal errors and will print them to stderr in
+                  addition to the log file.
 
 The user must specify the path to a Propane input grammar file and a path to an
 output file.
@@ -55,10 +62,10 @@ 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)
   {
@@ -67,38 +74,22 @@ token integer /\\d+/ <<
   }
   $$ = v;
 >>
-token lparen /\\(/;
-token rparen /\\)/;
+token lparen /\(/;
+token rparen /\)/;
 # Drop whitespace.
-drop /\\s+/;
+drop /\s+/;
 
-Start -> E1 <<
-  $$ = $1;
->>
-E1 -> E2 <<
-  $$ = $1;
->>
-E1 -> E1 plus E2 <<
-  $$ = $1 + $3;
->>
-E2 -> E3 <<
-  $$ = $1;
->>
-E2 -> E2 times E3 <<
-  $$ = $1 * $3;
->>
-E3 -> E4 <<
-  $$ = $1;
->>
+Start -> E1 << $$ = $1; >>
+E1 -> E2 << $$ = $1; >>
+E1 -> E1 plus E2 << $$ = $1 + $3; >>
+E2 -> E3 << $$ = $1; >>
+E2 -> E2 times E3 << $$ = $1 * $3; >>
+E3 -> E4 << $$ = $1; >>
 E3 -> E3 power E4 <<
   $$ = pow($1, $3);
 >>
-E4 -> integer <<
-  $$ = $1;
->>
-E4 -> lparen E1 rparen <<
-  $$ = $2;
->>
+E4 -> integer << $$ = $1; >>
+E4 -> lparen E1 rparen << $$ = $2; >>
 ```
 
 Grammar files can contain comment lines beginning with `#` which are ignored.

Rakefile.rb

@@ -1,5 +1,7 @@
 require "rake/clean"
 require "rspec/core/rake_task"
+require "simplecov"
+require "stringio"
 
 CLEAN.include %w[spec/run gen .yardoc yard coverage dist]
 
@@ -12,6 +14,18 @@ RSpec::Core::RakeTask.new(:spec, :example_pattern) do |task, args|
     task.rspec_opts = %W[-e "#{args.example_pattern}" -f documentation]
   end
 end
+task :spec do |task, args|
+  unless ENV["dist_specs"]
+    original_stdout = $stdout
+    sio = StringIO.new
+    $stdout = sio
+    SimpleCov.collate Dir["coverage/.resultset.json"]
+    $stdout = original_stdout
+    sio.string.lines.each do |line|
+      $stdout.write(line) unless line =~ /Coverage report generated for/
+    end
+  end
+end
 
 # dspec task is useful to test the distributable release script, but is not
 # useful for coverage information.

assets/parser.c.erb

@@ -3,6 +3,17 @@
 #include <stdlib.h>
 #include <string.h>
 
+/**************************************************************************
+ * Public data
+ *************************************************************************/
+
+/** Token names. */
+const char * <%= @grammar.prefix %>token_names[] = {
+<% @grammar.tokens.each_with_index do |token, index| %>
+    "<%= token.name %>",
+<% end %>
+};
+
 /**************************************************************************
  * User code blocks
  *************************************************************************/
@@ -44,11 +55,14 @@
 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;
     newcontext.input_length = input_length;
+    newcontext.text_position.row = 1u;
+    newcontext.text_position.col = 1u;
     newcontext.mode = <%= @lexer.mode_id("default") %>;
 
     /* Copy to the user's context structure. */
@@ -215,7 +229,10 @@ typedef struct
     /** Number of bytes of input text used to match. */
     size_t length;
 
-    /** Input text position delta. */
+    /** Input text position delta to end of token. */
+    <%= @grammar.prefix %>position_t end_delta_position;
+
+    /** Input text position delta to next code point after token end. */
     <%= @grammar.prefix %>position_t delta_position;
 
     /** Accepting lexer state from the match. */
@@ -309,9 +326,12 @@ static lexer_state_id_t check_lexer_transition(uint32_t current_state, uint32_t
  *
  * @param context
  *   Lexer/parser context structure.
- * @param[out] out_token_info
- *   The lexed token information is stored here if the return value is
- *   P_SUCCESS.
+ * @param[out] out_match_info
+ *   The longest match information is stored here if the return value is
+ *   P_SUCCESS or P_DECODE_ERROR.
+ * @param[out] out_unexpected_input_length
+ *   The unexpected input length is stored here if the return value is
+ *   P_UNEXPECTED_INPUT.
  *
  * @reval P_SUCCESS
  *   A token was successfully lexed.
@@ -325,8 +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 (;;)
@@ -340,14 +362,16 @@ 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)
             {
                 attempt_match.length += code_point_length;
+                attempt_match.end_delta_position = attempt_match.delta_position;
                 if (code_point == '\n')
                 {
                     attempt_match.delta_position.row++;
-                    attempt_match.delta_position.col = 0u;
+                    attempt_match.delta_position.col = 1u;
                 }
                 else
                 {
@@ -370,6 +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:
@@ -422,25 +447,34 @@ static size_t find_longest_match(<%= @grammar.prefix %>context_t * context,
  *   Input text does not match any lexer pattern.
  * @retval P_DROP
  *   A drop pattern was matched so the lexer should continue.
+ * @retval P_USER_TERMINATED
+ *   User code has requested to terminate the lexer.
  */
 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;
-    *out_token_info = token_info; // TODO: remove
     lexer_match_info_t match_info;
     size_t unexpected_input_length;
     size_t result = find_longest_match(context, &match_info, &unexpected_input_length);
     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)
         {
             uint8_t const * match = &context->input[context->input_index];
             <%= @grammar.prefix %>token_t user_code_token = lexer_user_code(context,
                 match_info.accepting_state->code_id, match, match_info.length, &token_info);
+            /* A TERMINATE_TOKEN_ID return code from lexer_user_code() means
+             * that the user code is requesting to terminate the lexer. */
+            if (user_code_token == TERMINATE_TOKEN_ID)
+            {
+                return P_USER_TERMINATED;
+            }
             /* An invalid token returned from lexer_user_code() means that the
              * user code did not explicitly return a token. So only override
              * the token to return if the user code does explicitly return a
@@ -469,11 +503,23 @@ static size_t attempt_lex_token(<%= @grammar.prefix %>context_t * context, <%= @
         }
         token_info.token = token_to_accept;
         token_info.length = match_info.length;
+        if (match_info.end_delta_position.row != 0u)
+        {
+            token_info.end_position.row = token_info.position.row + match_info.end_delta_position.row;
+            token_info.end_position.col = match_info.end_delta_position.col;
+        }
+        else
+        {
+            token_info.end_position.row = token_info.position.row;
+            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:
         token_info.token = TOKEN___EOF;
+        token_info.end_position = token_info.position;
         *out_token_info = token_info;
         return P_SUCCESS;
 
@@ -511,6 +557,8 @@ static size_t attempt_lex_token(<%= @grammar.prefix %>context_t * context, <%= @
  *   The decoder encountered invalid text encoding.
  * @reval P_UNEXPECTED_INPUT
  *   Input text does not match any lexer pattern.
+ * @retval P_USER_TERMINATED
+ *   User code has requested to terminate the lexer.
  */
 size_t <%= @grammar.prefix %>lex(<%= @grammar.prefix %>context_t * context, <%= @grammar.prefix %>token_info_t * out_token_info)
 {
@@ -528,6 +576,9 @@ size_t <%= @grammar.prefix %>lex(<%= @grammar.prefix %>context_t * context, <%=
  * Parser
  *************************************************************************/
 
+/** Invalid position value. */
+#define INVALID_POSITION (<%= @grammar.prefix %>position_t){0u, 0u}
+
 /** Reduce ID type. */
 typedef <%= get_type_for(@parser.reduce_table.size) %> reduce_id_t;
 
@@ -587,6 +638,25 @@ typedef struct
      * reduce action.
      */
     parser_state_id_t n_states;
+<% if @grammar.ast %>
+
+    /**
+     * Map of rule components to rule set child fields.
+     */
+    uint16_t const * rule_set_node_field_index_map;
+
+    /**
+     * Number of rule set AST node fields.
+     */
+    uint16_t rule_set_node_field_array_size;
+
+    /**
+     * Whether this rule was a generated optional rule that matched the
+     * optional target. In this case, propagate the matched target node up
+     * instead of making a new node for this rule.
+     */
+    bool propagate_optional_target;
+<% end %>
 } reduce_t;
 
 /** Parser state entry. */
@@ -617,19 +687,55 @@ typedef struct
 
     /** Parser value from this state. */
     <%= @grammar.prefix %>value_t pvalue;
+
+<% if @grammar.ast %>
+    /** AST node. */
+    void * ast_node;
+<% end %>
 } state_value_t;
 
+/** Common AST node structure. */
+typedef struct
+{
+    <%= @grammar.prefix %>position_t position;
+    <%= @grammar.prefix %>position_t end_position;
+    void * fields[];
+} ASTNode;
+
 /** Parser shift table. */
 static const shift_t parser_shift_table[] = {
 <%   @parser.shift_table.each do |shift| %>
-    {<%= shift[:symbol_id] %>u, <%= shift[:state_id] %>u},
+    {<%= shift[:symbol].id %>u, <%= shift[:state_id] %>u},
 <%   end %>
 };
 
+<% if @grammar.ast %>
+<%   @grammar.rules.each do |rule| %>
+<%     unless rule.flat_rule_set_node_field_index_map? %>
+const uint16_t r_<%= rule.name.gsub("$", "_") %><%= rule.id %>_node_field_index_map[<%= rule.rule_set_node_field_index_map.size %>] = {<%= rule.rule_set_node_field_index_map.map {|v| v.to_s}.join(", ") %>};
+<%     end %>
+<%   end %>
+<% end %>
+
 /** Parser reduce table. */
 static const reduce_t parser_reduce_table[] = {
 <%   @parser.reduce_table.each do |reduce| %>
-    {<%= reduce[:token_id] %>u, <%= reduce[:rule_id] %>u, <%= reduce[:rule_set_id] %>u, <%= reduce[:n_states] %>u},
+    {
+        <%= reduce[:token_id] %>u, /* Token: <%= reduce[:token] ? reduce[:token].name : "(any)" %> */
+        <%= reduce[:rule_id] %>u, /* Rule ID */
+        <%= reduce[:rule_set_id] %>u, /* Rule set ID (<%= reduce[:rule].rule_set.name %>) */
+<%     if @grammar.ast %>
+        <%= reduce[:n_states] %>u, /* Number of states */
+<%       if reduce[:rule].flat_rule_set_node_field_index_map? %>
+        NULL, /* No rule set node field index map (flat map) */
+<%       else %>
+        &r_<%= reduce[:rule].name.gsub("$", "_") %><%= reduce[:rule].id %>_node_field_index_map[0], /* Rule set node field index map */
+<%       end %>
+        <%= reduce[:rule].rule_set.ast_fields.size %>, /* Number of AST fields */
+        <%= reduce[:propagate_optional_target] %>}, /* Propagate optional target? */
+<%     else %>
+        <%= reduce[:n_states] %>u},
+<%     end %>
 <%   end %>
 };
 
@@ -699,7 +805,7 @@ static void state_values_stack_push(state_values_stack_t * stack)
     if (current_length >= current_capacity)
     {
         size_t const new_capacity = current_capacity * 2u;
-        state_value_t * new_entries = malloc(new_capacity * sizeof(state_value_t));
+        state_value_t * new_entries = (state_value_t *)malloc(new_capacity * sizeof(state_value_t));
         memcpy(new_entries, stack->entries, current_length * sizeof(state_value_t));
         free(stack->entries);
         stack->capacity = new_capacity;
@@ -733,6 +839,7 @@ static void state_values_stack_free(state_values_stack_t * stack)
      free(stack->entries);
 }
 
+<% unless @grammar.ast %>
 /**
  * Execute user code associated with a parser rule.
  *
@@ -747,18 +854,19 @@ static size_t parser_user_code(<%= @grammar.prefix %>value_t * _pvalue, uint32_t
 {
     switch (rule)
     {
-<% @grammar.rules.each do |rule| %>
-<%   if rule.code %>
+<%   @grammar.rules.each do |rule| %>
+<%     if rule.code %>
     case <%= rule.id %>u: {
 <%= expand_code(rule.code, true, rule, nil) %>
     } break;
+<%     end %>
 <%   end %>
-<% end %>
     default: break;
     }
 
     return P_SUCCESS;
 }
+<% end %>
 
 /**
  * Check if the parser should shift to a new state.
@@ -820,7 +928,7 @@ static size_t check_reduce(size_t state_id, <%= @grammar.prefix %>token_t token)
  *   can be accessed with <%= @grammar.prefix %>result().
  * @retval P_UNEXPECTED_TOKEN
  *   An unexpected token was encountered that does not match any grammar rule.
- *   The value context->token holds the unexpected token.
+ *   The function p_token(&context) can be used to get the unexpected token.
  * @reval P_DECODE_ERROR
  *   The decoder encountered invalid text encoding.
  * @reval P_UNEXPECTED_INPUT
@@ -832,7 +940,11 @@ size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * context)
     <%= @grammar.prefix %>token_t token = INVALID_TOKEN_ID;
     state_values_stack_t statevalues;
     size_t reduced_rule_set = INVALID_ID;
+<% if @grammar.ast %>
+    void * reduced_parser_node;
+<% else %>
     <%= @grammar.prefix %>value_t reduced_parser_value;
+<% end %>
     state_values_stack_init(&statevalues);
     state_values_stack_push(&statevalues);
     size_t result;
@@ -859,7 +971,11 @@ size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * context)
             if ((shift_state != INVALID_ID) && (token == TOKEN___EOF))
             {
                 /* Successful parse. */
+<% if @grammar.ast %>
+                context->parse_result = (<%= @grammar.ast_prefix %><%= @grammar.start_rule %><%= @grammar.ast_suffix %> *)state_values_stack_index(&statevalues, -1)->ast_node;
+<% else %>
                 context->parse_result = state_values_stack_index(&statevalues, -1)->pvalue;
+<% end %>
                 result = P_SUCCESS;
                 break;
             }
@@ -872,15 +988,29 @@ size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * context)
             if (reduced_rule_set == INVALID_ID)
             {
                 /* We shifted a token, mark it consumed. */
-                token = INVALID_TOKEN_ID;
+<% if @grammar.ast %>
+                <%= @grammar.ast_prefix %>Token<%= @grammar.ast_suffix %> * token_ast_node = (<%= @grammar.ast_prefix %>Token<%= @grammar.ast_suffix %> *)malloc(sizeof(<%= @grammar.ast_prefix %>Token<%= @grammar.ast_suffix %>));
+                token_ast_node->position = token_info.position;
+                token_ast_node->end_position = token_info.end_position;
+                token_ast_node->token = token;
+                token_ast_node->pvalue = token_info.pvalue;
+                state_values_stack_index(&statevalues, -1)->ast_node = token_ast_node;
+<% else %>
                 state_values_stack_index(&statevalues, -1)->pvalue = token_info.pvalue;
+<% end %>
+                token = INVALID_TOKEN_ID;
             }
             else
             {
                 /* We shifted a RuleSet. */
+<% if @grammar.ast %>
+                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;
             }
             continue;
@@ -890,12 +1020,64 @@ size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * context)
         if (reduce_index != INVALID_ID)
         {
             /* We have something to reduce. */
-            <%= @grammar.prefix %>value_t reduced_parser_value2 = {0};
+<% if @grammar.ast %>
+            if (parser_reduce_table[reduce_index].propagate_optional_target)
+            {
+                reduced_parser_node = state_values_stack_index(&statevalues, -1)->ast_node;
+            }
+            else if (parser_reduce_table[reduce_index].n_states > 0)
+            {
+                size_t n_fields = parser_reduce_table[reduce_index].rule_set_node_field_array_size;
+                ASTNode * node = (ASTNode *)malloc(sizeof(ASTNode) + n_fields * sizeof(void *));
+                node->position = INVALID_POSITION;
+                node->end_position = INVALID_POSITION;
+                for (size_t i = 0; i < n_fields; i++)
+                {
+                    node->fields[i] = NULL;
+                }
+                if (parser_reduce_table[reduce_index].rule_set_node_field_index_map == NULL)
+                {
+                    for (size_t i = 0; i < parser_reduce_table[reduce_index].n_states; i++)
+                    {
+                        node->fields[i] = state_values_stack_index(&statevalues, -(int)parser_reduce_table[reduce_index].n_states + (int)i)->ast_node;
+                    }
+                }
+                else
+                {
+                    for (size_t i = 0; i < parser_reduce_table[reduce_index].n_states; i++)
+                    {
+                        node->fields[parser_reduce_table[reduce_index].rule_set_node_field_index_map[i]] = state_values_stack_index(&statevalues, -(int)parser_reduce_table[reduce_index].n_states + (int)i)->ast_node;
+                    }
+                }
+                bool position_found = false;
+                for (size_t i = 0; i < n_fields; i++)
+                {
+                    ASTNode * child = (ASTNode *)node->fields[i];
+                    if ((child != NULL) && <%= @grammar.prefix %>position_valid(child->position))
+                    {
+                        if (!position_found)
+                        {
+                            node->position = child->position;
+                            position_found = true;
+                        }
+                        node->end_position = child->end_position;
+                    }
+                }
+                reduced_parser_node = node;
+            }
+            else
+            {
+                reduced_parser_node = NULL;
+            }
+<% else %>
+            <%= @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;
             }
             reduced_parser_value = reduced_parser_value2;
+<% end %>
             reduced_rule_set = parser_reduce_table[reduce_index].rule_set;
             state_values_stack_pop(&statevalues, parser_reduce_table[reduce_index].n_states);
             continue;
@@ -923,9 +1105,17 @@ size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * context)
  *
  * @return Parse result value.
  */
+<% if @grammar.ast %>
+<%= @grammar.ast_prefix %><%= @grammar.start_rule %><%= @grammar.ast_suffix %> * <%= @grammar.prefix %>result(<%= @grammar.prefix %>context_t * context)
+<% else %>
 <%= start_rule_type[1] %> <%= @grammar.prefix %>result(<%= @grammar.prefix %>context_t * context)
+<% end %>
 {
+<% if @grammar.ast %>
+    return context->parse_result;
+<% else %>
     return context->parse_result.v_<%= start_rule_type[0] %>;
+<% end %>
 }
 
 /**
@@ -953,3 +1143,13 @@ size_t <%= @grammar.prefix %>user_terminate_code(<%= @grammar.prefix %>context_t
 {
     return context->user_terminate_code;
 }
+
+/**
+ * Get the parse token.
+ *
+ * @return Parse token.
+ */
+<%= @grammar.prefix %>token_t <%= @grammar.prefix %>token(<%= @grammar.prefix %>context_t * context)
+{
+    return context->token;
+}

assets/parser.d.erb

@@ -8,6 +8,9 @@
 module <%= @grammar.modulename %>;
 <% end %>
 
+import core.memory;
+import core.stdc.stdlib : malloc;
+
 /**************************************************************************
  * User code blocks
  *************************************************************************/
@@ -31,7 +34,7 @@ public enum : size_t
 }
 
 /** Token type. */
-public alias <%= @grammar.prefix %>token_t = <%= get_type_for(@grammar.invalid_token_id) %>;
+public alias <%= @grammar.prefix %>token_t = <%= get_type_for(@grammar.terminate_token_id) %>;
 
 /** Token IDs. */
 public enum : <%= @grammar.prefix %>token_t
@@ -43,21 +46,14 @@ public enum : <%= @grammar.prefix %>token_t
 <%   end %>
 <% end %>
     INVALID_TOKEN_ID = <%= @grammar.invalid_token_id %>,
+    TERMINATE_TOKEN_ID = <%= @grammar.terminate_token_id %>,
 }
 
 /** Code point type. */
 public alias <%= @grammar.prefix %>code_point_t = uint;
 
-/** Parser values type(s). */
-public union <%= @grammar.prefix %>value_t
-{
-<% @grammar.ptypes.each do |name, typestring| %>
-    <%= typestring %> v_<%= name %>;
-<% end %>
-}
-
 /**
- * A structure to keep track of parser position.
+ * A structure to keep track of input position.
  *
  * This is useful for reporting errors, etc...
  */
@@ -68,14 +64,79 @@ public struct <%= @grammar.prefix %>position_t
 
     /** Input text column (0-based). */
     uint col;
+
+    /** Invalid position value. */
+    enum INVALID = <%= @grammar.prefix %>position_t(0u, 0u);
+
+    /** Return whether the position is valid. */
+    public @property bool valid()
+    {
+        return row != 0u;
+    }
 }
 
+<% if @grammar.ast %>
+/** Parser values type. */
+public alias <%= @grammar.prefix %>value_t = <%= @grammar.ptype %>;
+<% else %>
+/** Parser values type(s). */
+public union <%= @grammar.prefix %>value_t
+{
+<%   @grammar.ptypes.each do |name, typestring| %>
+    <%= typestring %> v_<%= name %>;
+<%   end %>
+}
+<% end %>
+
+<% if @grammar.ast %>
+/** Common AST node structure. */
+private struct ASTNode
+{
+    <%= @grammar.prefix %>position_t position;
+    <%= @grammar.prefix %>position_t end_position;
+    void *[0] fields;
+}
+
+/** AST node types. @{ */
+public struct <%= @grammar.ast_prefix %>Token<%= @grammar.ast_suffix %>
+{
+    /* ASTNode fields must be present in the same order here. */
+    <%= @grammar.prefix %>position_t position;
+    <%= @grammar.prefix %>position_t end_position;
+    <%= @grammar.prefix %>token_t token;
+    <%= @grammar.prefix %>value_t pvalue;
+}
+
+<%   @parser.rule_sets.each do |name, rule_set| %>
+<%     next if name.start_with?("$") %>
+<%     next if rule_set.optional? %>
+public struct <%= @grammar.ast_prefix %><%= name %><%= @grammar.ast_suffix %>
+{
+    <%= @grammar.prefix %>position_t position;
+    <%= @grammar.prefix %>position_t end_position;
+<%     rule_set.ast_fields.each do |fields| %>
+    union
+    {
+<%       fields.each do |field_name, type| %>
+        <%= type %> * <%= field_name %>;
+<%       end %>
+    }
+<%     end %>
+}
+
+<%   end %>
+/** @} */
+<% end %>
+
 /** Lexed token information. */
 public struct <%= @grammar.prefix %>token_info_t
 {
-    /** Text position where the token was found. */
+    /** Text position of first code point in token. */
     <%= @grammar.prefix %>position_t position;
 
+    /** Text position of last code point in token. */
+    <%= @grammar.prefix %>position_t end_position;
+
     /** Number of input bytes used by the token. */
     size_t length;
 
@@ -111,7 +172,11 @@ public struct <%= @grammar.prefix %>context_t
     /* Parser context data. */
 
     /** Parse result value. */
+<% if @grammar.ast %>
+    <%= @grammar.ast_prefix %><%= @grammar.start_rule %><%= @grammar.ast_suffix %> * parse_result;
+<% else %>
     <%= @grammar.prefix %>value_t parse_result;
+<% end %>
 
     /** Unexpected token received. */
     <%= @grammar.prefix %>token_t token;
@@ -171,6 +236,8 @@ public void <%= @grammar.prefix %>context_init(<%= @grammar.prefix %>context_t *
 
     /* Lexer initialization. */
     newcontext.input = input;
+    newcontext.text_position.row = 1u;
+    newcontext.text_position.col = 1u;
     newcontext.mode = <%= @lexer.mode_id("default") %>;
 
     /* Copy to the user's context structure. */
@@ -335,7 +402,10 @@ private struct lexer_match_info_t
     /** Number of bytes of input text used to match. */
     size_t length;
 
-    /** Input text position delta. */
+    /** Input text position delta to end of token. */
+    <%= @grammar.prefix %>position_t end_delta_position;
+
+    /** Input text position delta to next code point after token end. */
     <%= @grammar.prefix %>position_t delta_position;
 
     /** Accepting lexer state from the match. */
@@ -427,9 +497,12 @@ private lexer_state_id_t check_lexer_transition(uint current_state, uint code_po
  *
  * @param context
  *   Lexer/parser context structure.
- * @param[out] out_token_info
- *   The lexed token information is stored here if the return value is
- *   P_SUCCESS.
+ * @param[out] out_match_info
+ *   The longest match information is stored here if the return value is
+ *   P_SUCCESS or P_DECODE_ERROR.
+ * @param[out] out_unexpected_input_length
+ *   The unexpected input length is stored here if the return value is
+ *   P_UNEXPECTED_INPUT.
  *
  * @reval P_SUCCESS
  *   A token was successfully lexed.
@@ -460,10 +533,11 @@ private size_t find_longest_match(<%= @grammar.prefix %>context_t * context,
             if (transition_state != INVALID_LEXER_STATE_ID)
             {
                 attempt_match.length += code_point_length;
+                attempt_match.end_delta_position = attempt_match.delta_position;
                 if (code_point == '\n')
                 {
                     attempt_match.delta_position.row++;
-                    attempt_match.delta_position.col = 0u;
+                    attempt_match.delta_position.col = 1u;
                 }
                 else
                 {
@@ -538,13 +612,14 @@ private size_t find_longest_match(<%= @grammar.prefix %>context_t * context,
  *   Input text does not match any lexer pattern.
  * @retval P_DROP
  *   A drop pattern was matched so the lexer should continue.
+ * @retval P_USER_TERMINATED
+ *   User code has requested to terminate the lexer.
  */
 private 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;
     token_info.position = context.text_position;
     token_info.token = INVALID_TOKEN_ID;
-    *out_token_info = token_info; // TODO: remove
     lexer_match_info_t match_info;
     size_t unexpected_input_length;
     size_t result = find_longest_match(context, &match_info, &unexpected_input_length);
@@ -557,6 +632,12 @@ private size_t attempt_lex_token(<%= @grammar.prefix %>context_t * context, <%=
             string match = context.input[context.input_index..(context.input_index + match_info.length)];
             <%= @grammar.prefix %>token_t user_code_token = lexer_user_code(context,
                 match_info.accepting_state.code_id, match, &token_info);
+            /* A TERMINATE_TOKEN_ID return code from lexer_user_code() means
+             * that the user code is requesting to terminate the lexer. */
+            if (user_code_token == TERMINATE_TOKEN_ID)
+            {
+                return P_USER_TERMINATED;
+            }
             /* An invalid token returned from lexer_user_code() means that the
              * user code did not explicitly return a token. So only override
              * the token to return if the user code does explicitly return a
@@ -585,11 +666,22 @@ private size_t attempt_lex_token(<%= @grammar.prefix %>context_t * context, <%=
         }
         token_info.token = token_to_accept;
         token_info.length = match_info.length;
+        if (match_info.end_delta_position.row != 0u)
+        {
+            token_info.end_position.row = token_info.position.row + match_info.end_delta_position.row;
+            token_info.end_position.col = match_info.end_delta_position.col;
+        }
+        else
+        {
+            token_info.end_position.row = token_info.position.row;
+            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:
         token_info.token = TOKEN___EOF;
+        token_info.end_position = token_info.position;
         *out_token_info = token_info;
         return P_SUCCESS;
 
@@ -627,6 +719,8 @@ private size_t attempt_lex_token(<%= @grammar.prefix %>context_t * context, <%=
  *   The decoder encountered invalid text encoding.
  * @reval P_UNEXPECTED_INPUT
  *   Input text does not match any lexer pattern.
+ * @retval P_USER_TERMINATED
+ *   User code has requested to terminate the lexer.
  */
 public size_t <%= @grammar.prefix %>lex(<%= @grammar.prefix %>context_t * context, <%= @grammar.prefix %>token_info_t * out_token_info)
 {
@@ -703,6 +797,25 @@ private struct reduce_t
      * reduce action.
      */
     parser_state_id_t n_states;
+<% if @grammar.ast %>
+
+    /**
+     * Map of rule components to rule set child fields.
+     */
+    immutable(ushort) * rule_set_node_field_index_map;
+
+    /**
+     * Number of rule set AST node fields.
+     */
+    ushort rule_set_node_field_array_size;
+
+    /**
+     * Whether this rule was a generated optional rule that matched the
+     * optional target. In this case, propagate the matched target node up
+     * instead of making a new node for this rule.
+     */
+    bool propagate_optional_target;
+<% end %>
 }
 
 /** Parser state entry. */
@@ -734,6 +847,11 @@ private struct state_value_t
     /** Parser value from this state. */
     <%= @grammar.prefix %>value_t pvalue;
 
+<% if @grammar.ast %>
+    /** AST node. */
+    void * ast_node;
+<% end %>
+
     this(size_t state_id)
     {
         this.state_id = state_id;
@@ -743,14 +861,37 @@ private struct state_value_t
 /** Parser shift table. */
 private immutable shift_t[] parser_shift_table = [
 <%   @parser.shift_table.each do |shift| %>
-    shift_t(<%= shift[:symbol_id] %>u, <%= shift[:state_id] %>u),
+    shift_t(<%= shift[:symbol].id %>u, <%= shift[:state_id] %>u),
 <%   end %>
 ];
 
+<% if @grammar.ast %>
+<%   @grammar.rules.each do |rule| %>
+<%     unless rule.flat_rule_set_node_field_index_map? %>
+immutable ushort[<%= rule.rule_set_node_field_index_map.size %>] r_<%= rule.name.gsub("$", "_") %><%= rule.id %>_node_field_index_map = [<%= rule.rule_set_node_field_index_map.map {|v| v.to_s}.join(", ") %>];
+<%     end %>
+<%   end %>
+<% end %>
+
 /** Parser reduce table. */
 private immutable reduce_t[] parser_reduce_table = [
 <%   @parser.reduce_table.each do |reduce| %>
-    reduce_t(<%= reduce[:token_id] %>u, <%= reduce[:rule_id] %>u, <%= reduce[:rule_set_id] %>u, <%= reduce[:n_states] %>u),
+    reduce_t(
+        <%= reduce[:token_id] %>u, /* Token: <%= reduce[:token] ? reduce[:token].name : "(any)" %> */
+        <%= reduce[:rule_id] %>u, /* Rule ID */
+        <%= reduce[:rule_set_id] %>u, /* Rule set ID (<%= reduce[:rule].rule_set.name %>) */
+<%     if @grammar.ast %>
+        <%= reduce[:n_states] %>u, /* Number of states */
+<%       if reduce[:rule].flat_rule_set_node_field_index_map? %>
+        null, /* No rule set node field index map (flat map) */
+<%       else %>
+        &r_<%= reduce[:rule].name.gsub("$", "_") %><%= reduce[:rule].id %>_node_field_index_map[0], /* Rule set node field index map */
+<%       end %>
+        <%= reduce[:rule].rule_set.ast_fields.size %>, /* Number of AST fields */
+        <%= reduce[:propagate_optional_target] %>), /* Propagate optional target? */
+<%     else %>
+        <%= reduce[:n_states] %>u), /* Number of states */
+<%     end %>
 <%   end %>
 ];
 
@@ -761,6 +902,7 @@ private immutable parser_state_t[] parser_state_table = [
 <%   end %>
 ];
 
+<% unless @grammar.ast %>
 /**
  * Execute user code associated with a parser rule.
  *
@@ -775,18 +917,19 @@ private size_t parser_user_code(<%= @grammar.prefix %>value_t * _pvalue, uint ru
 {
     switch (rule)
     {
-<% @grammar.rules.each do |rule| %>
-<%   if rule.code %>
+<%   @grammar.rules.each do |rule| %>
+<%     if rule.code %>
     case <%= rule.id %>u: {
 <%= expand_code(rule.code, true, rule, nil) %>
     } break;
+<%     end %>
 <%   end %>
-<% end %>
     default: break;
     }
 
     return P_SUCCESS;
 }
+<% end %>
 
 /**
  * Check if the parser should shift to a new state.
@@ -848,7 +991,7 @@ private size_t check_reduce(size_t state_id, <%= @grammar.prefix %>token_t token
  *   can be accessed with <%= @grammar.prefix %>result().
  * @retval P_UNEXPECTED_TOKEN
  *   An unexpected token was encountered that does not match any grammar rule.
- *   The value context.token holds the unexpected token.
+ *   The function p_token(&context) can be used to get the unexpected token.
  * @reval P_DECODE_ERROR
  *   The decoder encountered invalid text encoding.
  * @reval P_UNEXPECTED_INPUT
@@ -860,7 +1003,11 @@ public size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * cont
     <%= @grammar.prefix %>token_t token = INVALID_TOKEN_ID;
     state_value_t[] statevalues = new state_value_t[](1);
     size_t reduced_rule_set = INVALID_ID;
+<% if @grammar.ast %>
+    void * reduced_parser_node;
+<% else %>
     <%= @grammar.prefix %>value_t reduced_parser_value;
+<% end %>
     for (;;)
     {
         if (token == INVALID_TOKEN_ID)
@@ -883,7 +1030,11 @@ public size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * cont
             if ((shift_state != INVALID_ID) && (token == TOKEN___EOF))
             {
                 /* Successful parse. */
+<% if @grammar.ast %>
+                context.parse_result = cast(<%= @grammar.ast_prefix %><%= @grammar.start_rule %><%= @grammar.ast_suffix %> *)statevalues[$-1].ast_node;
+<% else %>
                 context.parse_result = statevalues[$-1].pvalue;
+<% end %>
                 return P_SUCCESS;
             }
         }
@@ -894,15 +1045,24 @@ public size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * cont
             if (reduced_rule_set == INVALID_ID)
             {
                 /* We shifted a token, mark it consumed. */
-                token = INVALID_TOKEN_ID;
+<% if @grammar.ast %>
+                <%= @grammar.ast_prefix %>Token<%= @grammar.ast_suffix %> * token_ast_node = new <%= @grammar.ast_prefix %>Token<%= @grammar.ast_suffix %>(token_info.position, token_info.end_position, token, token_info.pvalue);
+                statevalues[$-1].ast_node = token_ast_node;
+<% else %>
                 statevalues[$-1].pvalue = token_info.pvalue;
+<% end %>
+                token = INVALID_TOKEN_ID;
             }
             else
             {
                 /* We shifted a RuleSet. */
+<% if @grammar.ast %>
+                statevalues[$-1].ast_node = reduced_parser_node;
+<% else %>
                 statevalues[$-1].pvalue = reduced_parser_value;
                 <%= @grammar.prefix %>value_t new_parse_result;
                 reduced_parser_value = new_parse_result;
+<% end %>
                 reduced_rule_set = INVALID_ID;
             }
             continue;
@@ -912,12 +1072,65 @@ public size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * cont
         if (reduce_index != INVALID_ID)
         {
             /* We have something to reduce. */
+<% if @grammar.ast %>
+            if (parser_reduce_table[reduce_index].propagate_optional_target)
+            {
+                reduced_parser_node = statevalues[$ - 1].ast_node;
+            }
+            else if (parser_reduce_table[reduce_index].n_states > 0)
+            {
+                size_t n_fields = parser_reduce_table[reduce_index].rule_set_node_field_array_size;
+                size_t node_size = ASTNode.sizeof + n_fields * (void *).sizeof;
+                ASTNode * node = cast(ASTNode *)malloc(node_size);
+                GC.addRange(node, node_size);
+                node.position = <%= @grammar.prefix %>position_t.INVALID;
+                node.end_position = <%= @grammar.prefix %>position_t.INVALID;
+                foreach (i; 0..n_fields)
+                {
+                    node.fields[i] = null;
+                }
+                if (parser_reduce_table[reduce_index].rule_set_node_field_index_map is null)
+                {
+                    foreach (i; 0..parser_reduce_table[reduce_index].n_states)
+                    {
+                        node.fields[i] = statevalues[$ - parser_reduce_table[reduce_index].n_states + i].ast_node;
+                    }
+                }
+                else
+                {
+                    foreach (i; 0..parser_reduce_table[reduce_index].n_states)
+                    {
+                        node.fields[parser_reduce_table[reduce_index].rule_set_node_field_index_map[i]] = statevalues[$ - parser_reduce_table[reduce_index].n_states + i].ast_node;
+                    }
+                }
+                bool position_found = false;
+                foreach (i; 0..n_fields)
+                {
+                    ASTNode * child = cast(ASTNode *)node.fields[i];
+                    if (child && child.position.valid)
+                    {
+                        if (!position_found)
+                        {
+                            node.position = child.position;
+                            position_found = true;
+                        }
+                        node.end_position = child.end_position;
+                    }
+                }
+                reduced_parser_node = node;
+            }
+            else
+            {
+                reduced_parser_node = null;
+            }
+<% else %>
             <%= @grammar.prefix %>value_t 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;
             }
             reduced_parser_value = reduced_parser_value2;
+<% end %>
             reduced_rule_set = parser_reduce_table[reduce_index].rule_set;
             statevalues.length -= parser_reduce_table[reduce_index].n_states;
             continue;
@@ -942,9 +1155,17 @@ public size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * cont
  *
  * @return Parse result value.
  */
+<% if @grammar.ast %>
+public <%= @grammar.ast_prefix %><%= @grammar.start_rule %><%= @grammar.ast_suffix %> * <%= @grammar.prefix %>result(<%= @grammar.prefix %>context_t * context)
+<% else %>
 public <%= start_rule_type[1] %> <%= @grammar.prefix %>result(<%= @grammar.prefix %>context_t * context)
+<% end %>
 {
+<% if @grammar.ast %>
+    return context.parse_result;
+<% else %>
     return context.parse_result.v_<%= start_rule_type[0] %>;
+<% end %>
 }
 
 /**
@@ -972,3 +1193,13 @@ public size_t <%= @grammar.prefix %>user_terminate_code(<%= @grammar.prefix %>co
 {
     return context.user_terminate_code;
 }
+
+/**
+ * Get the parse token.
+ *
+ * @return Parse token.
+ */
+public <%= @grammar.prefix %>token_t <%= @grammar.prefix %>token(<%= @grammar.prefix %>context_t * context)
+{
+    return context.token;
+}

assets/parser.h.erb

@@ -23,7 +23,7 @@
 #define <%= @grammar.prefix.upcase %>USER_TERMINATED 6u
 
 /** Token type. */
-typedef <%= get_type_for(@grammar.invalid_token_id) %> <%= @grammar.prefix %>token_t;
+typedef <%= get_type_for(@grammar.terminate_token_id) %> <%= @grammar.prefix %>token_t;
 
 /** Token IDs. */
 <% @grammar.tokens.each_with_index do |token, index| %>
@@ -33,23 +33,13 @@ typedef <%= get_type_for(@grammar.invalid_token_id) %> <%= @grammar.prefix %>tok
 <%   end %>
 <% end %>
 #define INVALID_TOKEN_ID <%= @grammar.invalid_token_id %>u
+#define TERMINATE_TOKEN_ID <%= @grammar.terminate_token_id %>u
 
 /** Code point type. */
 typedef uint32_t <%= @grammar.prefix %>code_point_t;
 
-/** User header code blocks. */
-<%= @grammar.code_blocks.fetch("header", "") %>
-
-/** Parser values type(s). */
-typedef union
-{
-<% @grammar.ptypes.each do |name, typestring| %>
-    <%= typestring %> v_<%= name %>;
-<% end %>
-} <%= @grammar.prefix %>value_t;
-
 /**
- * A structure to keep track of parser position.
+ * A structure to keep track of input position.
  *
  * This is useful for reporting errors, etc...
  */
@@ -62,12 +52,72 @@ typedef struct
     uint32_t col;
 } <%= @grammar.prefix %>position_t;
 
+/** Return whether the position is valid. */
+#define <%= @grammar.prefix %>position_valid(p) ((p).row != 0u)
+
+/** User header code blocks. */
+<%= @grammar.code_blocks.fetch("header", "") %>
+
+<% if @grammar.ast %>
+/** Parser values type. */
+typedef <%= @grammar.ptype %> <%= @grammar.prefix %>value_t;
+<% else %>
+/** Parser values type(s). */
+typedef union
+{
+<%   @grammar.ptypes.each do |name, typestring| %>
+    <%= typestring %> v_<%= name %>;
+<%   end %>
+} <%= @grammar.prefix %>value_t;
+<% end %>
+
+<% if @grammar.ast %>
+/** AST node types. @{ */
+typedef struct <%= @grammar.ast_prefix %>Token<%= @grammar.ast_suffix %>
+{
+    /* ASTNode fields must be present in the same order here. */
+    <%= @grammar.prefix %>position_t position;
+    <%= @grammar.prefix %>position_t end_position;
+    <%= @grammar.prefix %>token_t token;
+    <%= @grammar.prefix %>value_t pvalue;
+} <%= @grammar.ast_prefix %>Token<%= @grammar.ast_suffix %>;
+
+<%   @parser.rule_sets.each do |name, rule_set| %>
+<%     next if name.start_with?("$") %>
+<%     next if rule_set.optional? %>
+struct <%= name %>;
+<%   end %>
+
+<%   @parser.rule_sets.each do |name, rule_set| %>
+<%     next if name.start_with?("$") %>
+<%     next if rule_set.optional? %>
+typedef struct <%= @grammar.ast_prefix %><%= name %><%= @grammar.ast_suffix %>
+{
+    <%= @grammar.prefix %>position_t position;
+    <%= @grammar.prefix %>position_t end_position;
+<%     rule_set.ast_fields.each do |fields| %>
+    union
+    {
+<%       fields.each do |field_name, type| %>
+        struct <%= type %> * <%= field_name %>;
+<%       end %>
+    };
+<%     end %>
+} <%= @grammar.ast_prefix %><%= name %><%= @grammar.ast_suffix %>;
+
+<%   end %>
+/** @} */
+<% end %>
+
 /** Lexed token information. */
 typedef struct
 {
-    /** Text position where the token was found. */
+    /** Text position of first code point in token. */
     <%= @grammar.prefix %>position_t position;
 
+    /** Text position of last code point in token. */
+    <%= @grammar.prefix %>position_t end_position;
+
     /** Number of input bytes used by the token. */
     size_t length;
 
@@ -106,7 +156,11 @@ typedef struct
     /* Parser context data. */
 
     /** Parse result value. */
+<% if @grammar.ast %>
+    <%= @grammar.ast_prefix %><%= @grammar.start_rule %><%= @grammar.ast_suffix %> * parse_result;
+<% else %>
     <%= @grammar.prefix %>value_t parse_result;
+<% end %>
 
     /** Unexpected token received. */
     <%= @grammar.prefix %>token_t token;
@@ -115,6 +169,13 @@ typedef struct
     size_t user_terminate_code;
 } <%= @grammar.prefix %>context_t;
 
+/**************************************************************************
+ * Public data
+ *************************************************************************/
+
+/** Token names. */
+extern const char * <%= @grammar.prefix %>token_names[];
+
 void <%= @grammar.prefix %>context_init(<%= @grammar.prefix %>context_t * context, uint8_t const * input, size_t input_length);
 
 size_t <%= @grammar.prefix %>decode_code_point(uint8_t const * input, size_t input_length,
@@ -124,8 +185,14 @@ size_t <%= @grammar.prefix %>lex(<%= @grammar.prefix %>context_t * context, <%=
 
 size_t <%= @grammar.prefix %>parse(<%= @grammar.prefix %>context_t * context);
 
+<% if @grammar.ast %>
+<%= @grammar.ast_prefix %><%= @grammar.start_rule %><%= @grammar.ast_suffix %> * <%= @grammar.prefix %>result(<%= @grammar.prefix %>context_t * context);
+<% else %>
 <%= start_rule_type[1] %> <%= @grammar.prefix %>result(<%= @grammar.prefix %>context_t * context);
+<% end %>
 
 <%= @grammar.prefix %>position_t <%= @grammar.prefix %>position(<%= @grammar.prefix %>context_t * context);
 
 size_t <%= @grammar.prefix %>user_terminate_code(<%= @grammar.prefix %>context_t * context);
+
+<%= @grammar.prefix %>token_t <%= @grammar.prefix %>token(<%= @grammar.prefix %>context_t * context);

doc/user_guide.md

@@ -13,7 +13,9 @@ Propane is a LALR Parser Generator (LPG) which:
   * generates a built-in lexer to tokenize input
   * supports UTF-8 lexer inputs
   * generates a table-driven shift/reduce parser to parse input in linear time
-  * targets C or D language outputs
+  * targets C, C++, or D language outputs
+  * optionally supports automatic full AST generation
+  * tracks input text start and end positions for all matched tokens/rules
   * is MIT-licensed
   * is distributable as a standalone Ruby script
 
@@ -34,9 +36,14 @@ Propane is typically invoked from the command-line as `./propane`.
 
     Usage: ./propane [options] <input-file> <output-file>
     Options:
-      --log LOG   Write log file
-      --version   Show program version and exit
-      -h, --help  Show this usage and exit
+      -h, --help  Show this usage and exit.
+      --log LOG   Write log file. This will show all parser states and their
+                  associated shifts and reduces. It can be helpful when
+                  debugging a grammar.
+      --version   Show program version and exit.
+      -w          Treat warnings as errors. This option will treat shift/reduce
+                  conflicts as fatal errors and will print them to stderr in
+                  addition to the log file.
 
 The user must specify the path to a Propane input grammar file and a path to an
 output file.
@@ -60,10 +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)
   {
@@ -72,38 +79,20 @@ token integer /\\d+/ <<
   }
   $$ = v;
 >>
-token lparen /\\(/;
-token rparen /\\)/;
+token lparen /\(/;
+token rparen /\)/;
 # Drop whitespace.
-drop /\\s+/;
+drop /\s+/;
 
-Start -> E1 <<
-  $$ = $1;
->>
-E1 -> E2 <<
-  $$ = $1;
->>
-E1 -> E1 plus E2 <<
-  $$ = $1 + $3;
->>
-E2 -> E3 <<
-  $$ = $1;
->>
-E2 -> E2 times E3 <<
-  $$ = $1 * $3;
->>
-E3 -> E4 <<
-  $$ = $1;
->>
-E3 -> E3 power E4 <<
-  $$ = pow($1, $3);
->>
-E4 -> integer <<
-  $$ = $1;
->>
-E4 -> lparen E1 rparen <<
-  $$ = $2;
->>
+Start -> E1 << $$ = $1; >>
+E1 -> E2 << $$ = $1; >>
+E1 -> E1 plus E2 << $$ = $1 + $3; >>
+E2 -> E3 << $$ = $1; >>
+E2 -> E2 times E3 << $$ = $1 * $3; >>
+E3 -> E4 << $$ = $1; >>
+E3 -> E3 power E4 << $$ = pow($1, $3); >>
+E4 -> integer << $$ = $1; >>
+E4 -> lparen E1 rparen << $$ = $2; >>
 ```
 
 Grammar files can contain comment lines beginning with `#` which are ignored.
@@ -117,8 +106,8 @@ lowercase character and beginning a rule name with an uppercase character.
 
 ##> User Code Blocks
 
-User code blocks begin with the line following a "<<" token and end with the
-line preceding a grammar line consisting of solely the ">>" token.
+User code blocks begin following a "<<" token and end with a ">>" token found
+at the end of a line.
 All text lines in the code block are copied verbatim into the output file.
 
 ### Standalone Code Blocks
@@ -166,7 +155,7 @@ Example:
 ```
 ptype ulong;
 
-token integer /\\d+/ <<
+token integer /\d+/ <<
   ulong v;
   foreach (c; match)
   {
@@ -189,9 +178,7 @@ This parser value can then be used later in a parser rule.
 Example:
 
 ```
-E1 -> E1 plus E2 <<
-  $$ = $1 + $3;
->>
+E1 -> E1 plus E2 << $$ = $1 + $3; >>
 ```
 
 Parser rule code blocks appear following a rule expression.
@@ -202,6 +189,143 @@ rule.
 Parser values for the rules or tokens in the rule pattern can be accessed
 positionally with tokens `$1`, `$2`, `$3`, etc...
 
+Parser rule code blocks are not available in AST generation mode.
+In AST generation mode, a full parse tree is automatically constructed in
+memory for user code to traverse after parsing is complete.
+
+##> AST generation mode - the `ast` statement
+
+To activate AST generation mode, place the `ast` statement in your grammar file:
+
+```
+ast;
+```
+
+It is recommended to place this statement early in the grammar.
+
+In AST generation mode various aspects of propane's behavior are changed:
+
+  * Only one `ptype` is allowed.
+  * Parser user code blocks are not supported.
+  * Structure types are generated to represent the parsed tokens and rules as
+  defined in the grammar.
+  * The parse result from `p_result()` points to a `Start` struct containing
+  the entire parse tree for the input. If the user has changed the start rule
+  with the `start` grammar statement, the name of the start struct will be
+  given by the user-specified start rule instead of `Start`.
+
+Example AST generation grammar:
+
+```
+ast;
+
+ptype int;
+
+token a << $$ = 11; >>
+token b << $$ = 22; >>
+token one /1/;
+token two /2/;
+token comma /,/ <<
+  $$ = 42;
+>>
+token lparen /\(/;
+token rparen /\)/;
+drop /\s+/;
+
+Start -> Items;
+
+Items -> Item:item ItemsMore;
+Items -> ;
+
+ItemsMore -> comma Item:item ItemsMore;
+ItemsMore -> ;
+
+Item -> a;
+Item -> b;
+Item -> lparen Item:item rparen;
+Item -> Dual;
+
+Dual -> One Two;
+Dual -> Two One;
+One -> one;
+Two -> two;
+```
+
+The following unit test describes the fields that will be present for an
+example parse:
+
+```
+string input = "a, ((b)), b";
+p_context_t context;
+p_context_init(&context, input);
+assert_eq(P_SUCCESS, p_parse(&context));
+Start * start = p_result(&context);
+assert(start.pItems1 !is null);
+assert(start.pItems !is null);
+Items * items = start.pItems;
+assert(items.item !is null);
+assert(items.item.pToken1 !is null);
+assert_eq(TOKEN_a, items.item.pToken1.token);
+assert_eq(11, items.item.pToken1.pvalue);
+assert(items.pItemsMore !is null);
+ItemsMore * itemsmore = items.pItemsMore;
+assert(itemsmore.item !is null);
+assert(itemsmore.item.item !is null);
+assert(itemsmore.item.item.item !is null);
+assert(itemsmore.item.item.item.pToken1 !is null);
+assert_eq(TOKEN_b, itemsmore.item.item.item.pToken1.token);
+assert_eq(22, itemsmore.item.item.item.pToken1.pvalue);
+assert(itemsmore.pItemsMore !is null);
+itemsmore = itemsmore.pItemsMore;
+assert(itemsmore.item !is null);
+assert(itemsmore.item.pToken1 !is null);
+assert_eq(TOKEN_b, itemsmore.item.pToken1.token);
+assert_eq(22, itemsmore.item.pToken1.pvalue);
+assert(itemsmore.pItemsMore is null);
+```
+
+## `ast_prefix` and `ast_suffix` statements
+
+In AST generation mode, structure types are defined and named based on the
+rules in the grammar.
+Additionally, a structure type called `Token` is generated to hold parsed
+token information.
+
+These structure names can be modified by using the `ast_prefix` or `ast_suffix`
+statements in the grammar file.
+The field names that point to instances of the structures are not affected by
+the `ast_prefix` or `ast_suffix` values.
+
+For example, if the following two lines were added to the example above:
+
+```
+ast_prefix ABC;
+ast_suffix XYZ;
+```
+
+Then the types would be used as such instead:
+
+```
+string input = "a, ((b)), b";
+p_context_t context;
+p_context_init(&context, input);
+assert_eq(P_SUCCESS, p_parse(&context));
+ABCStartXYZ * start = p_result(&context);
+assert(start.pItems1 !is null);
+assert(start.pItems !is null);
+ABCItemsXYZ * items = start.pItems;
+assert(items.pItem !is null);
+assert(items.pItem.pToken1 !is null);
+assert_eq(TOKEN_a, items.pItem.pToken1.token);
+assert_eq(11, items.pItem.pToken1.pvalue);
+assert(items.pItemsMore !is null);
+ABCItemsMoreXYZ * itemsmore = items.pItemsMore;
+assert(itemsmore.pItem !is null);
+assert(itemsmore.pItem.pItem !is null);
+assert(itemsmore.pItem.pItem.pItem !is null);
+assert(itemsmore.pItem.pItem.pItem.pToken1 !is null);
+```
+
 ##> Specifying tokens - the `token` statement
 
 The `token` statement allows defining a lexer token and a pattern to match that
@@ -238,9 +362,7 @@ lexer.
 Example:
 
 ```
-token if <<
-  writeln("'if' keyword lexed");
->>
+token if << writeln("'if' keyword lexed"); >>
 ```
 
 The `token` statement is actually a shortcut statement for a combination of a
@@ -277,9 +399,7 @@ code but may not result in a matched token.
 Example:
 
 ```
-/foo+/ <<
-  writeln("saw a foo pattern");
->>
+/foo+/ << writeln("saw a foo pattern"); >>
 ```
 
 This can be especially useful with ${#Lexer modes}.
@@ -312,10 +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.
@@ -325,9 +445,19 @@ Regular expressions can include many special characters:
   * The `(` character begins a matching group.
   * The `{` character begins a count qualifier.
   * The `\` character escapes the following character and changes its meaning:
-    * The `\d` sequence matches any character `0` through `9`.
-    * The `\s` sequence matches a space, horizontal tab `\t`, carriage return
-    `\r`, a form feed `\f`, or a vertical tab `\v` character.
+    * The `\a` sequence matches an ASCII bell character (0x07).
+    * The `\b` sequence matches an ASCII backspace character (0x08).
+    * 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 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.
 
@@ -381,9 +511,7 @@ tokenid str;
   mystringvalue = "";
   $mode(string);
 >>
-string: /[^"]+/ <<
-  mystringvalue += match;
->>
+string: /[^"]+/ << mystringvalue ~= match; >>
 string: /"/ <<
   $mode(default);
   return $token(str);
@@ -411,6 +539,28 @@ It also returns the `str` token now that the token is complete.
 Note that the token name `str` above could have been `string` instead - the
 namespace for token names is distinct from the namespace for lexer modes.
 
+Multiple modes can be specified for a token or pattern or drop statement.
+For example, if the grammar wanted to only recognize an identifier following
+a `.` token and not other keywords, it could switch to an `identonly` mode
+when matching a `.`
+The `ident` token pattern will be matched in either the `default` or
+`identonly` mode.
+
+```
+ptype char;
+token abc;
+token def;
+default, identonly: token ident /[a-z]+/ <<
+  $$ = match[0];
+  $mode(default);
+  return $token(ident);
+>>
+token dot /\./ <<
+  $mode(identonly);
+>>
+default, identonly: drop /\s+/;
+```
+
 ##> Specifying parser value types - the `ptype` statement
 
 The `ptype` statement is used to define parser value type(s).
@@ -440,20 +590,12 @@ ptype Value;
 ptype array = Value[];
 ptype dict = Value[string];
 
-Object -> lbrace rbrace <<
-  $$ = new Value();
->>
+Object -> lbrace rbrace << $$ = new Value(); >>
 
-Values (array) -> Value <<
-  $$ = [$1];
->>
-Values -> Values comma Value <<
-  $$ = $1 ~ [$3];
->>
+Values (array) -> Value << $$ = [$1]; >>
+Values -> Values comma Value << $$ = $1 ~ [$3]; >>
 
-KeyValue (dict) -> string colon Value <<
-  $$ = [$1: $3];
->>
+KeyValue (dict) -> string colon Value << $$ = [$1: $3]; >>
 ```
 
 In this example, the default parser value type is `Value`.
@@ -469,6 +611,12 @@ In this example:
   * a reduced `Values`'s parser value has a type of `Value[]`.
   * a reduced `KeyValue`'s parser value has a type of `Value[string]`.
 
+When AST generation mode is active, the `ptype` functionality works differently.
+In this mode, only one `ptype` is used by the parser.
+Lexer user code blocks may assign a parse value to the generated `Token` node
+by assigning to `$$` within a lexer code block.
+The type of the parse value `$$` is given by the global `ptype` type.
+
 ##> Specifying a parser rule - the rule statement
 
 Rule statements create parser rules which define the grammar that will be
@@ -479,66 +627,87 @@ Rules with the same name define a rule set for that name and act as
 alternatives that the parser can accept when attempting to match a reference to
 that rule.
 
-The grammar file must define a rule with the name `Start` which will be used as
-the top-level starting rule that the parser attempts to reduce.
+The default start rule name is `Start`.
+This can be changed with the `start` statement.
+The grammar file must define a rule with the name of the start rule name which
+will be used as the top-level starting rule that the parser attempts to reduce.
+
+Rule statements are composed of the name of the rule, a `->` token, the fields
+defining the rule pattern that must be matched, and a terminating semicolon or
+user code block.
 
 Example:
 
 ```
 ptype ulong;
-token word /[a-z]+/ <<
-  $$ = match.length;
->>
-Start -> word <<
-  $$ = $1;
->>
+start Top;
+token word /[a-z]+/ << $$ = match.length; >>
+Top -> word << $$ = $1; >>
 ```
 
-In the above example the `Start` rule is defined to match a single `word`
+In the above example the `Top` rule is defined to match a single `word`
 token.
 
+Another example:
+
+```
+Start -> E1 << $$ = $1; >>
+E1 -> E2 << $$ = $1; >>
+E1 -> E1 plus E2 << $$ = $1 + $3; >>
+E2 -> E3 << $$ = $1; >>
+E2 -> E2 times E3 << $$ = $1 * $3; >>
+E3 -> E4 << $$ = $1; >>
+E3 -> E3 power E4 << $$ = pow($1, $3); >>
+E4 -> integer << $$ = $1; >>
+E4 -> lparen E1 rparen << $$ = $2; >>
+```
+
+This example uses the default start rule name of `Start`.
+
+A parser rule has zero or more fields on the right side of its definition.
+Each of these fields is either a token name or a rule name.
+A field can be immediately followed by a `?` character to signify that it is
+optional.
+A field can optionally be followed by a `:` and then a field alias name.
+If present, the field alias name is used to refer to the field value in user
+code blocks, or if AST mode is active, the field alias name is used as the
+field name in the generated AST node structure.
+An optional and named field must use the format `field?:name`.
 Example:
 
 ```
-Start -> E1 <<
-  $$ = $1;
->>
-E1 -> E2 <<
-  $$ = $1;
->>
-E1 -> E1 plus E2 <<
-  $$ = $1 + $3;
->>
-E2 -> E3 <<
-  $$ = $1;
->>
-E2 -> E2 times E3 <<
-  $$ = $1 * $3;
->>
-E3 -> E4 <<
-  $$ = $1;
->>
-E3 -> E3 power E4 <<
-  $$ = pow($1, $3);
->>
-E4 -> integer <<
-  $$ = $1;
->>
-E4 -> lparen E1 rparen <<
-  $$ = $2;
->>
+token public;
+token private;
+token int;
+token ident /[a-zA-Z_][a-zA-Z_0-9]*/;
+token semicolon /;/;
+IntegerDeclaration -> Visibility?:visibility int ident:name semicolon;
+Visibility -> public;
+Visibility -> private;
 ```
 
-A parser rule has zero or more terms on the right side of its definition.
-Each of these terms is either a token name or a rule name.
-
-In a parser rule code block, parser values for the right side terms are
-accessible as `$1` for the first term's parser value, `$2` for the second
-term's parser value, etc...
+In a parser rule code block, parser values for the right side fields are
+accessible as `$1` for the first field's parser value, `$2` for the second
+field's parser value, etc...
+For the `IntegerDeclaration` rule, the first field value can also be referred to as `${visibility}` and the third field value can also be referred
+to as `${name}`.
 The `$$` symbol accesses the output parser value for this rule.
 The above examples demonstrate how the parser values for the rule components
 can be used to produce the parser value for the accepted rule.
 
+Parser rule code blocks are not allowed and not used when AST generation mode
+is active.
+
+##> Specifying the parser start rule name - the `start` statement
+
+The start rule can be changed from the default of `Start` by using the `start`
+statement.
+Example:
+
+```
+start MyStartRule;
+```
+
 ##> Specifying the parser module name - the `module` statement
 
 The `module` statement can be used to specify the module name for a generated
@@ -574,23 +743,22 @@ default.
 It can also be used when generating multiple lexers/parsers to be used in the
 same program to avoid symbol collisions.
 
-##> User termination of the parser
+##> User termination of the lexer or parser
 
-Propane supports allowing parser user code blocks to terminate execution of the
-parser.
-One example use of this functionality is to detect and report an error before
-the parser continues parsing the remainder of the input.
-Another use of this features is to begin parsing input and determine whether a
-different parser should be used instead.
+Propane supports allowing lexer or parser user code blocks to terminate
+execution of the parser.
+Some example uses of this functionality could be to:
 
-To terminate parsing from a parser user code block, use the `$terminate(code)`
-function, passing an integer expression argument.
+  * Detect integer overflow when lexing an integer literal constant.
+  * Detect and report an error as soon as possible during parsing before continuing to parse any more of the input.
+  * Determine whether parsing should stop and instead be retried using a different parser version.
+
+To terminate parsing from a lexer or parser user code block, use the
+`$terminate(code)` function, passing an integer expression argument.
 For example:
 
 ```
-NewExpression -> new Expression <<
-  $terminate(42);
->>
+NewExpression -> new Expression << $terminate(42); >>
 ```
 
 The value passed to the `$terminate()` function is known as the "user terminate
@@ -619,10 +787,16 @@ 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_input()`, `p_lex()`, and `p_parse()`.
+Result codes are returned by the API functions `p_decode_code_point()`, `p_lex()`, and `p_parse()`.
 
 ##> Types
 
+### `p_code_point_t`
+
+The `p_code_point_t` type is aliased to a 32-bit unsigned integer.
+It is used to store decoded code points from the input text and perform
+lexing based on the grammar's lexer patterns.
+
 ### `p_context_t`
 
 Propane defines a `p_context_t` structure type.
@@ -633,8 +807,112 @@ 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`.
-These fields contain the 0-based row and column describing a parser position.
+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
+querying the `valid` property.
+
+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
+generated.
+The name of the structure type is given by the name of the rule.
+Additionally a structure type called `Token` is generated to represent an
+AST node which refers to a raw parser token rather than a composite rule.
+
+#### AST Node Fields
+
+All AST nodes have a `position` field specifying the text position of the
+beginning of the matched token or rule, and an `end_position` field specifying
+the text position of the end of the matched token or rule.
+Each of these fields are instances of the `p_position_t` structure.
+
+A `Token` node will always have a valid `position` and `end_position`.
+A rule node may not have valid positions if the rule allows for an empty match.
+In this case the `position` structure should be checked for validity before
+using it.
+For C targets this can be accomplished with
+`if (p_position_valid(node->position))` and for D targets this can be
+accomplished with `if (node.position.valid)`.
+
+A `Token` node has the following additional fields:
+
+  * `token` which specifies which token was parsed (one of `TOKEN_*`)
+  * `pvalue` which specifies the parser value for the token. If a lexer user
+  code block assigned to `$$`, the assigned value will be stored here.
+
+AST node structures for rules contain generated fields based on the
+right hand side components specified for all rules of a given name.
+
+In this example:
+
+```
+Start -> Items;
+
+Items -> Item ItemsMore;
+Items -> ;
+```
+
+The `Start` structure will have a field called `pItems` and another field of
+the same name but with a positional suffix (`pItems1`) which both point to the
+parsed `Items` node.
+Their value will be null if the parsed `Items` rule was empty.
+
+The `Items` structure will have fields:
+
+  * `pItem` and `pItem1` which point to the parsed `Item` structure.
+  * `pItemsMore` and `pItemsMore2` which point to the parsed `ItemsMore` structure.
+
+If a rule can be empty (for example in the second `Items` rule above), then
+an instance of a pointer to that rule's generated AST node will be null if the
+parser matches the empty rule pattern.
+
+The non-positional AST node field pointer will not be generated if there are
+multiple positions in which an instance of the node it points to could be
+present.
+For example, in the below rules:
+
+```
+Dual -> One Two;
+Dual -> Two One;
+```
+
+The generated `Dual` structure will contain `pOne1`, `pTwo2`, `pTwo1`, and
+`pOne2` fields.
+However, a `pOne` field and `pTwo` field will not be generated since it would
+be ambiguous which one was matched.
+
+If the first rule is matched, then `pOne1` and `pTwo2` will be non-null while
+`pTwo1` and `pOne2` will be null.
+If the second rule is matched instead, then the opposite would be the case.
+
+If a field alias is present in a rule definition, an additional field will be
+generated in the AST node with the field alias name.
+For example:
+
+```
+Exp -> Exp:left plus ExpB:right;
+```
+
+In the generated `Exp` structure, the fields `pExp`, `pExp1`, and `left` will
+all point to the same child node (an instance of the `Exp` structure), and the
+fields `pExpB`, `pExpB3`, and `right` will all point to the same child node
+(an instance of the `ExpB` structure).
 
 ##> Functions
 
@@ -645,11 +923,56 @@ structure.
 The input to be used for lexing/parsing is passed in when initializing the
 context structure.
 
+C example:
+
+```
+p_context_t context;
+p_context_init(&context, input, input_length);
+```
+
+D example:
+
+```
+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`
@@ -665,6 +988,24 @@ p_context_init(&context, input, input_length);
 size_t result = p_parse(&context);
 ```
 
+### `p_position_valid`
+
+The `p_position_valid()` function is only generated for C targets.
+it is used to determine whether or not a `p_position_t` structure is valid.
+
+Example:
+
+```
+if (p_position_valid(node->position))
+{
+    ....
+}
+```
+
+For D targets, rather than using `p_position_valid()`, the `valid` property
+function of the `p_position_t` structure can be queried
+(e.g. `if (node.position.valid)`).
+
 ### `p_result`
 
 The `p_result()` function can be used to retrieve the final parse value after
@@ -682,6 +1023,9 @@ if (p_parse(&context) == P_SUCCESS)
 }
 ```
 
+If AST generation mode is active, then the `p_result()` function returns a
+`Start *` pointing to the `Start` AST structure.
+
 ### `p_position`
 
 The `p_position()` function can be used to retrieve the parser position where
@@ -718,6 +1062,64 @@ if (p_parse(&context) == P_USER_TERMINATED)
 }
 ```
 
+### `p_token`
+
+The `p_token()` function can be used to retrieve the current parse token.
+This is useful after `p_parse()` returns a `P_UNEXPECTED_TOKEN` value.
+terminate code after `p_parse()` returns a `P_USER_TERMINATED` value to
+indicate what token the parser was not expecting.
+
+Example:
+
+```
+if (p_parse(&context) == P_UNEXPECTED_TOKEN)
+{
+    p_token_t unexpected_token = p_token(&context);
+}
+```
+
+### `p_decode_code_point`
+
+The `p_decode_code_point()` function can be used to decode code points from a
+UTF-8 string.
+It does not require a lexer/parser context structure and can be used as a
+standalone UTF-8 decoder or from within a lexer or parser user code block.
+
+D Example:
+
+```
+size_t result;
+p_code_point_t code_point;
+ubyte code_point_length;
+
+result = p_decode_code_point("\xf0\x9f\xa7\xa1", &code_point, &code_point_length);
+assert(result == P_SUCCESS);
+assert(code_point == 0x1F9E1u);
+assert(code_point_length == 4u);
+```
+
+##> Data
+
+### `p_token_names`
+
+The `p_token_names` array contains the grammar-specified token names.
+It is indexed by the token ID.
+
+C example:
+
+```
+p_context_t context;
+p_context_init(&context, input, input_length);
+size_t result = p_parse(&context);
+if (p_parse(&context) == P_UNEXPECTED_TOKEN)
+{
+    p_position_t error_position = p_position(&context);
+    fprintf(stderr, "Error: unexpected token `%s' at row %u column %u\n",
+        p_token_names[context.token],
+        error_position.row + 1, error_position.col + 1);
+}
+```
+
 #> License
 
 Propane is licensed under the terms of the MIT License:

extra/vim/ftdetect/propane.vim

@@ -0,0 +1 @@
+au BufNewFile,BufRead *.propane set filetype=propane

extra/vim/syntax/propane.vim

@@ -0,0 +1,33 @@
+" Vim syntax file for Propane
+" Language: propane
+" Maintainer: Josh Holtrop
+" URL: https://github.com/holtrop/propane
+
+if exists("b:current_syntax")
+  finish
+endif
+
+if !exists("b:propane_subtype")
+  let b:propane_subtype = "d"
+endif
+
+exe "syn include @propaneTarget syntax/".b:propane_subtype.".vim"
+
+syn region propaneTarget matchgroup=propaneDelimiter start="<<" end=">>$" contains=@propaneTarget keepend
+
+syn match propaneComment "#.*"
+syn match propaneOperator "->"
+syn match propaneFieldAlias ":[a-zA-Z0-9_]\+" contains=propaneFieldOperator
+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="\v\\\\|\\/"
+
+hi def link propaneComment Comment
+hi def link propaneKeyword Keyword
+hi def link propaneRegex String
+hi def link propaneOperator Operator
+hi def link propaneFieldOperator Operator
+hi def link propaneDelimiter Delimiter
+hi def link propaneFieldAlias Identifier

lib/propane.rb

@@ -31,10 +31,10 @@ class Propane
 
   class << self
 
-    def run(input_file, output_file, log_file)
+    def run(input_file, output_file, log_file, options)
       begin
         grammar = Grammar.new(File.read(input_file))
-        generator = Generator.new(grammar, output_file, log_file)
+        generator = Generator.new(grammar, output_file, log_file, options)
         generator.generate
       rescue Error => e
         $stderr.puts e.message

lib/propane/cli.rb

@@ -4,15 +4,21 @@ class Propane
     USAGE = <<EOF
 Usage: #{$0} [options] <input-file> <output-file>
 Options:
-  --log LOG   Write log file
-  --version   Show program version and exit
-  -h, --help  Show this usage and exit
+  -h, --help  Show this usage and exit.
+  --log LOG   Write log file. This will show all parser states and their
+              associated shifts and reduces. It can be helpful when
+              debugging a grammar.
+  --version   Show program version and exit.
+  -w          Treat warnings as errors. This option will treat shift/reduce
+              conflicts as fatal errors and will print them to stderr in
+              addition to the log file.
 EOF
 
     class << self
 
       def run(args)
         params = []
+        options = {}
         log_file = nil
         i = 0
         while i < args.size
@@ -24,11 +30,13 @@ EOF
               log_file = args[i]
             end
           when "--version"
-            puts "propane v#{VERSION}"
+            puts "propane version #{VERSION}"
             return 0
           when "-h", "--help"
             puts USAGE
             return 0
+          when "-w"
+            options[:warnings_as_errors] = true
           when /^-/
             $stderr.puts "Error: unknown option #{arg}"
             return 1
@@ -45,7 +53,7 @@ EOF
           $stderr.puts "Error: cannot read #{params[0]}"
           return 2
         end
-        Propane.run(*params, log_file)
+        Propane.run(*params, log_file, options)
       end
 
     end

lib/propane/generator.rb

@@ -2,7 +2,7 @@ class Propane
 
   class Generator
 
-    def initialize(grammar, output_file, log_file)
+    def initialize(grammar, output_file, log_file, options)
       @grammar = grammar
       @output_file = output_file
       if log_file
@@ -11,23 +11,28 @@ class Propane
         @log = StringIO.new
       end
       @language =
-        if output_file =~ /\.([a-z]+)$/
-          $1
-        else
+        if output_file.end_with?(".d")
           "d"
+        else
+          "c"
         end
+      @options = options
       process_grammar!
     end
 
     def generate
-      extensions = [@language]
+      extensions = [nil]
       if @language == "c"
         extensions += %w[h]
       end
       extensions.each do |extension|
-        template = Assets.get("parser.#{extension}.erb")
+        template = Assets.get("parser.#{extension || @language}.erb")
+        if extension
+          output_file = @output_file.sub(%r{\.[a-z]+$}, ".#{extension}")
+        else
+          output_file = @output_file
+        end
         erb = ERB.new(template, trim_mode: "<>")
-        output_file = @output_file.sub(%r{\.[a-z]+$}, ".#{extension}")
         result = erb.result(binding.clone)
         File.open(output_file, "wb") do |fh|
           fh.write(result)
@@ -42,8 +47,8 @@ class Propane
       # Assign default pattern mode to patterns without a mode assigned.
       found_default = false
       @grammar.patterns.each do |pattern|
-        if pattern.mode.nil?
-          pattern.mode = "default"
+        if pattern.modes.empty?
+          pattern.modes << "default"
           found_default = true
         end
         pattern.ptypename ||= "default"
@@ -51,6 +56,7 @@ class Propane
       unless found_default
         raise Error.new("No patterns found for default mode")
       end
+      check_ptypes!
       # Add EOF token.
       @grammar.tokens << Token.new("$EOF", nil, nil)
       tokens_by_name = {}
@@ -66,11 +72,14 @@ class Propane
         tokens_by_name[token.name] = token
       end
       # Check for user start rule.
-      unless @grammar.rules.find {|rule| rule.name == "Start"}
-        raise Error.new("Start rule not found")
+      unless @grammar.rules.find {|rule| rule.name == @grammar.start_rule}
+        raise Error.new("Start rule `#{@grammar.start_rule}` not found")
       end
       # Add "real" start rule.
-      @grammar.rules.unshift(Rule.new("$Start", ["Start", "$EOF"], nil, nil, nil))
+      @grammar.rules.unshift(Rule.new("$Start", [@grammar.start_rule, "$EOF"], nil, nil, nil))
+      # Generate and add rules for optional components.
+      generate_optional_component_rules!(tokens_by_name)
+      # Build rule sets.
       rule_sets = {}
       rule_set_id = @grammar.tokens.size
       @grammar.rules.each_with_index do |rule, rule_id|
@@ -119,10 +128,55 @@ class Propane
         end
       end
       determine_possibly_empty_rulesets!(rule_sets)
+      rule_sets.each do |name, rule_set|
+        rule_set.finalize(@grammar)
+      end
       # Generate the lexer.
       @lexer = Lexer.new(@grammar)
       # Generate the parser.
-      @parser = Parser.new(@grammar, rule_sets, @log)
+      @parser = Parser.new(@grammar, rule_sets, @log, @options)
+    end
+
+    # Check that any referenced ptypes have been defined.
+    def check_ptypes!
+      (@grammar.patterns + @grammar.tokens + @grammar.rules).each do |potor|
+        if potor.ptypename
+          unless @grammar.ptypes.include?(potor.ptypename)
+            raise Error.new("Error: Line #{potor.line_number}: ptype #{potor.ptypename} not declared. Declare with `ptype` statement.")
+          end
+        end
+      end
+    end
+
+    # Generate and add rules for any optional components.
+    def generate_optional_component_rules!(tokens_by_name)
+      optional_rules_added = Set.new
+      @grammar.rules.each do |rule|
+        rule.components.each do |component|
+          if component =~ /^(.*)\?$/
+            c = $1
+            unless optional_rules_added.include?(component)
+              # Create two rules for the optional component: one empty and
+              # one just matching the component.
+              # We need to find the ptypename for the optional component in
+              # order to copy it to the generated rules.
+              if tokens_by_name[c]
+                # The optional component is a token.
+                ptypename = tokens_by_name[c].ptypename
+              else
+                # The optional component must be a rule, so find any instance
+                # of that rule that specifies a ptypename.
+                ptypename = @grammar.rules.reduce(nil) do |result, rule|
+                  rule.name == c && rule.ptypename ? rule.ptypename : result
+                end
+              end
+              @grammar.rules << Rule.new(component, [], nil, ptypename, rule.line_number)
+              @grammar.rules << Rule.new(component, [c], "$$ = $1;\n", ptypename, rule.line_number)
+              optional_rules_added << component
+            end
+          end
+        end
+      end
     end
 
     # Determine which grammar rules could expand to empty sequences.
@@ -198,6 +252,16 @@ class Propane
       code = code.gsub(/\$token\(([$\w]+)\)/) do |match|
         "TOKEN_#{Token.code_name($1)}"
       end
+      code = code.gsub(/\$terminate\((.*)\);/) do |match|
+        user_terminate_code = $1
+        retval = rule ? "P_USER_TERMINATED" : "TERMINATE_TOKEN_ID"
+        case @language
+        when "c"
+          "context->user_terminate_code = (#{user_terminate_code}); return #{retval};"
+        when "d"
+          "context.user_terminate_code = (#{user_terminate_code}); return #{retval};"
+        end
+      end
       if parser
         code = code.gsub(/\$\$/) do |match|
           case @language
@@ -216,22 +280,35 @@ class Propane
             "statevalues[$-1-n_states+#{index}].pvalue.v_#{rule.components[index - 1].ptypename}"
           end
         end
-        code = code.gsub(/\$terminate\((.*)\);/) do |match|
-          user_terminate_code = $1
-          case @language
-          when "c"
-            "context->user_terminate_code = (#{user_terminate_code}); return P_USER_TERMINATED;"
-          when "d"
-            "context.user_terminate_code = (#{user_terminate_code}); return P_USER_TERMINATED;"
+        code = code.gsub(/\$\{(\w+)\}/) do |match|
+          aliasname = $1
+          if index = rule.aliases[aliasname]
+            case @language
+            when "c"
+              "state_values_stack_index(statevalues, -(int)n_states + #{index})->pvalue.v_#{rule.components[index].ptypename}"
+            when "d"
+              "statevalues[$-n_states+#{index}].pvalue.v_#{rule.components[index].ptypename}"
+            end
+          else
+            raise Error.new("Field alias '#{aliasname}' not found")
           end
         end
       else
         code = code.gsub(/\$\$/) do |match|
-          case @language
-          when "c"
-            "out_token_info->pvalue.v_#{pattern.ptypename}"
-          when "d"
-            "out_token_info.pvalue.v_#{pattern.ptypename}"
+          if @grammar.ast
+            case @language
+            when "c"
+              "out_token_info->pvalue"
+            when "d"
+              "out_token_info.pvalue"
+            end
+          else
+            case @language
+            when "c"
+              "out_token_info->pvalue.v_#{pattern.ptypename}"
+            when "d"
+              "out_token_info.pvalue.v_#{pattern.ptypename}"
+            end
           end
         end
         code = code.gsub(/\$mode\(([a-zA-Z_][a-zA-Z_0-9]*)\)/) do |match|
@@ -257,7 +334,7 @@ class Propane
     #   Start rule parser value type name and type string.
     def start_rule_type
       start_rule = @grammar.rules.find do |rule|
-        rule.name == "Start"
+        rule.name == @grammar.start_rule
       end
       [start_rule.ptypename, @grammar.ptypes[start_rule.ptypename]]
     end

lib/propane/grammar.rb

@@ -5,9 +5,13 @@ class Propane
     # Reserve identifiers beginning with a double-underscore for internal use.
     IDENTIFIER_REGEX = /(?:[a-zA-Z]|_[a-zA-Z0-9])[a-zA-Z_0-9]*/
 
+    attr_reader :ast
+    attr_reader :ast_prefix
+    attr_reader :ast_suffix
     attr_reader :modulename
     attr_reader :patterns
     attr_reader :rules
+    attr_reader :start_rule
     attr_reader :tokens
     attr_reader :code_blocks
     attr_reader :ptypes
@@ -15,15 +19,19 @@ class Propane
 
     def initialize(input)
       @patterns = []
+      @start_rule = "Start"
       @tokens = []
       @rules = []
       @code_blocks = {}
       @line_number = 1
       @next_line_number = @line_number
-      @mode = nil
+      @modeline = nil
       @input = input.gsub("\r\n", "\n")
       @ptypes = {"default" => "void *"}
       @prefix = "p_"
+      @ast = false
+      @ast_prefix = ""
+      @ast_suffix = ""
       parse_grammar!
     end
 
@@ -35,6 +43,10 @@ class Propane
       @tokens.size
     end
 
+    def terminate_token_id
+      @tokens.size + 1
+    end
+
     private
 
     def parse_grammar!
@@ -46,10 +58,14 @@ class Propane
     def parse_statement!
       if parse_white_space!
       elsif parse_comment_line!
-      elsif @mode.nil? && parse_mode_label!
+      elsif @modeline.nil? && parse_mode_label!
+      elsif parse_ast_statement!
+      elsif parse_ast_prefix_statement!
+      elsif parse_ast_suffix_statement!
       elsif parse_module_statement!
       elsif parse_ptype_statement!
       elsif parse_pattern_statement!
+      elsif parse_start_statement!
       elsif parse_token_statement!
       elsif parse_tokenid_statement!
       elsif parse_drop_statement!
@@ -65,8 +81,8 @@ class Propane
     end
 
     def parse_mode_label!
-      if md = consume!(/(#{IDENTIFIER_REGEX})\s*:/)
-        @mode = md[1]
+      if md = consume!(/(#{IDENTIFIER_REGEX}(?:\s*,\s*#{IDENTIFIER_REGEX})*)\s*:/)
+        @modeline = md[1]
       end
     end
 
@@ -78,12 +94,30 @@ class Propane
       consume!(/#.*\n/)
     end
 
+    def parse_ast_statement!
+      if consume!(/ast\s*;/)
+        @ast = true
+      end
+    end
+
+    def parse_ast_prefix_statement!
+      if md = consume!(/ast_prefix\s+(\w+)\s*;/)
+        @ast_prefix = md[1]
+      end
+    end
+
+    def parse_ast_suffix_statement!
+      if md = consume!(/ast_suffix\s+(\w+)\s*;/)
+        @ast_suffix = md[1]
+      end
+    end
+
     def parse_module_statement!
       if consume!(/module\s+/)
         md = consume!(/([\w.]+)\s*/, "expected module name")
         @modulename = md[1]
         consume!(/;/, "expected `;'")
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -92,6 +126,9 @@ class Propane
       if consume!(/ptype\s+/)
         name = "default"
         if md = consume!(/(#{IDENTIFIER_REGEX})\s*=\s*/)
+          if @ast
+            raise Error.new("Multiple ptypes are unsupported in AST mode")
+          end
           name = md[1]
         end
         md = consume!(/([^;]+);/, "expected parser result type expression")
@@ -104,18 +141,21 @@ class Propane
         md = consume!(/(#{IDENTIFIER_REGEX})\s*/, "expected token name")
         name = md[1]
         if md = consume!(/\((#{IDENTIFIER_REGEX})\)\s*/)
+          if @ast
+            raise Error.new("Multiple ptypes are unsupported in AST mode")
+          end
           ptypename = md[1]
         end
         pattern = parse_pattern! || name
         consume!(/\s+/)
         unless code = parse_code_block!
-          consume!(/;/, "expected pattern or `;' or code block")
+          consume!(/;/, "expected `;' or code block")
         end
         token = Token.new(name, ptypename, @line_number)
         @tokens << token
-        pattern = Pattern.new(pattern: pattern, token: token, line_number: @line_number, code: code, mode: @mode, ptypename: ptypename)
+        pattern = Pattern.new(pattern: pattern, token: token, line_number: @line_number, code: code, modes: get_modes_from_modeline, ptypename: ptypename)
         @patterns << pattern
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -125,12 +165,15 @@ class Propane
         md = consume!(/(#{IDENTIFIER_REGEX})\s*/, "expected token name")
         name = md[1]
         if md = consume!(/\((#{IDENTIFIER_REGEX})\)\s*/)
+          if @ast
+            raise Error.new("Multiple ptypes are unsupported in AST mode")
+          end
           ptypename = md[1]
         end
         consume!(/;/, "expected `;'");
         token = Token.new(name, ptypename, @line_number)
         @tokens << token
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -143,8 +186,8 @@ class Propane
         end
         consume!(/\s+/)
         consume!(/;/, "expected `;'")
-        @patterns << Pattern.new(pattern: pattern, line_number: @line_number, mode: @mode)
-        @mode = nil
+        @patterns << Pattern.new(pattern: pattern, line_number: @line_number, modes: get_modes_from_modeline)
+        @modeline = nil
         true
       end
     end
@@ -152,13 +195,20 @@ class Propane
     def parse_rule_statement!
       if md = consume!(/(#{IDENTIFIER_REGEX})\s*(?:\((#{IDENTIFIER_REGEX})\))?\s*->\s*/)
         rule_name, ptypename = *md[1, 2]
-        md = consume!(/((?:#{IDENTIFIER_REGEX}\s*)*)\s*/, "expected rule component list")
+        if @ast && ptypename
+          raise Error.new("Multiple ptypes are unsupported in AST mode")
+        end
+        md = consume!(/((?:#{IDENTIFIER_REGEX}\??(?::#{IDENTIFIER_REGEX})?\s*)*)\s*/, "expected rule component list")
         components = md[1].strip.split(/\s+/)
-        unless code = parse_code_block!
-          consume!(/;/, "expected pattern or `;' or code block")
+        if @ast
+          consume!(/;/, "expected `;'")
+        else
+          unless code = parse_code_block!
+            consume!(/;/, "expected `;' or code block")
+          end
         end
         @rules << Rule.new(rule_name, components, code, ptypename, @line_number)
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -167,26 +217,37 @@ class Propane
       if pattern = parse_pattern!
         consume!(/\s+/)
         if md = consume!(/\((#{IDENTIFIER_REGEX})\)\s*/)
+          if @ast
+            raise Error.new("Multiple ptypes are unsupported in AST mode")
+          end
           ptypename = md[1]
         end
         unless code = parse_code_block!
           raise Error.new("Line #{@line_number}: expected code block to follow pattern")
         end
-        @patterns << Pattern.new(pattern: pattern, line_number: @line_number, code: code, mode: @mode, ptypename: ptypename)
-        @mode = nil
+        @patterns << Pattern.new(pattern: pattern, line_number: @line_number, code: code, modes: get_modes_from_modeline, ptypename: ptypename)
+        @modeline = nil
         true
       end
     end
 
+    def parse_start_statement!
+      if md = consume!(/start\s+(\w+)\s*;/)
+        @start_rule = md[1]
+      end
+    end
+
     def parse_code_block_statement!
-      if md = consume!(/<<([a-z]*)\n(.*?)^>>\n/m)
+      if md = consume!(/<<([a-z]*)(.*?)>>\n/m)
         name, code = md[1..2]
+        code.sub!(/\A\n/, "")
+        code += "\n" unless code.end_with?("\n")
         if @code_blocks[name]
           @code_blocks[name] += code
         else
           @code_blocks[name] = code
         end
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -218,8 +279,11 @@ class Propane
     end
 
     def parse_code_block!
-      if md = consume!(/<<\n(.*?)^>>\n/m)
-        md[1]
+      if md = consume!(/<<(.*?)>>\n/m)
+        code = md[1]
+        code.sub!(/\A\n/, "")
+        code += "\n" unless code.end_with?("\n")
+        code
       end
     end
 
@@ -251,6 +315,14 @@ class Propane
       end
     end
 
+    def get_modes_from_modeline
+      if @modeline
+        Set[*@modeline.split(",").map(&:strip)]
+      else
+        Set.new
+      end
+    end
+
   end
 
 end

lib/propane/lexer.rb

@@ -26,8 +26,14 @@ class Propane
     private
 
     def build_tables!
-      @modes = @grammar.patterns.group_by do |pattern|
-        pattern.mode
+      modenames = @grammar.patterns.reduce(Set.new) do |result, pattern|
+        result + pattern.modes
+      end
+      @modes = modenames.reduce({}) do |result, modename|
+        result[modename] = @grammar.patterns.select do |pattern|
+          pattern.modes.include?(modename)
+        end
+        result
       end.transform_values do |patterns|
         {dfa: DFA.new(patterns)}
       end

lib/propane/parser.rb

@@ -7,12 +7,15 @@ class Propane
     attr_reader :reduce_table
     attr_reader :rule_sets
 
-    def initialize(grammar, rule_sets, log)
+    def initialize(grammar, rule_sets, log, options)
       @grammar = grammar
       @rule_sets = rule_sets
       @log = log
       @item_sets = []
       @item_sets_set = {}
+      @warnings = Set.new
+      @errors = Set.new
+      @options = options
       start_item = Item.new(grammar.rules.first, 0)
       eval_item_sets = Set[ItemSet.new([start_item])]
 
@@ -23,10 +26,10 @@ class Propane
           item_set.id = @item_sets.size
           @item_sets << item_set
           @item_sets_set[item_set] = item_set
-          item_set.following_symbols.each do |following_symbol|
-            unless following_symbol.name == "$EOF"
-              following_set = item_set.build_following_item_set(following_symbol)
-              eval_item_sets << following_set
+          item_set.next_symbols.each do |next_symbol|
+            unless next_symbol.name == "$EOF"
+              next_item_set = item_set.build_next_item_set(next_symbol)
+              eval_item_sets << next_item_set
             end
           end
         end
@@ -37,8 +40,21 @@ class Propane
       end
 
       build_reduce_actions!
-      write_log!
       build_tables!
+      write_log!
+      errormessage = ""
+      if @errors.size > 0
+        errormessage += @errors.join("\n")
+      end
+      if @warnings.size > 0 && @options[:warnings_as_errors]
+        if errormessage != ""
+          errormessage += "\n"
+        end
+        errormessage += "Fatal errors (-w):\n" + @warnings.join("\n")
+      end
+      if errormessage != ""
+        raise Error.new(errormessage)
+      end
     end
 
     private
@@ -48,48 +64,47 @@ class Propane
       @shift_table = []
       @reduce_table = []
       @item_sets.each do |item_set|
-        shift_entries = item_set.following_symbols.map do |following_symbol|
-          state_id =
-            if following_symbol.name == "$EOF"
-              0
-            else
-              item_set.following_item_set[following_symbol].id
+        unless item_set.reduce_rules.empty?
+          item_set.shift_entries.each do |shift_entry|
+            token = shift_entry[:symbol]
+            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
-          {
-            symbol_id: following_symbol.id,
-            state_id: state_id,
-          }
+          end
         end
         reduce_entries =
-          case ra = item_set.reduce_actions
-          when Rule
-            [{token_id: @grammar.invalid_token_id, rule_id: ra.id,
-              rule_set_id: ra.rule_set.id, n_states: ra.components.size}]
-          when Hash
-            ra.map do |token, rule|
-              {token_id: token.id, rule_id: rule.id,
-               rule_set_id: rule.rule_set.id, n_states: rule.components.size}
+          if rule = item_set.reduce_rule
+            [{token_id: @grammar.invalid_token_id, rule_id: rule.id, rule: rule,
+              rule_set_id: rule.rule_set.id, n_states: rule.components.size,
+              propagate_optional_target: rule.optional? && rule.components.size == 1}]
+          elsif reduce_actions = item_set.reduce_actions
+            reduce_actions.map do |token, rule|
+              {token: token, token_id: token.id, rule_id: rule.id, rule: rule,
+               rule_set_id: rule.rule_set.id, n_states: rule.components.size,
+               propagate_optional_target: rule.optional? && rule.components.size == 1}
             end
           else
             []
           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
 
     def process_item_set(item_set)
-      item_set.following_symbols.each do |following_symbol|
-        unless following_symbol.name == "$EOF"
-          following_set = @item_sets_set[item_set.build_following_item_set(following_symbol)]
-          item_set.following_item_set[following_symbol] = following_set
-          following_set.in_sets << item_set
+      item_set.next_symbols.each do |next_symbol|
+        unless next_symbol.name == "$EOF"
+          next_item_set = @item_sets_set[item_set.build_next_item_set(next_symbol)]
+          item_set.next_item_set[next_symbol] = next_item_set
+          next_item_set.in_sets << item_set
         end
       end
     end
@@ -99,7 +114,109 @@ class Propane
     # @return [void]
     def build_reduce_actions!
       @item_sets.each do |item_set|
-        item_set.reduce_actions = build_reduce_actions_for_item_set(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
 
@@ -108,40 +225,45 @@ class Propane
     # @param item_set [ItemSet]
     #   ItemSet (parser state)
     #
-    # @return [nil, Rule, Hash]
-    #   If no reduce actions are possible for the given item set, nil.
-    #   If only one reduce action is possible for the given item set, the Rule
-    #   to reduce.
-    #   Otherwise, a mapping of lookahead Tokens to the Rules to reduce.
+    # @return [void]
     def build_reduce_actions_for_item_set(item_set)
       # To build the reduce actions, we start by looking at any
       # "complete" items, i.e., items where the parse position is at the
       # end of a rule. These are the only rules that are candidates for
       # reduction in the current ItemSet.
-      reduce_rules = Set.new(item_set.items.select(&:complete?).map(&:rule))
+      item_set.reduce_rules = Set.new(item_set.items.select(&:complete?).map(&:rule))
 
-      # If there are no rules to reduce for this ItemSet, we're done here.
-      return nil if reduce_rules.size == 0
-
-      # If there is exactly one rule to reduce for this ItemSet, then do not
-      # figure out the lookaheads; just reduce it.
-      return reduce_rules.first if reduce_rules.size == 1
-
-      # Otherwise, we have more than one possible rule to reduce.
+      if item_set.reduce_rules.size == 1
+        item_set.reduce_rule = item_set.reduce_rules.first
+      end
+    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, 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
       # up to this one. This restriction gives us a more precise lookahead set,
       # and allows us to parse LALR grammars.
-      item_sets = item_set.leading_item_sets
-      reduce_rules.reduce({}) do |reduce_actions, reduce_rule|
+      item_sets = Set[item_set] + item_set.leading_item_sets
+      item_set.reduce_rules.reduce({}) do |reduce_actions, reduce_rule|
         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]
-            raise Error.new("Error: reduce/reduce conflict between rule #{existing_reduce_rule.id} (#{existing_reduce_rule.name}) and rule #{reduce_rule.id} (#{reduce_rule.name})")
+            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
@@ -181,13 +303,14 @@ class Propane
         # tokens to form the lookahead token set.
         item_sets.each do |item_set|
           item_set.items.each do |item|
-            if item.following_symbol == rule_set
+            if item.next_symbol == rule_set
               (1..).each do |offset|
-                case symbol = item.following_symbol(offset)
+                case symbol = item.next_symbol(offset)
                 when nil
                   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
@@ -240,20 +363,26 @@ class Propane
         @log.puts
         @log.puts "  Incoming states: #{incoming_ids.join(", ")}"
         @log.puts "  Outgoing states:"
-        item_set.following_item_set.each do |following_symbol, following_item_set|
-          @log.puts "    #{following_symbol.name} => #{following_item_set.id}"
+        item_set.next_item_set.each do |next_symbol, next_item_set|
+          @log.puts "    #{next_symbol.name} => #{next_item_set.id}"
         end
         @log.puts
         @log.puts "  Reduce actions:"
-        case item_set.reduce_actions
-        when Rule
-          @log.puts "    * => rule #{item_set.reduce_actions.id}, rule set #{@rule_sets[item_set.reduce_actions.name].id} (#{item_set.reduce_actions.name})"
-        when Hash
+        if item_set.reduce_rule
+          @log.puts "    * => rule #{item_set.reduce_rule.id}, rule set #{@rule_sets[item_set.reduce_rule.name].id} (#{item_set.reduce_rule.name})"
+        elsif item_set.reduce_actions
           item_set.reduce_actions.each do |token, rule|
             @log.puts "    lookahead #{token.name} => #{rule.name} (#{rule.id}), rule set ##{rule.rule_set.id}"
           end
         end
       end
+      if @warnings.size > 0
+        @log.puts
+        @log.puts "Warnings:"
+        @warnings.each do |warning|
+          @log.puts "  #{warning}"
+        end
+      end
     end
 
   end

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.
@@ -56,7 +57,7 @@ class Propane
 
       # Return the set of Items obtained by "closing" the current item.
       #
-      # If the following symbol for the current item is another Rule name, then
+      # If the next symbol for the current item is another Rule name, then
       # this method will return all Items for that Rule with a position of 0.
       # Otherwise, an empty Array is returned.
       #
@@ -81,17 +82,17 @@ class Propane
         @position == @rule.components.size
       end
 
-      # Get the following symbol for the Item.
+      # Get the next symbol for the Item.
       #
-      # That is, the symbol which follows the parse position marker in the
+      # That is, the symbol which is after the parse position marker in the
       # current Item.
       #
       # @param offset [Integer]
       #   Offset from current parse position to examine.
       #
       # @return [Token, RuleSet, nil]
-      #   Following symbol for the Item.
-      def following_symbol(offset = 0)
+      #   Next symbol for the Item.
+      def next_symbol(offset = 0)
         @rule.components[@position + offset]
       end
 
@@ -108,25 +109,25 @@ class Propane
         end
       end
 
-      # Get whether this Item is followed by the provided symbol.
+      # Get whether this Item's next symbol is the given symbol.
       #
       # @param symbol [Token, RuleSet]
       #   Symbol to query.
       #
       # @return [Boolean]
-      #   Whether this Item is followed by the provided symbol.
-      def followed_by?(symbol)
-        following_symbol == symbol
+      #   Whether this Item's next symbol is the given symbol.
+      def next_symbol?(symbol)
+        next_symbol == symbol
       end
 
-      # Get the following item for this Item.
+      # Get the next item for this Item.
       #
       # That is, the Item formed by moving the parse position marker one place
       # forward from its position in this Item.
       #
       # @return [Item]
-      #   The following item for this Item.
-      def following_item
+      #   The next item for this Item.
+      def next_item
         Item.new(@rule, @position + 1)
       end
 

lib/propane/parser/item_set.rb

@@ -2,7 +2,7 @@ class Propane
   class Parser
 
     # Represent a parser "item set", which is a set of possible items that the
-    # parser could currently be parsing.
+    # parser could currently be parsing. This is equivalent to a parser state.
     class ItemSet
 
       # @return [Set<Item>]
@@ -14,45 +14,58 @@ class Propane
       attr_accessor :id
 
       # @return [Hash]
-      #   Maps a following symbol to its ItemSet.
-      attr_reader :following_item_set
+      #   Maps a next symbol to its ItemSet.
+      attr_reader :next_item_set
 
       # @return [Set<ItemSet>]
       #   ItemSets leading to this item set.
       attr_reader :in_sets
 
-      # @return [nil, Rule, Hash]
-      #   Reduce actions, mapping lookahead tokens to rules.
+      # @return [nil, Rule]
+      #   Rule to reduce if there is only one possibility.
+      attr_accessor :reduce_rule
+
+      # @return [Set<Rule>]
+      #   Set of rules that could be reduced in this parser state.
+      attr_accessor :reduce_rules
+
+      # @return [nil, Hash]
+      #   Reduce actions, mapping lookahead tokens to rules, if there is
+      #   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>]
       #   Items in this ItemSet.
       def initialize(items)
         @items = Set.new(items)
-        @following_item_set = {}
+        @next_item_set = {}
         @in_sets = Set.new
         close!
       end
 
-      # Get the set of following symbols for all Items in this ItemSet.
+      # Get the set of next symbols for all Items in this ItemSet.
       #
       # @return [Set<Token, RuleSet>]
-      #   Set of following symbols for all Items in this ItemSet.
-      def following_symbols
-        Set.new(@items.map(&:following_symbol).compact)
+      #   Set of next symbols for all Items in this ItemSet.
+      def next_symbols
+        @_next_symbols ||= Set.new(@items.map(&:next_symbol).compact)
       end
 
-      # Build a following ItemSet for the given following symbol.
+      # Build a next ItemSet for the given next symbol.
       #
       # @param symbol [Token, RuleSet]
-      #   Following symbol to build the following ItemSet for.
+      #   Next symbol to build the next ItemSet for.
       #
       # @return [ItemSet]
-      #   Following ItemSet for the given following symbol.
-      def build_following_item_set(symbol)
-        ItemSet.new(items_followed_by(symbol).map(&:following_item))
+      #   Next ItemSet for the given next symbol.
+      def build_next_item_set(symbol)
+        ItemSet.new(items_with_next(symbol).map(&:next_item))
       end
 
       # Hash function.
@@ -87,14 +100,27 @@ class Propane
 
       # Set of ItemSets that lead to this ItemSet.
       #
-      # This set includes this ItemSet.
-      #
       # @return [Set<ItemSet>]
       #   Set of all ItemSets that lead up to this ItemSet.
       def leading_item_sets
-        @in_sets.reduce(Set[self]) do |result, item_set|
-          result + item_set.leading_item_sets
-        end
+        @_leading_item_sets ||=
+          begin
+            result = Set.new
+            eval_sets = Set[self]
+            evaled = Set.new
+            while eval_sets.size > 0
+              eval_set = eval_sets.first
+              eval_sets.delete(eval_set)
+              evaled << eval_set
+              eval_set.in_sets.each do |in_set|
+                result << in_set
+                unless evaled.include?(in_set)
+                  eval_sets << in_set
+                end
+              end
+            end
+            result
+          end
       end
 
       # Represent the ItemSet as a String.
@@ -127,16 +153,16 @@ class Propane
         end
       end
 
-      # Get the Items followed by the given following symbol.
+      # Get the Items with the given next symbol.
       #
       # @param symbol [Token, RuleSet]
-      #   Following symbol.
+      #   Next symbol.
       #
       # @return [Array<Item>]
-      #   Items followed by the given following symbol.
-      def items_followed_by(symbol)
+      #   Items with the given next symbol.
+      def items_with_next(symbol)
         @items.select do |item|
-          item.followed_by?(symbol)
+          item.next_symbol?(symbol)
         end
       end
 

lib/propane/pattern.rb

@@ -26,9 +26,9 @@ class Propane
     #   Regex NFA for matching the pattern.
     attr_reader :nfa
 
-    # @return [String, nil]
-    #   Lexer mode for this pattern.
-    attr_accessor :mode
+    # @return [Set]
+    #   Lexer modes for this pattern.
+    attr_accessor :modes
 
     # @return [String, nil]
     #   Parser value type name.
@@ -46,16 +46,16 @@ class Propane
     #   Token to be returned by this pattern.
     # @option options [Integer, nil] :line_number
     #   Line number where the token was defined in the input grammar.
-    # @option options [String, nil] :mode
-    #   Lexer mode for this pattern.
+    # @option options [String, nil] :modes
+    #   Lexer modes for this pattern.
     def initialize(options)
       @code = options[:code]
       @pattern = options[:pattern]
       @token = options[:token]
       @line_number = options[:line_number]
-      @mode = options[:mode]
+      @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
@@ -124,18 +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")
+        when "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")
+        when "n"
+          CharacterRangeUnit.new("\n")
+        when "r"
+          CharacterRangeUnit.new("\r")
         when "s"
           ccu = CharacterClassUnit.new
           ccu << CharacterRangeUnit.new(" ")
@@ -145,6 +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")
+        when "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

@@ -92,16 +92,19 @@ 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)
-          thing.each do |ccu_unit|
-            @units << ccu_unit
+          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

lib/propane/rule.rb

@@ -6,6 +6,10 @@ class Propane
     #   Rule components.
     attr_reader :components
 
+    # @return [Hash]
+    #   Field aliases.
+    attr_reader :aliases
+
     # @return [String]
     #   User code associated with the rule.
     attr_reader :code
@@ -30,6 +34,11 @@ class Propane
     #   The RuleSet that this Rule is a part of.
     attr_accessor :rule_set
 
+    # @return [Array<Integer>]
+    #   Map this rule's components to their positions in the parent RuleSet's
+    #   node field pointer array. This is used for AST construction.
+    attr_accessor :rule_set_node_field_index_map
+
     # Construct a Rule.
     #
     # @param name [String]
@@ -44,7 +53,20 @@ class Propane
     #   Line number where the rule was defined in the input grammar.
     def initialize(name, components, code, ptypename, line_number)
       @name = name
-      @components = components
+      @aliases = {}
+      @components = components.each_with_index.map do |component, i|
+        if component =~ /(\S+):(\S+)/
+          c, aliasname = $1, $2
+          if @aliases[aliasname]
+            raise Error.new("Error: duplicate field alias `#{aliasname}` for rule #{name} defined on line #{line_number}")
+          end
+          @aliases[aliasname] = i
+          c
+        else
+          component
+        end
+      end
+      @rule_set_node_field_index_map = components.map {0}
       @code = code
       @ptypename = ptypename
       @line_number = line_number
@@ -60,6 +82,14 @@ class Propane
       @components.empty?
     end
 
+    # Return whether this is an optional Rule.
+    #
+    # @return [Boolean]
+    #   Whether this is an optional Rule.
+    def optional?
+      @name.end_with?("?")
+    end
+
     # Represent the Rule as a String.
     #
     # @return [String]
@@ -68,6 +98,17 @@ class Propane
       "#{@name} -> #{@components.map(&:name).join(" ")}"
     end
 
+    # Check whether the rule set node field index map is just a 1:1 mapping.
+    #
+    # @return [Boolean]
+    #   Boolean indicating whether the rule set node field index map is just a
+    #   1:1 mapping.
+    def flat_rule_set_node_field_index_map?
+      @rule_set_node_field_index_map.each_with_index.all? do |v, i|
+        v == i
+      end
+    end
+
   end
 
 end

lib/propane/rule_set.rb

@@ -1,7 +1,12 @@
 class Propane
 
+  # A RuleSet collects all grammar rules of the same name.
   class RuleSet
 
+    # @return [Array<Hash>]
+    #   AST fields.
+    attr_reader :ast_fields
+
     # @return [Integer]
     #   ID of the RuleSet.
     attr_reader :id
@@ -51,6 +56,24 @@ class Propane
       @could_be_empty
     end
 
+    # Return whether this is an optional RuleSet.
+    #
+    # @return [Boolean]
+    #   Whether this is an optional RuleSet.
+    def optional?
+      @name.end_with?("?")
+    end
+
+    # For optional rule sets, return the underlying component that is optional.
+    def option_target
+      @rules.each do |rule|
+        if rule.components.size > 0
+          return rule.components[0]
+        end
+      end
+      raise "Optional rule target not found"
+    end
+
     # Build the start token set for the RuleSet.
     #
     # @return [Set<Token>]
@@ -75,6 +98,72 @@ class Propane
       @_start_token_set
     end
 
+    # Finalize a RuleSet after adding all Rules to it.
+    def finalize(grammar)
+      if grammar.ast
+        build_ast_fields(grammar)
+      end
+    end
+
+    private
+
+    # Build the set of AST fields for this RuleSet.
+    #
+    # This is an Array of Hashes. Each entry in the Array corresponds to a
+    # field location in the AST node. The entry is a Hash. It could have one or
+    # two keys. It will always have the field name with a positional suffix as
+    # a key. It may also have the field name without the positional suffix if
+    # that field only exists in one position across all Rules in the RuleSet.
+    #
+    # @return [void]
+    def build_ast_fields(grammar)
+      field_ast_node_indexes = {}
+      field_indexes_across_all_rules = {}
+      # Stores the index into @ast_fields by field alias name.
+      field_aliases = {}
+      @ast_fields = []
+      @rules.each do |rule|
+        rule.components.each_with_index do |component, i|
+          if component.is_a?(RuleSet) && component.optional?
+            component = component.option_target
+          end
+          if component.is_a?(Token)
+            node_name = "Token"
+          else
+            node_name = component.name
+          end
+          struct_name = "#{grammar.ast_prefix}#{node_name}#{grammar.ast_suffix}"
+          field_name = "p#{node_name}#{i + 1}"
+          unless field_ast_node_indexes[field_name]
+            field_ast_node_indexes[field_name] = @ast_fields.size
+            @ast_fields << {field_name => struct_name}
+          end
+          rule.aliases.each do |alias_name, index|
+            if index == i
+              alias_ast_fields_index = field_ast_node_indexes[field_name]
+              if field_aliases[alias_name] && field_aliases[alias_name] != alias_ast_fields_index
+                raise Error.new("Error: conflicting AST node field positions for alias `#{alias_name}` in rule #{rule.name} defined on line #{rule.line_number}")
+              end
+              field_aliases[alias_name] = alias_ast_fields_index
+              @ast_fields[alias_ast_fields_index][alias_name] = @ast_fields[alias_ast_fields_index].first[1]
+            end
+          end
+          field_indexes_across_all_rules[node_name] ||= Set.new
+          field_indexes_across_all_rules[node_name] << field_ast_node_indexes[field_name]
+          rule.rule_set_node_field_index_map[i] = field_ast_node_indexes[field_name]
+        end
+      end
+      field_indexes_across_all_rules.each do |node_name, indexes_across_all_rules|
+        if indexes_across_all_rules.size == 1
+          # If this field was only seen in one position across all rules,
+          # then add an alias to the positional field name that does not
+          # include the position.
+          @ast_fields[indexes_across_all_rules.first]["p#{node_name}"] =
+            "#{grammar.ast_prefix}#{node_name}#{grammar.ast_suffix}"
+        end
+      end
+    end
+
   end
 
 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 = "1.1.0"
+  VERSION = "2.3.0"
 end

spec/ast_node_memory_remains.c.propane

@@ -0,0 +1,151 @@
+ast;
+ast_prefix P;
+
+<<header
+#include <stdio.h>
+#include <stdint.h>
+
+typedef union
+{
+    uint64_t i64;
+    const uint8_t * s;
+    double dou;
+} TokenVal;
+>>
+
+ptype TokenVal;
+
+# Keywords.
+token byte;
+token def;
+token int;
+token long;
+token module;
+token return;
+token short;
+token size_t;
+token ssize_t;
+token ubyte;
+token uint;
+token ulong;
+token ushort;
+
+# Symbols.
+token arrow /->/;
+token comma /,/;
+token lbrace /\{/;
+token lparen /\(/;
+token rbrace /\}/;
+token rparen /\)/;
+token semicolon /;/;
+
+# Integer literals.
+token hex_int_l /0[xX][0-9a-fA-F][0-9a-fA-F_]*/ <<
+  $$.i64 = 64u;
+>>
+
+# Identifier.
+token ident /\$?[a-zA-Z_][a-zA-Z_0-9]*\??/ <<
+  $$.s = match;
+  $mode(default);
+  return $token(ident);
+>>
+
+# Comments.
+drop /#.*/;
+
+# Whitespace.
+drop /[ \r\n]*/;
+
+start Module;
+
+# Assignment operators - right associative
+Expression -> Expression_Or:exp0;
+
+# Logical OR operator - left associative
+Expression_Or -> Expression_And:exp0;
+
+# Logical AND operator - left associative
+Expression_And -> Expression_Comp:exp0;
+
+# Equality operators - left associative
+Expression_Comp -> Expression_Relational:exp0;
+
+# Relational operators - left associative
+Expression_Relational -> Expression_REMatch:exp0;
+
+# Regular expression - left associative
+Expression_REMatch -> Expression_BinOr:exp0;
+
+# Binary OR operator - left associative
+Expression_BinOr -> Expression_Xor:exp0;
+
+# Binary XOR operator - left associative
+Expression_Xor -> Expression_BinAnd:exp0;
+
+# Binary AND operator - left associative
+Expression_BinAnd -> Expression_BitShift:exp0;
+
+# Bit shift operators - left associative
+Expression_BitShift -> Expression_Plus:exp0;
+
+# Add/subtract operators - left associative
+Expression_Plus -> Expression_Mul:exp0;
+
+# Multiplication/divide/modulus operators - left associative
+Expression_Mul -> Expression_Range:exp0;
+
+# Range construction operators - left associative
+Expression_Range -> Expression_UnaryPrefix:exp0;
+
+# Unary prefix operators
+Expression_UnaryPrefix -> Expression_Dot:exp0;
+
+# Postfix operators
+Expression_Dot -> Expression_Ident:exp0;
+Expression_Dot -> Expression_Dot:exp1 lparen rparen;
+
+# Literals, identifiers, and parenthesized expressions
+Expression_Ident -> Literal;
+Expression_Ident -> ident;
+
+FunctionDefinition -> def ident:name lparen FunctionParameterList?:parameters rparen FunctionReturnType?:returntype lbrace Statements rbrace;
+
+FunctionParameterList -> ident:name Type:type FunctionParameterListMore?:more;
+FunctionParameterListMore -> comma ident:name Type:type FunctionParameterListMore?:more;
+
+FunctionReturnType -> arrow Type;
+
+Literal -> LiteralInteger;
+LiteralInteger -> hex_int_l;
+
+Module -> ModuleStatement? ModuleItems;
+
+ModuleItem -> FunctionDefinition;
+
+ModuleItems -> ;
+ModuleItems -> ModuleItems ModuleItem;
+
+ModulePath -> ident;
+
+ModuleStatement -> module ModulePath semicolon;
+
+ReturnStatement -> return Expression?:exp0 semicolon;
+
+Statements -> ;
+Statements -> Statements Statement;
+Statement -> Expression semicolon;
+Statement -> ReturnStatement;
+
+Type -> TypeBase;
+
+TypeBase -> byte;
+TypeBase -> ubyte;
+TypeBase -> short;
+TypeBase -> ushort;
+TypeBase -> int;
+TypeBase -> uint;
+TypeBase -> long;
+TypeBase -> ulong;
+TypeBase -> size_t;
+TypeBase -> ssize_t;

spec/ast_node_memory_remains.d.propane

@@ -0,0 +1,177 @@
+ast;
+ast_prefix P;
+
+<<
+import std.bigint;
+
+private string stringvalue;
+
+union TokenVal
+{
+    BigInt bi;
+    string s;
+    double dou;
+}
+>>
+
+ptype TokenVal;
+
+# Keywords.
+token byte;
+token def;
+token int;
+token long;
+token module;
+token return;
+token short;
+token size_t;
+token ssize_t;
+token ubyte;
+token uint;
+token ulong;
+token ushort;
+
+# Symbols.
+token arrow /->/;
+token comma /,/;
+token lbrace /\{/;
+token lparen /\(/;
+token rbrace /\}/;
+token rparen /\)/;
+token semicolon /;/;
+
+# Integer literals.
+token hex_int_l /0[xX][0-9a-fA-F][0-9a-fA-F_]*/ <<
+  $$.bi = BigInt(match[0..3]);
+  foreach (c; match[3..$])
+  {
+    if (('0' <= c) && (c <= '9'))
+    {
+      $$.bi *= 0x10;
+      $$.bi += (c - '0');
+    }
+    if (('a' <= c) && (c <= 'f'))
+    {
+      $$.bi *= 0x10;
+      $$.bi += (c - 'a' + 10);
+    }
+    if (('A' <= c) && (c <= 'F'))
+    {
+      $$.bi *= 0x10;
+      $$.bi += (c - 'A' + 10);
+    }
+  }
+>>
+
+# Identifier.
+token ident /\$?[a-zA-Z_][a-zA-Z_0-9]*\??/ <<
+  if (match[0] == '$')
+  {
+    $$.s = match[1..$];
+  }
+  else
+  {
+    $$.s = match;
+  }
+  $mode(default);
+  return $token(ident);
+>>
+
+# Comments.
+drop /#.*/;
+
+# Whitespace.
+drop /[ \r\n]*/;
+
+start Module;
+
+# Assignment operators - right associative
+Expression -> Expression_Or:exp0;
+
+# Logical OR operator - left associative
+Expression_Or -> Expression_And:exp0;
+
+# Logical AND operator - left associative
+Expression_And -> Expression_Comp:exp0;
+
+# Equality operators - left associative
+Expression_Comp -> Expression_Relational:exp0;
+
+# Relational operators - left associative
+Expression_Relational -> Expression_REMatch:exp0;
+
+# Regular expression - left associative
+Expression_REMatch -> Expression_BinOr:exp0;
+
+# Binary OR operator - left associative
+Expression_BinOr -> Expression_Xor:exp0;
+
+# Binary XOR operator - left associative
+Expression_Xor -> Expression_BinAnd:exp0;
+
+# Binary AND operator - left associative
+Expression_BinAnd -> Expression_BitShift:exp0;
+
+# Bit shift operators - left associative
+Expression_BitShift -> Expression_Plus:exp0;
+
+# Add/subtract operators - left associative
+Expression_Plus -> Expression_Mul:exp0;
+
+# Multiplication/divide/modulus operators - left associative
+Expression_Mul -> Expression_Range:exp0;
+
+# Range construction operators - left associative
+Expression_Range -> Expression_UnaryPrefix:exp0;
+
+# Unary prefix operators
+Expression_UnaryPrefix -> Expression_Dot:exp0;
+
+# Postfix operators
+Expression_Dot -> Expression_Ident:exp0;
+Expression_Dot -> Expression_Dot:exp1 lparen rparen;
+
+# Literals, identifiers, and parenthesized expressions
+Expression_Ident -> Literal;
+Expression_Ident -> ident;
+
+FunctionDefinition -> def ident:name lparen FunctionParameterList?:parameters rparen FunctionReturnType?:returntype lbrace Statements rbrace;
+
+FunctionParameterList -> ident:name Type:type FunctionParameterListMore?:more;
+FunctionParameterListMore -> comma ident:name Type:type FunctionParameterListMore?:more;
+
+FunctionReturnType -> arrow Type;
+
+Literal -> LiteralInteger;
+LiteralInteger -> hex_int_l;
+
+Module -> ModuleStatement? ModuleItems;
+
+ModuleItem -> FunctionDefinition;
+
+ModuleItems -> ;
+ModuleItems -> ModuleItems ModuleItem;
+
+ModulePath -> ident;
+
+ModuleStatement -> module ModulePath semicolon;
+
+ReturnStatement -> return Expression?:exp0 semicolon;
+
+Statements -> ;
+Statements -> Statements Statement;
+Statement -> Expression semicolon;
+Statement -> ReturnStatement;
+
+Type -> TypeBase;
+
+TypeBase -> byte;
+TypeBase -> ubyte;
+TypeBase -> short;
+TypeBase -> ushort;
+TypeBase -> int;
+TypeBase -> uint;
+TypeBase -> long;
+TypeBase -> ulong;
+TypeBase -> size_t;
+TypeBase -> ssize_t;

spec/json_parser.c.propane

@@ -120,11 +120,11 @@ string: /\\t/ <<
 >>
 string: /\\u[0-9a-fA-F]{4}/ <<
   /* Not actually going to encode the code point for this example... */
-  char s[] = {'{', match[2], match[3], match[4], match[5], '}', 0};
+  char s[] = {'{', (char)match[2], (char)match[3], (char)match[4], (char)match[5], '}', 0};
   str_append(&string_value, s);
 >>
 string: /[^\\]/ <<
-  char s[] = {match[0], 0};
+  char s[] = {(char)match[0], 0};
   str_append(&string_value, s);
 >>
 Start -> Value <<

spec/json_types.c

@@ -5,7 +5,7 @@
 
 JSONValue * JSONValue_new(size_t id)
 {
-    JSONValue * jv = calloc(1, sizeof(JSONValue));
+    JSONValue * jv = (JSONValue *)calloc(1, sizeof(JSONValue));
     jv->id = id;
     return jv;
 }
@@ -29,7 +29,7 @@ void JSONObject_append(JSONValue * object, char const * name, JSONValue * value)
         }
     }
     size_t const new_size = size + 1;
-    void * new_entries = malloc(sizeof(object->object.entries[0]) * new_size);
+    JSONObjectEntry * new_entries = (JSONObjectEntry *)malloc(sizeof(object->object.entries[0]) * new_size);
     if (size > 0)
     {
         memcpy(new_entries, object->object.entries, size * sizeof(object->object.entries[0]));
@@ -52,7 +52,7 @@ void JSONArray_append(JSONValue * array, JSONValue * value)
 {
     size_t const size = array->array.size;
     size_t const new_size = size + 1;
-    JSONValue ** new_entries = malloc(sizeof(JSONValue *) * new_size);
+    JSONValue ** new_entries = (JSONValue **)malloc(sizeof(JSONValue *) * new_size);
     if (array->array.size > 0)
     {
         memcpy(new_entries, array->array.entries, sizeof(JSONValue *) * size);

spec/json_types.h

@@ -11,6 +11,12 @@
 #define JSON_FALSE 5u
 #define JSON_NULL 6u
 
+typedef struct JSONObjectEntry_s
+{
+    char const * name;
+    struct JSONValue_s * value;
+} JSONObjectEntry;
+
 typedef struct JSONValue_s
 {
     size_t id;
@@ -19,11 +25,7 @@ typedef struct JSONValue_s
         struct
         {
             size_t size;
-            struct
-            {
-                char const * name;
-                struct JSONValue_s * value;
-            } * entries;
+            JSONObjectEntry * entries;
         } object;
         struct
         {

spec/propane/grammar_spec.rb

@@ -151,30 +151,30 @@ EOF
 
       o = grammar.patterns.find {|pattern| pattern.token == o}
       expect(o).to_not be_nil
-      expect(o.mode).to be_nil
+      expect(o.modes).to be_empty
 
       o = grammar.tokens.find {|token| token.name == "b"}
       expect(o).to_not be_nil
 
       o = grammar.patterns.find {|pattern| pattern.token == o}
       expect(o).to_not be_nil
-      expect(o.mode).to eq "m1"
+      expect(o.modes).to eq Set["m1"]
 
       o = grammar.patterns.find {|pattern| pattern.pattern == "foo"}
       expect(o).to_not be_nil
-      expect(o.mode).to be_nil
+      expect(o.modes).to be_empty
 
       o = grammar.patterns.find {|pattern| pattern.pattern == "bar"}
       expect(o).to_not be_nil
-      expect(o.mode).to eq "m2"
+      expect(o.modes).to eq Set["m2"]
 
       o = grammar.patterns.find {|pattern| pattern.pattern == "q"}
       expect(o).to_not be_nil
-      expect(o.mode).to be_nil
+      expect(o.modes).to be_empty
 
       o = grammar.patterns.find {|pattern| pattern.pattern == "r"}
       expect(o).to_not be_nil
-      expect(o.mode).to eq "m3"
+      expect(o.modes).to eq Set["m3"]
     end
 
     it "allows assigning ptypes to tokens and rules" do

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/spec_helper.rb

@@ -2,6 +2,10 @@ unless ENV["dist_specs"]
   require "bundler/setup"
   require "simplecov"
 
+  class MyFormatter
+    def format(*args)
+    end
+  end
   SimpleCov.start do
     add_filter "/spec/"
     add_filter "/.bundle/"
@@ -12,6 +16,7 @@ unless ENV["dist_specs"]
     end
     project_name "Propane"
     merge_timeout 3600
+    formatter(MyFormatter)
   end
 
   RSpec.configure do |config|

spec/test_ast.c

@@ -0,0 +1,55 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "a, ((b)), b";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert_eq(P_SUCCESS, p_parse(&context));
+    Start * start = p_result(&context);
+    assert(start->pItems1 != NULL);
+    assert(start->pItems != NULL);
+    Items * items = start->pItems;
+    assert(items->pItem != NULL);
+    assert(items->pItem->pToken1 != NULL);
+    assert_eq(TOKEN_a, items->pItem->pToken1->token);
+    assert_eq(11, items->pItem->pToken1->pvalue);
+    assert(items->pItemsMore != NULL);
+    ItemsMore * itemsmore = items->pItemsMore;
+    assert(itemsmore->pItem != NULL);
+    assert(itemsmore->pItem->pItem != NULL);
+    assert(itemsmore->pItem->pItem->pItem != NULL);
+    assert(itemsmore->pItem->pItem->pItem->pToken1 != NULL);
+    assert_eq(TOKEN_b, itemsmore->pItem->pItem->pItem->pToken1->token);
+    assert_eq(22, itemsmore->pItem->pItem->pItem->pToken1->pvalue);
+    assert(itemsmore->pItemsMore != NULL);
+    itemsmore = itemsmore->pItemsMore;
+    assert(itemsmore->pItem != NULL);
+    assert(itemsmore->pItem->pToken1 != NULL);
+    assert_eq(TOKEN_b, itemsmore->pItem->pToken1->token);
+    assert_eq(22, itemsmore->pItem->pToken1->pvalue);
+    assert(itemsmore->pItemsMore == NULL);
+
+    input = "";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert_eq(P_SUCCESS, p_parse(&context));
+    start = p_result(&context);
+    assert(start->pItems == NULL);
+
+    input = "2 1";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert_eq(P_SUCCESS, p_parse(&context));
+    start = p_result(&context);
+    assert(start->pItems != NULL);
+    assert(start->pItems->pItem != NULL);
+    assert(start->pItems->pItem->pDual != NULL);
+    assert(start->pItems->pItem->pDual->pTwo1 != NULL);
+    assert(start->pItems->pItem->pDual->pOne2 != NULL);
+    assert(start->pItems->pItem->pDual->pTwo2 == NULL);
+    assert(start->pItems->pItem->pDual->pOne1 == NULL);
+
+    return 0;
+}

spec/test_ast.d

@@ -0,0 +1,57 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "a, ((b)), b";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert_eq(P_SUCCESS, p_parse(&context));
+    Start * start = p_result(&context);
+    assert(start.pItems1 !is null);
+    assert(start.pItems !is null);
+    Items * items = start.pItems;
+    assert(items.pItem !is null);
+    assert(items.pItem.pToken1 !is null);
+    assert_eq(TOKEN_a, items.pItem.pToken1.token);
+    assert_eq(11, items.pItem.pToken1.pvalue);
+    assert(items.pItemsMore !is null);
+    ItemsMore * itemsmore = items.pItemsMore;
+    assert(itemsmore.pItem !is null);
+    assert(itemsmore.pItem.pItem !is null);
+    assert(itemsmore.pItem.pItem.pItem !is null);
+    assert(itemsmore.pItem.pItem.pItem.pToken1 !is null);
+    assert_eq(TOKEN_b, itemsmore.pItem.pItem.pItem.pToken1.token);
+    assert_eq(22, itemsmore.pItem.pItem.pItem.pToken1.pvalue);
+    assert(itemsmore.pItemsMore !is null);
+    itemsmore = itemsmore.pItemsMore;
+    assert(itemsmore.pItem !is null);
+    assert(itemsmore.pItem.pToken1 !is null);
+    assert_eq(TOKEN_b, itemsmore.pItem.pToken1.token);
+    assert_eq(22, itemsmore.pItem.pToken1.pvalue);
+    assert(itemsmore.pItemsMore is null);
+
+    input = "";
+    p_context_init(&context, input);
+    assert_eq(P_SUCCESS, p_parse(&context));
+    start = p_result(&context);
+    assert(start.pItems is null);
+
+    input = "2 1";
+    p_context_init(&context, input);
+    assert_eq(P_SUCCESS, p_parse(&context));
+    start = p_result(&context);
+    assert(start.pItems !is null);
+    assert(start.pItems.pItem !is null);
+    assert(start.pItems.pItem.pDual !is null);
+    assert(start.pItems.pItem.pDual.pTwo1 !is null);
+    assert(start.pItems.pItem.pDual.pOne2 !is null);
+    assert(start.pItems.pItem.pDual.pTwo2 is null);
+    assert(start.pItems.pItem.pDual.pOne1 is null);
+}

spec/test_ast_field_aliases.c

@@ -0,0 +1,19 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "\na\nb\nc";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+
+    assert_eq(TOKEN_a, start->first->pToken->token);
+    assert_eq(TOKEN_b, start->second->pToken->token);
+    assert_eq(TOKEN_c, start->third->pToken->token);
+
+    return 0;
+}

spec/test_ast_field_aliases.d

@@ -0,0 +1,21 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "\na\nb\nc";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+
+    assert_eq(TOKEN_a, start.first.pToken.token);
+    assert_eq(TOKEN_b, start.second.pToken.token);
+    assert_eq(TOKEN_c, start.third.pToken.token);
+}

spec/test_ast_invalid_positions.c

@@ -0,0 +1,102 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "\na\n  bb ccc";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+
+    assert_eq(2, start->pT1->pToken->position.row);
+    assert_eq(1, start->pT1->pToken->position.col);
+    assert_eq(2, start->pT1->pToken->end_position.row);
+    assert_eq(1, start->pT1->pToken->end_position.col);
+    assert(p_position_valid(start->pT1->pA->position));
+    assert_eq(3, start->pT1->pA->position.row);
+    assert_eq(3, start->pT1->pA->position.col);
+    assert_eq(3, start->pT1->pA->end_position.row);
+    assert_eq(8, start->pT1->pA->end_position.col);
+    assert_eq(2, start->pT1->position.row);
+    assert_eq(1, start->pT1->position.col);
+    assert_eq(3, start->pT1->end_position.row);
+    assert_eq(8, start->pT1->end_position.col);
+
+    assert_eq(2, start->position.row);
+    assert_eq(1, start->position.col);
+    assert_eq(3, start->end_position.row);
+    assert_eq(8, start->end_position.col);
+
+    input = "a\nbb";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+
+    assert_eq(1, start->pT1->pToken->position.row);
+    assert_eq(1, start->pT1->pToken->position.col);
+    assert_eq(1, start->pT1->pToken->end_position.row);
+    assert_eq(1, start->pT1->pToken->end_position.col);
+    assert(p_position_valid(start->pT1->pA->position));
+    assert_eq(2, start->pT1->pA->position.row);
+    assert_eq(1, start->pT1->pA->position.col);
+    assert_eq(2, start->pT1->pA->end_position.row);
+    assert_eq(2, start->pT1->pA->end_position.col);
+    assert_eq(1, start->pT1->position.row);
+    assert_eq(1, start->pT1->position.col);
+    assert_eq(2, start->pT1->end_position.row);
+    assert_eq(2, start->pT1->end_position.col);
+
+    assert_eq(1, start->position.row);
+    assert_eq(1, start->position.col);
+    assert_eq(2, start->end_position.row);
+    assert_eq(2, start->end_position.col);
+
+    input = "a\nc\nc";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+
+    assert_eq(1, start->pT1->pToken->position.row);
+    assert_eq(1, start->pT1->pToken->position.col);
+    assert_eq(1, start->pT1->pToken->end_position.row);
+    assert_eq(1, start->pT1->pToken->end_position.col);
+    assert(p_position_valid(start->pT1->pA->position));
+    assert_eq(2, start->pT1->pA->position.row);
+    assert_eq(1, start->pT1->pA->position.col);
+    assert_eq(3, start->pT1->pA->end_position.row);
+    assert_eq(1, start->pT1->pA->end_position.col);
+    assert_eq(1, start->pT1->position.row);
+    assert_eq(1, start->pT1->position.col);
+    assert_eq(3, start->pT1->end_position.row);
+    assert_eq(1, start->pT1->end_position.col);
+
+    assert_eq(1, start->position.row);
+    assert_eq(1, start->position.col);
+    assert_eq(3, start->end_position.row);
+    assert_eq(1, start->end_position.col);
+
+    input = "a";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+
+    assert_eq(1, start->pT1->pToken->position.row);
+    assert_eq(1, start->pT1->pToken->position.col);
+    assert_eq(1, start->pT1->pToken->end_position.row);
+    assert_eq(1, start->pT1->pToken->end_position.col);
+    assert(!p_position_valid(start->pT1->pA->position));
+    assert_eq(1, start->pT1->position.row);
+    assert_eq(1, start->pT1->position.col);
+    assert_eq(1, start->pT1->end_position.row);
+    assert_eq(1, start->pT1->end_position.col);
+
+    assert_eq(1, start->position.row);
+    assert_eq(1, start->position.col);
+    assert_eq(1, start->end_position.row);
+    assert_eq(1, start->end_position.col);
+
+    return 0;
+}

spec/test_ast_invalid_positions.d

@@ -0,0 +1,104 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "\na\n  bb ccc";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+
+    assert_eq(2, start.pT1.pToken.position.row);
+    assert_eq(1, start.pT1.pToken.position.col);
+    assert_eq(2, start.pT1.pToken.end_position.row);
+    assert_eq(1, start.pT1.pToken.end_position.col);
+    assert(start.pT1.pA.position.valid);
+    assert_eq(3, start.pT1.pA.position.row);
+    assert_eq(3, start.pT1.pA.position.col);
+    assert_eq(3, start.pT1.pA.end_position.row);
+    assert_eq(8, start.pT1.pA.end_position.col);
+    assert_eq(2, start.pT1.position.row);
+    assert_eq(1, start.pT1.position.col);
+    assert_eq(3, start.pT1.end_position.row);
+    assert_eq(8, start.pT1.end_position.col);
+
+    assert_eq(2, start.position.row);
+    assert_eq(1, start.position.col);
+    assert_eq(3, start.end_position.row);
+    assert_eq(8, start.end_position.col);
+
+    input = "a\nbb";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+
+    assert_eq(1, start.pT1.pToken.position.row);
+    assert_eq(1, start.pT1.pToken.position.col);
+    assert_eq(1, start.pT1.pToken.end_position.row);
+    assert_eq(1, start.pT1.pToken.end_position.col);
+    assert(start.pT1.pA.position.valid);
+    assert_eq(2, start.pT1.pA.position.row);
+    assert_eq(1, start.pT1.pA.position.col);
+    assert_eq(2, start.pT1.pA.end_position.row);
+    assert_eq(2, start.pT1.pA.end_position.col);
+    assert_eq(1, start.pT1.position.row);
+    assert_eq(1, start.pT1.position.col);
+    assert_eq(2, start.pT1.end_position.row);
+    assert_eq(2, start.pT1.end_position.col);
+
+    assert_eq(1, start.position.row);
+    assert_eq(1, start.position.col);
+    assert_eq(2, start.end_position.row);
+    assert_eq(2, start.end_position.col);
+
+    input = "a\nc\nc";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+
+    assert_eq(1, start.pT1.pToken.position.row);
+    assert_eq(1, start.pT1.pToken.position.col);
+    assert_eq(1, start.pT1.pToken.end_position.row);
+    assert_eq(1, start.pT1.pToken.end_position.col);
+    assert(start.pT1.pA.position.valid);
+    assert_eq(2, start.pT1.pA.position.row);
+    assert_eq(1, start.pT1.pA.position.col);
+    assert_eq(3, start.pT1.pA.end_position.row);
+    assert_eq(1, start.pT1.pA.end_position.col);
+    assert_eq(1, start.pT1.position.row);
+    assert_eq(1, start.pT1.position.col);
+    assert_eq(3, start.pT1.end_position.row);
+    assert_eq(1, start.pT1.end_position.col);
+
+    assert_eq(1, start.position.row);
+    assert_eq(1, start.position.col);
+    assert_eq(3, start.end_position.row);
+    assert_eq(1, start.end_position.col);
+
+    input = "a";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+
+    assert_eq(1, start.pT1.pToken.position.row);
+    assert_eq(1, start.pT1.pToken.position.col);
+    assert_eq(1, start.pT1.pToken.end_position.row);
+    assert_eq(1, start.pT1.pToken.end_position.col);
+    assert(!start.pT1.pA.position.valid);
+    assert_eq(1, start.pT1.position.row);
+    assert_eq(1, start.pT1.position.col);
+    assert_eq(1, start.pT1.end_position.row);
+    assert_eq(1, start.pT1.end_position.col);
+
+    assert_eq(1, start.position.row);
+    assert_eq(1, start.position.col);
+    assert_eq(1, start.end_position.row);
+    assert_eq(1, start.end_position.col);
+}

spec/test_ast_node_memory_remains.c

@@ -0,0 +1,415 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include <stdlib.h>
+#include "testutils.h"
+
+int main(int argc, char * argv[])
+{
+    const char * input =
+        "# 0\n"
+        "def byte_val() -> byte\n"
+        "{\n"
+        "    return 0x42;\n"
+        "}\n"
+        "\n"
+        "# 1\n"
+        "def short_val() -> short\n"
+        "{\n"
+        "    return 0x4242;\n"
+        "}\n"
+        "\n"
+        "# 2\n"
+        "def int_val() -> int\n"
+        "{\n"
+        "    return 0x42424242;\n"
+        "}\n"
+        "\n"
+        "# 3\n"
+        "def long_val() -> long\n"
+        "{\n"
+        "    return 0x4242_4242_4242_4242;\n"
+        "}\n"
+        "\n"
+        "# 4\n"
+        "def ssize_t_val() -> ssize_t\n"
+        "{\n"
+        "    return 0x42424242;\n"
+        "}\n"
+        "\n"
+        "# 5\n"
+        "def byte_to_short() -> short\n"
+        "{\n"
+        "    return byte_val();\n"
+        "}\n"
+        "\n"
+        "# 6\n"
+        "def byte_to_int() -> int\n"
+        "{\n"
+        "    return byte_val();\n"
+        "}\n"
+        "\n"
+        "# 7\n"
+        "def byte_to_long() -> long\n"
+        "{\n"
+        "    return byte_val();\n"
+        "}\n"
+        "\n"
+        "# 8\n"
+        "def byte_to_ssize_t() -> ssize_t\n"
+        "{\n"
+        "    return byte_val();\n"
+        "}\n"
+        "\n"
+        "# 9\n"
+        "def short_to_byte() -> byte\n"
+        "{\n"
+        "    return short_val();\n"
+        "}\n"
+        "\n"
+        "# 10\n"
+        "def short_to_int() -> int\n"
+        "{\n"
+        "    return short_val();\n"
+        "}\n"
+        "\n"
+        "# 11\n"
+        "def short_to_long() -> long\n"
+        "{\n"
+        "    return short_val();\n"
+        "}\n"
+        "\n"
+        "# 12\n"
+        "def short_to_ssize_t() -> ssize_t\n"
+        "{\n"
+        "    return short_val();\n"
+        "}\n"
+        "\n"
+        "# 13\n"
+        "def int_to_byte() -> byte\n"
+        "{\n"
+        "    return int_val();\n"
+        "}\n"
+        "\n"
+        "# 14\n"
+        "def int_to_short() -> short\n"
+        "{\n"
+        "    return int_val();\n"
+        "}\n"
+        "\n"
+        "# 15\n"
+        "def int_to_long() -> long\n"
+        "{\n"
+        "    return int_val();\n"
+        "}\n"
+        "\n"
+        "# 16\n"
+        "def int_to_ssize_t() -> ssize_t\n"
+        "{\n"
+        "    return int_val();\n"
+        "}\n"
+        "\n"
+        "# 17\n"
+        "def long_to_byte() -> byte\n"
+        "{\n"
+        "    return long_val();\n"
+        "}\n"
+        "\n"
+        "# 18\n"
+        "def long_to_short() -> short\n"
+        "{\n"
+        "    return long_val();\n"
+        "}\n"
+        "\n"
+        "# 19\n"
+        "def long_to_int() -> int\n"
+        "{\n"
+        "    return long_val();\n"
+        "}\n"
+        "\n"
+        "# 20\n"
+        "def long_to_ssize_t() -> ssize_t\n"
+        "{\n"
+        "    return long_val();\n"
+        "}\n"
+        "\n"
+        "# 21\n"
+        "def ssize_t_to_byte() -> byte\n"
+        "{\n"
+        "    return ssize_t_val();\n"
+        "}\n"
+        "\n"
+        "# 22\n"
+        "def ssize_t_to_short() -> short\n"
+        "{\n"
+        "    return ssize_t_val();\n"
+        "}\n"
+        "\n"
+        "# 23\n"
+        "def ssize_t_to_int() -> int\n"
+        "{\n"
+        "    return ssize_t_val();\n"
+        "}\n"
+        "\n"
+        "# 24\n"
+        "def ssize_t_to_long() -> long\n"
+        "{\n"
+        "    return ssize_t_val();\n"
+        "}\n"
+        "\n"
+        "# 25\n"
+        "def ubyte_val() -> ubyte\n"
+        "{\n"
+        "    return 0x42;\n"
+        "}\n"
+        "\n"
+        "# 26\n"
+        "def ushort_val() -> ushort\n"
+        "{\n"
+        "    return 0x4242;\n"
+        "}\n"
+        "\n"
+        "# 27\n"
+        "def uint_val() -> uint\n"
+        "{\n"
+        "    return 0x42424242;\n"
+        "}\n"
+        "\n"
+        "# 28\n"
+        "def ulong_val() -> ulong\n"
+        "{\n"
+        "    return 0x4242_4242_4242_4242;\n"
+        "}\n"
+        "\n"
+        "# 29\n"
+        "def size_t_val() -> size_t\n"
+        "{\n"
+        "    return 0x42424242;\n"
+        "}\n"
+        "\n"
+        "# 30\n"
+        "def ubyte_to_ushort() -> ushort\n"
+        "{\n"
+        "    return ubyte_val();\n"
+        "}\n"
+        "\n"
+        "# 31\n"
+        "def ubyte_to_uint() -> uint\n"
+        "{\n"
+        "    return ubyte_val();\n"
+        "}\n"
+        "\n"
+        "# 32\n"
+        "def ubyte_to_ulong() -> ulong\n"
+        "{\n"
+        "    return ubyte_val();\n"
+        "}\n"
+        "\n"
+        "# 33\n"
+        "def ubyte_to_size_t() -> size_t\n"
+        "{\n"
+        "    return ubyte_val();\n"
+        "}\n"
+        "\n"
+        "# 34\n"
+        "def ushort_to_ubyte() -> ubyte\n"
+        "{\n"
+        "    return ushort_val();\n"
+        "}\n"
+        "\n"
+        "# 35\n"
+        "def ushort_to_uint() -> uint\n"
+        "{\n"
+        "    return ushort_val();\n"
+        "}\n"
+        "\n"
+        "# 36\n"
+        "def ushort_to_ulong() -> ulong\n"
+        "{\n"
+        "    return ushort_val();\n"
+        "}\n"
+        "\n"
+        "# 37\n"
+        "def ushort_to_size_t() -> size_t\n"
+        "{\n"
+        "    return ushort_val();\n"
+        "}\n"
+        "\n"
+        "# 38\n"
+        "def uint_to_ubyte() -> ubyte\n"
+        "{\n"
+        "    return uint_val();\n"
+        "}\n"
+        "\n"
+        "# 39\n"
+        "def uint_to_ushort() -> ushort\n"
+        "{\n"
+        "    return uint_val();\n"
+        "}\n"
+        "\n"
+        "# 40\n"
+        "def uint_to_ulong() -> ulong\n"
+        "{\n"
+        "    return uint_val();\n"
+        "}\n"
+        "\n"
+        "# 41\n"
+        "def uint_to_size_t() -> size_t\n"
+        "{\n"
+        "    return uint_val();\n"
+        "}\n"
+        "\n"
+        "# 42\n"
+        "def ulong_to_ubyte() -> ubyte\n"
+        "{\n"
+        "    return ulong_val();\n"
+        "}\n"
+        "\n"
+        "# 43\n"
+        "def ulong_to_ushort() -> ushort\n"
+        "{\n"
+        "    return ulong_val();\n"
+        "}\n"
+        "\n"
+        "# 44\n"
+        "def ulong_to_uint() -> uint\n"
+        "{\n"
+        "    return ulong_val();\n"
+        "}\n"
+        "\n"
+        "# 45\n"
+        "def ulong_to_size_t() -> size_t\n"
+        "{\n"
+        "    return ulong_val();\n"
+        "}\n"
+        "\n"
+        "# 46\n"
+        "def size_t_to_ubyte() -> ubyte\n"
+        "{\n"
+        "    return size_t_val();\n"
+        "}\n"
+        "\n"
+        "# 47\n"
+        "def size_t_to_ushort() -> ushort\n"
+        "{\n"
+        "    return size_t_val();\n"
+        "}\n"
+        "\n"
+        "# 48\n"
+        "def size_t_to_int() -> int\n"
+        "{\n"
+        "    return size_t_val();\n"
+        "}\n"
+        "\n"
+        "# 49\n"
+        "def size_t_to_ulong() -> ulong\n"
+        "{\n"
+        "    return size_t_val();\n"
+        "}\n"
+        "\n"
+        "# 50\n"
+        "def main() -> int\n"
+        "{\n"
+        "    return int_val();\n"
+        "}\n";
+    struct
+    {
+        const char * name;
+        p_token_t token;
+    } expected[] = {
+        {"byte_val", TOKEN_byte},
+        {"short_val", TOKEN_short},
+        {"int_val", TOKEN_int},
+        {"long_val", TOKEN_long},
+        {"ssize_t_val", TOKEN_ssize_t},
+        {"byte_to_short", TOKEN_short},
+        {"byte_to_int", TOKEN_int},
+        {"byte_to_long", TOKEN_long},
+        {"byte_to_ssize_t", TOKEN_ssize_t},
+        {"short_to_byte", TOKEN_byte},
+        {"short_to_int", TOKEN_int},
+        {"short_to_long", TOKEN_long},
+        {"short_to_ssize_t", TOKEN_ssize_t},
+        {"int_to_byte", TOKEN_byte},
+        {"int_to_short", TOKEN_short},
+        {"int_to_long", TOKEN_long},
+        {"int_to_ssize_t", TOKEN_ssize_t},
+        {"long_to_byte", TOKEN_byte},
+        {"long_to_short", TOKEN_short},
+        {"long_to_int", TOKEN_int},
+        {"long_to_ssize_t", TOKEN_ssize_t},
+        {"ssize_t_to_byte", TOKEN_byte},
+        {"ssize_t_to_short", TOKEN_short},
+        {"ssize_t_to_int", TOKEN_int},
+        {"ssize_t_to_long", TOKEN_long},
+        {"ubyte_val", TOKEN_ubyte},
+        {"ushort_val", TOKEN_ushort},
+        {"uint_val", TOKEN_uint},
+        {"ulong_val", TOKEN_ulong},
+        {"size_t_val", TOKEN_size_t},
+        {"ubyte_to_ushort", TOKEN_ushort},
+        {"ubyte_to_uint", TOKEN_uint},
+        {"ubyte_to_ulong", TOKEN_ulong},
+        {"ubyte_to_size_t", TOKEN_size_t},
+        {"ushort_to_ubyte", TOKEN_ubyte},
+        {"ushort_to_uint", TOKEN_uint},
+        {"ushort_to_ulong", TOKEN_ulong},
+        {"ushort_to_size_t", TOKEN_size_t},
+        {"uint_to_ubyte", TOKEN_ubyte},
+        {"uint_to_ushort", TOKEN_ushort},
+        {"uint_to_ulong", TOKEN_ulong},
+        {"uint_to_size_t", TOKEN_size_t},
+        {"ulong_to_ubyte", TOKEN_ubyte},
+        {"ulong_to_ushort", TOKEN_ushort},
+        {"ulong_to_uint", TOKEN_uint},
+        {"ulong_to_size_t", TOKEN_size_t},
+        {"size_t_to_ubyte", TOKEN_ubyte},
+        {"size_t_to_ushort", TOKEN_ushort},
+        {"size_t_to_int", TOKEN_int},
+        {"size_t_to_ulong", TOKEN_ulong},
+        {"main", TOKEN_int},
+    };
+    p_context_t context;
+    p_context_init(&context, (const uint8_t *)input, strlen(input));
+    size_t result = p_parse(&context);
+    assert_eq(P_SUCCESS, result);
+    PModule * pmod = p_result(&context);
+    PModuleItems * pmis = pmod->pModuleItems;
+    PFunctionDefinition ** pfds;
+    size_t n_pfds = 0u;
+    while (pmis != NULL)
+    {
+        PModuleItem * pmi = pmis->pModuleItem;
+        if (pmi->pFunctionDefinition != NULL)
+        {
+            n_pfds++;
+        }
+        pmis = pmis->pModuleItems;
+    }
+    pfds = (PFunctionDefinition **)malloc(n_pfds * sizeof(PModuleItems *));
+    pmis = pmod->pModuleItems;
+    size_t pfd_i = n_pfds;
+    while (pmis != NULL)
+    {
+        PModuleItem * pmi = pmis->pModuleItem;
+        PFunctionDefinition * pfd = pmi->pFunctionDefinition;
+        if (pfd != NULL)
+        {
+            pfd_i--;
+            assert(pfd_i < n_pfds);
+            pfds[pfd_i] = pfd;
+        }
+        pmis = pmis->pModuleItems;
+    }
+    assert_eq(51, n_pfds);
+    for (size_t i = 0; i < n_pfds; i++)
+    {
+        if (strncmp(expected[i].name, (const char *)pfds[i]->name->pvalue.s, strlen(expected[i].name)) != 0 ||
+            (expected[i].token != pfds[i]->returntype->pType->pTypeBase->pToken1->token))
+        {
+            fprintf(stderr, "Index %lu: expected %s/%u, got %u\n", i, expected[i].name, expected[i].token, pfds[i]->returntype->pType->pTypeBase->pToken1->token);
+        }
+    }
+
+    return 0;
+}

spec/test_ast_node_memory_remains.d

@@ -0,0 +1,408 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "
+# 0
+def byte_val() -> byte
+{
+    return 0x42;
+}
+
+# 1
+def short_val() -> short
+{
+    return 0x4242;
+}
+
+# 2
+def int_val() -> int
+{
+    return 0x42424242;
+}
+
+# 3
+def long_val() -> long
+{
+    return 0x4242_4242_4242_4242;
+}
+
+# 4
+def ssize_t_val() -> ssize_t
+{
+    return 0x42424242;
+}
+
+# 5
+def byte_to_short() -> short
+{
+    return byte_val();
+}
+
+# 6
+def byte_to_int() -> int
+{
+    return byte_val();
+}
+
+# 7
+def byte_to_long() -> long
+{
+    return byte_val();
+}
+
+# 8
+def byte_to_ssize_t() -> ssize_t
+{
+    return byte_val();
+}
+
+# 9
+def short_to_byte() -> byte
+{
+    return short_val();
+}
+
+# 10
+def short_to_int() -> int
+{
+    return short_val();
+}
+
+# 11
+def short_to_long() -> long
+{
+    return short_val();
+}
+
+# 12
+def short_to_ssize_t() -> ssize_t
+{
+    return short_val();
+}
+
+# 13
+def int_to_byte() -> byte
+{
+    return int_val();
+}
+
+# 14
+def int_to_short() -> short
+{
+    return int_val();
+}
+
+# 15
+def int_to_long() -> long
+{
+    return int_val();
+}
+
+# 16
+def int_to_ssize_t() -> ssize_t
+{
+    return int_val();
+}
+
+# 17
+def long_to_byte() -> byte
+{
+    return long_val();
+}
+
+# 18
+def long_to_short() -> short
+{
+    return long_val();
+}
+
+# 19
+def long_to_int() -> int
+{
+    return long_val();
+}
+
+# 20
+def long_to_ssize_t() -> ssize_t
+{
+    return long_val();
+}
+
+# 21
+def ssize_t_to_byte() -> byte
+{
+    return ssize_t_val();
+}
+
+# 22
+def ssize_t_to_short() -> short
+{
+    return ssize_t_val();
+}
+
+# 23
+def ssize_t_to_int() -> int
+{
+    return ssize_t_val();
+}
+
+# 24
+def ssize_t_to_long() -> long
+{
+    return ssize_t_val();
+}
+
+# 25
+def ubyte_val() -> ubyte
+{
+    return 0x42;
+}
+
+# 26
+def ushort_val() -> ushort
+{
+    return 0x4242;
+}
+
+# 27
+def uint_val() -> uint
+{
+    return 0x42424242;
+}
+
+# 28
+def ulong_val() -> ulong
+{
+    return 0x4242_4242_4242_4242;
+}
+
+# 29
+def size_t_val() -> size_t
+{
+    return 0x42424242;
+}
+
+# 30
+def ubyte_to_ushort() -> ushort
+{
+    return ubyte_val();
+}
+
+# 31
+def ubyte_to_uint() -> uint
+{
+    return ubyte_val();
+}
+
+# 32
+def ubyte_to_ulong() -> ulong
+{
+    return ubyte_val();
+}
+
+# 33
+def ubyte_to_size_t() -> size_t
+{
+    return ubyte_val();
+}
+
+# 34
+def ushort_to_ubyte() -> ubyte
+{
+    return ushort_val();
+}
+
+# 35
+def ushort_to_uint() -> uint
+{
+    return ushort_val();
+}
+
+# 36
+def ushort_to_ulong() -> ulong
+{
+    return ushort_val();
+}
+
+# 37
+def ushort_to_size_t() -> size_t
+{
+    return ushort_val();
+}
+
+# 38
+def uint_to_ubyte() -> ubyte
+{
+    return uint_val();
+}
+
+# 39
+def uint_to_ushort() -> ushort
+{
+    return uint_val();
+}
+
+# 40
+def uint_to_ulong() -> ulong
+{
+    return uint_val();
+}
+
+# 41
+def uint_to_size_t() -> size_t
+{
+    return uint_val();
+}
+
+# 42
+def ulong_to_ubyte() -> ubyte
+{
+    return ulong_val();
+}
+
+# 43
+def ulong_to_ushort() -> ushort
+{
+    return ulong_val();
+}
+
+# 44
+def ulong_to_uint() -> uint
+{
+    return ulong_val();
+}
+
+# 45
+def ulong_to_size_t() -> size_t
+{
+    return ulong_val();
+}
+
+# 46
+def size_t_to_ubyte() -> ubyte
+{
+    return size_t_val();
+}
+
+# 47
+def size_t_to_ushort() -> ushort
+{
+    return size_t_val();
+}
+
+# 48
+def size_t_to_int() -> int
+{
+    return size_t_val();
+}
+
+# 49
+def size_t_to_ulong() -> ulong
+{
+    return size_t_val();
+}
+
+# 50
+def main() -> int
+{
+    return int_val();
+}
+";
+    struct Expected
+    {
+        string name;
+        p_token_t token;
+    }
+    Expected[] expected = [
+        Expected("byte_val", TOKEN_byte),
+        Expected("short_val", TOKEN_short),
+        Expected("int_val", TOKEN_int),
+        Expected("long_val", TOKEN_long),
+        Expected("ssize_t_val", TOKEN_ssize_t),
+        Expected("byte_to_short", TOKEN_short),
+        Expected("byte_to_int", TOKEN_int),
+        Expected("byte_to_long", TOKEN_long),
+        Expected("byte_to_ssize_t", TOKEN_ssize_t),
+        Expected("short_to_byte", TOKEN_byte),
+        Expected("short_to_int", TOKEN_int),
+        Expected("short_to_long", TOKEN_long),
+        Expected("short_to_ssize_t", TOKEN_ssize_t),
+        Expected("int_to_byte", TOKEN_byte),
+        Expected("int_to_short", TOKEN_short),
+        Expected("int_to_long", TOKEN_long),
+        Expected("int_to_ssize_t", TOKEN_ssize_t),
+        Expected("long_to_byte", TOKEN_byte),
+        Expected("long_to_short", TOKEN_short),
+        Expected("long_to_int", TOKEN_int),
+        Expected("long_to_ssize_t", TOKEN_ssize_t),
+        Expected("ssize_t_to_byte", TOKEN_byte),
+        Expected("ssize_t_to_short", TOKEN_short),
+        Expected("ssize_t_to_int", TOKEN_int),
+        Expected("ssize_t_to_long", TOKEN_long),
+        Expected("ubyte_val", TOKEN_ubyte),
+        Expected("ushort_val", TOKEN_ushort),
+        Expected("uint_val", TOKEN_uint),
+        Expected("ulong_val", TOKEN_ulong),
+        Expected("size_t_val", TOKEN_size_t),
+        Expected("ubyte_to_ushort", TOKEN_ushort),
+        Expected("ubyte_to_uint", TOKEN_uint),
+        Expected("ubyte_to_ulong", TOKEN_ulong),
+        Expected("ubyte_to_size_t", TOKEN_size_t),
+        Expected("ushort_to_ubyte", TOKEN_ubyte),
+        Expected("ushort_to_uint", TOKEN_uint),
+        Expected("ushort_to_ulong", TOKEN_ulong),
+        Expected("ushort_to_size_t", TOKEN_size_t),
+        Expected("uint_to_ubyte", TOKEN_ubyte),
+        Expected("uint_to_ushort", TOKEN_ushort),
+        Expected("uint_to_ulong", TOKEN_ulong),
+        Expected("uint_to_size_t", TOKEN_size_t),
+        Expected("ulong_to_ubyte", TOKEN_ubyte),
+        Expected("ulong_to_ushort", TOKEN_ushort),
+        Expected("ulong_to_uint", TOKEN_uint),
+        Expected("ulong_to_size_t", TOKEN_size_t),
+        Expected("size_t_to_ubyte", TOKEN_ubyte),
+        Expected("size_t_to_ushort", TOKEN_ushort),
+        Expected("size_t_to_int", TOKEN_int),
+        Expected("size_t_to_ulong", TOKEN_ulong),
+        Expected("main", TOKEN_int),
+    ];
+    p_context_t context;
+    p_context_init(&context, input);
+    size_t result = p_parse(&context);
+    assert_eq(P_SUCCESS, result);
+    PModule * pmod = p_result(&context);
+    PModuleItems * pmis = pmod.pModuleItems;
+    PFunctionDefinition *[] pfds;
+    while (pmis !is null)
+    {
+        PModuleItem * pmi = pmis.pModuleItem;
+        if (pmi is null)
+        {
+            stderr.writeln("pmi is null!!!?");
+            assert(0);
+        }
+        PFunctionDefinition * pfd = pmi.pFunctionDefinition;
+        if (pfd !is null)
+        {
+            pfds = [pfd] ~ pfds;
+        }
+        pmis = pmis.pModuleItems;
+    }
+    assert_eq(51, pfds.length);
+    for (size_t i = 0; i < pfds.length; i++)
+    {
+        if ((expected[i].name != pfds[i].name.pvalue.s) ||
+            (expected[i].token != pfds[i].returntype.pType.pTypeBase.pToken1.token))
+        {
+            stderr.writeln("Index ", i, ": expected ", expected[i].name, "/", expected[i].token, ", got ", pfds[i].name.pvalue.s, "/", pfds[i].returntype.pType.pTypeBase.pToken1.token);
+        }
+    }
+}

spec/test_ast_ps.c

@@ -0,0 +1,55 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "a, ((b)), b";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert_eq(P_SUCCESS, p_parse(&context));
+    PStartS * start = p_result(&context);
+    assert(start->pItems1 != NULL);
+    assert(start->pItems != NULL);
+    PItemsS * items = start->pItems;
+    assert(items->pItem != NULL);
+    assert(items->pItem->pToken1 != NULL);
+    assert_eq(TOKEN_a, items->pItem->pToken1->token);
+    assert_eq(11, items->pItem->pToken1->pvalue);
+    assert(items->pItemsMore != NULL);
+    PItemsMoreS * itemsmore = items->pItemsMore;
+    assert(itemsmore->pItem != NULL);
+    assert(itemsmore->pItem->pItem != NULL);
+    assert(itemsmore->pItem->pItem->pItem != NULL);
+    assert(itemsmore->pItem->pItem->pItem->pToken1 != NULL);
+    assert_eq(TOKEN_b, itemsmore->pItem->pItem->pItem->pToken1->token);
+    assert_eq(22, itemsmore->pItem->pItem->pItem->pToken1->pvalue);
+    assert(itemsmore->pItemsMore != NULL);
+    itemsmore = itemsmore->pItemsMore;
+    assert(itemsmore->pItem != NULL);
+    assert(itemsmore->pItem->pToken1 != NULL);
+    assert_eq(TOKEN_b, itemsmore->pItem->pToken1->token);
+    assert_eq(22, itemsmore->pItem->pToken1->pvalue);
+    assert(itemsmore->pItemsMore == NULL);
+
+    input = "";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert_eq(P_SUCCESS, p_parse(&context));
+    start = p_result(&context);
+    assert(start->pItems == NULL);
+
+    input = "2 1";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert_eq(P_SUCCESS, p_parse(&context));
+    start = p_result(&context);
+    assert(start->pItems != NULL);
+    assert(start->pItems->pItem != NULL);
+    assert(start->pItems->pItem->pDual != NULL);
+    assert(start->pItems->pItem->pDual->pTwo1 != NULL);
+    assert(start->pItems->pItem->pDual->pOne2 != NULL);
+    assert(start->pItems->pItem->pDual->pTwo2 == NULL);
+    assert(start->pItems->pItem->pDual->pOne1 == NULL);
+
+    return 0;
+}

spec/test_ast_ps.d

@@ -0,0 +1,57 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "a, ((b)), b";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert_eq(P_SUCCESS, p_parse(&context));
+    PStartS * start = p_result(&context);
+    assert(start.pItems1 !is null);
+    assert(start.pItems !is null);
+    PItemsS * items = start.pItems;
+    assert(items.pItem !is null);
+    assert(items.pItem.pToken1 !is null);
+    assert_eq(TOKEN_a, items.pItem.pToken1.token);
+    assert_eq(11, items.pItem.pToken1.pvalue);
+    assert(items.pItemsMore !is null);
+    PItemsMoreS * itemsmore = items.pItemsMore;
+    assert(itemsmore.pItem !is null);
+    assert(itemsmore.pItem.pItem !is null);
+    assert(itemsmore.pItem.pItem.pItem !is null);
+    assert(itemsmore.pItem.pItem.pItem.pToken1 !is null);
+    assert_eq(TOKEN_b, itemsmore.pItem.pItem.pItem.pToken1.token);
+    assert_eq(22, itemsmore.pItem.pItem.pItem.pToken1.pvalue);
+    assert(itemsmore.pItemsMore !is null);
+    itemsmore = itemsmore.pItemsMore;
+    assert(itemsmore.pItem !is null);
+    assert(itemsmore.pItem.pToken1 !is null);
+    assert_eq(TOKEN_b, itemsmore.pItem.pToken1.token);
+    assert_eq(22, itemsmore.pItem.pToken1.pvalue);
+    assert(itemsmore.pItemsMore is null);
+
+    input = "";
+    p_context_init(&context, input);
+    assert_eq(P_SUCCESS, p_parse(&context));
+    start = p_result(&context);
+    assert(start.pItems is null);
+
+    input = "2 1";
+    p_context_init(&context, input);
+    assert_eq(P_SUCCESS, p_parse(&context));
+    start = p_result(&context);
+    assert(start.pItems !is null);
+    assert(start.pItems.pItem !is null);
+    assert(start.pItems.pItem.pDual !is null);
+    assert(start.pItems.pItem.pDual.pTwo1 !is null);
+    assert(start.pItems.pItem.pDual.pOne2 !is null);
+    assert(start.pItems.pItem.pDual.pTwo2 is null);
+    assert(start.pItems.pItem.pDual.pOne1 is null);
+}

spec/test_ast_token_positions.c

@@ -0,0 +1,84 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "abbccc";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+
+    assert_eq(1, start->pT1->pToken->position.row);
+    assert_eq(1, start->pT1->pToken->position.col);
+    assert_eq(1, start->pT1->pToken->end_position.row);
+    assert_eq(1, start->pT1->pToken->end_position.col);
+    assert_eq(1, start->pT1->position.row);
+    assert_eq(1, start->pT1->position.col);
+    assert_eq(1, start->pT1->end_position.row);
+    assert_eq(1, start->pT1->end_position.col);
+
+    assert_eq(1, start->pT2->pToken->position.row);
+    assert_eq(2, start->pT2->pToken->position.col);
+    assert_eq(1, start->pT2->pToken->end_position.row);
+    assert_eq(3, start->pT2->pToken->end_position.col);
+    assert_eq(1, start->pT2->position.row);
+    assert_eq(2, start->pT2->position.col);
+    assert_eq(1, start->pT2->end_position.row);
+    assert_eq(3, start->pT2->end_position.col);
+
+    assert_eq(1, start->pT3->pToken->position.row);
+    assert_eq(4, start->pT3->pToken->position.col);
+    assert_eq(1, start->pT3->pToken->end_position.row);
+    assert_eq(6, start->pT3->pToken->end_position.col);
+    assert_eq(1, start->pT3->position.row);
+    assert_eq(4, start->pT3->position.col);
+    assert_eq(1, start->pT3->end_position.row);
+    assert_eq(6, start->pT3->end_position.col);
+
+    assert_eq(1, start->position.row);
+    assert_eq(1, start->position.col);
+    assert_eq(1, start->end_position.row);
+    assert_eq(6, start->end_position.col);
+
+    input = "\n\n  bb\nc\ncc\n\n     a";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+
+    assert_eq(3, start->pT1->pToken->position.row);
+    assert_eq(3, start->pT1->pToken->position.col);
+    assert_eq(3, start->pT1->pToken->end_position.row);
+    assert_eq(4, start->pT1->pToken->end_position.col);
+    assert_eq(3, start->pT1->position.row);
+    assert_eq(3, start->pT1->position.col);
+    assert_eq(3, start->pT1->end_position.row);
+    assert_eq(4, start->pT1->end_position.col);
+
+    assert_eq(4, start->pT2->pToken->position.row);
+    assert_eq(1, start->pT2->pToken->position.col);
+    assert_eq(5, start->pT2->pToken->end_position.row);
+    assert_eq(2, start->pT2->pToken->end_position.col);
+    assert_eq(4, start->pT2->position.row);
+    assert_eq(1, start->pT2->position.col);
+    assert_eq(5, start->pT2->end_position.row);
+    assert_eq(2, start->pT2->end_position.col);
+
+    assert_eq(7, start->pT3->pToken->position.row);
+    assert_eq(6, start->pT3->pToken->position.col);
+    assert_eq(7, start->pT3->pToken->end_position.row);
+    assert_eq(6, start->pT3->pToken->end_position.col);
+    assert_eq(7, start->pT3->position.row);
+    assert_eq(6, start->pT3->position.col);
+    assert_eq(7, start->pT3->end_position.row);
+    assert_eq(6, start->pT3->end_position.col);
+
+    assert_eq(3, start->position.row);
+    assert_eq(3, start->position.col);
+    assert_eq(7, start->end_position.row);
+    assert_eq(6, start->end_position.col);
+
+    return 0;
+}

spec/test_ast_token_positions.d

@@ -0,0 +1,86 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "abbccc";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+
+    assert_eq(1, start.pT1.pToken.position.row);
+    assert_eq(1, start.pT1.pToken.position.col);
+    assert_eq(1, start.pT1.pToken.end_position.row);
+    assert_eq(1, start.pT1.pToken.end_position.col);
+    assert_eq(1, start.pT1.position.row);
+    assert_eq(1, start.pT1.position.col);
+    assert_eq(1, start.pT1.end_position.row);
+    assert_eq(1, start.pT1.end_position.col);
+
+    assert_eq(1, start.pT2.pToken.position.row);
+    assert_eq(2, start.pT2.pToken.position.col);
+    assert_eq(1, start.pT2.pToken.end_position.row);
+    assert_eq(3, start.pT2.pToken.end_position.col);
+    assert_eq(1, start.pT2.position.row);
+    assert_eq(2, start.pT2.position.col);
+    assert_eq(1, start.pT2.end_position.row);
+    assert_eq(3, start.pT2.end_position.col);
+
+    assert_eq(1, start.pT3.pToken.position.row);
+    assert_eq(4, start.pT3.pToken.position.col);
+    assert_eq(1, start.pT3.pToken.end_position.row);
+    assert_eq(6, start.pT3.pToken.end_position.col);
+    assert_eq(1, start.pT3.position.row);
+    assert_eq(4, start.pT3.position.col);
+    assert_eq(1, start.pT3.end_position.row);
+    assert_eq(6, start.pT3.end_position.col);
+
+    assert_eq(1, start.position.row);
+    assert_eq(1, start.position.col);
+    assert_eq(1, start.end_position.row);
+    assert_eq(6, start.end_position.col);
+
+    input = "\n\n  bb\nc\ncc\n\n     a";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+
+    assert_eq(3, start.pT1.pToken.position.row);
+    assert_eq(3, start.pT1.pToken.position.col);
+    assert_eq(3, start.pT1.pToken.end_position.row);
+    assert_eq(4, start.pT1.pToken.end_position.col);
+    assert_eq(3, start.pT1.position.row);
+    assert_eq(3, start.pT1.position.col);
+    assert_eq(3, start.pT1.end_position.row);
+    assert_eq(4, start.pT1.end_position.col);
+
+    assert_eq(4, start.pT2.pToken.position.row);
+    assert_eq(1, start.pT2.pToken.position.col);
+    assert_eq(5, start.pT2.pToken.end_position.row);
+    assert_eq(2, start.pT2.pToken.end_position.col);
+    assert_eq(4, start.pT2.position.row);
+    assert_eq(1, start.pT2.position.col);
+    assert_eq(5, start.pT2.end_position.row);
+    assert_eq(2, start.pT2.end_position.col);
+
+    assert_eq(7, start.pT3.pToken.position.row);
+    assert_eq(6, start.pT3.pToken.position.col);
+    assert_eq(7, start.pT3.pToken.end_position.row);
+    assert_eq(6, start.pT3.pToken.end_position.col);
+    assert_eq(7, start.pT3.position.row);
+    assert_eq(6, start.pT3.position.col);
+    assert_eq(7, start.pT3.end_position.row);
+    assert_eq(6, start.pT3.end_position.col);
+
+    assert_eq(3, start.position.row);
+    assert_eq(3, start.position.col);
+    assert_eq(7, start.end_position.row);
+    assert_eq(6, start.end_position.col);
+}

spec/test_error_positions.c

@@ -12,28 +12,31 @@ int main()
     input = "a\n123\na  a";
     p_context_init(&context, (uint8_t const *)input, strlen(input));
     assert(p_parse(&context) == P_UNEXPECTED_TOKEN);
-    assert(p_position(&context).row == 2);
-    assert(p_position(&context).col == 3);
-    assert(context.token == TOKEN_a);
+    assert(p_position(&context).row == 3);
+    assert(p_position(&context).col == 4);
+    assert(p_token(&context) == TOKEN_a);
 
     input = "12";
     p_context_init(&context, (uint8_t const *)input, strlen(input));
     assert(p_parse(&context) == P_UNEXPECTED_TOKEN);
-    assert(p_position(&context).row == 0);
-    assert(p_position(&context).col == 0);
-    assert(context.token == TOKEN_num);
+    assert(p_position(&context).row == 1);
+    assert(p_position(&context).col == 1);
+    assert(p_token(&context) == TOKEN_num);
 
     input = "a 12\n\nab";
     p_context_init(&context, (uint8_t const *)input, strlen(input));
     assert(p_parse(&context) == P_UNEXPECTED_INPUT);
-    assert(p_position(&context).row == 2);
-    assert(p_position(&context).col == 1);
+    assert(p_position(&context).row == 3);
+    assert(p_position(&context).col == 2);
 
     input = "a 12\n\na\n\n77\na   \xAA";
     p_context_init(&context, (uint8_t const *)input, strlen(input));
     assert(p_parse(&context) == P_DECODE_ERROR);
-    assert(p_position(&context).row == 5);
-    assert(p_position(&context).col == 4);
+    assert(p_position(&context).row == 6);
+    assert(p_position(&context).col == 5);
+
+    assert(strcmp(p_token_names[TOKEN_a], "a") == 0);
+    assert(strcmp(p_token_names[TOKEN_num], "num") == 0);
 
     return 0;
 }

spec/test_error_positions.d

@@ -16,22 +16,25 @@ unittest
     input = "a\n123\na  a";
     p_context_init(&context, input);
     assert(p_parse(&context) == P_UNEXPECTED_TOKEN);
-    assert(p_position(&context) == p_position_t(2, 3));
-    assert(context.token == TOKEN_a);
+    assert(p_position(&context) == p_position_t(3, 4));
+    assert(p_token(&context) == TOKEN_a);
 
     input = "12";
     p_context_init(&context, input);
     assert(p_parse(&context) == P_UNEXPECTED_TOKEN);
-    assert(p_position(&context) == p_position_t(0, 0));
-    assert(context.token == TOKEN_num);
+    assert(p_position(&context) == p_position_t(1, 1));
+    assert(p_token(&context) == TOKEN_num);
 
     input = "a 12\n\nab";
     p_context_init(&context, input);
     assert(p_parse(&context) == P_UNEXPECTED_INPUT);
-    assert(p_position(&context) == p_position_t(2, 1));
+    assert(p_position(&context) == p_position_t(3, 2));
 
     input = "a 12\n\na\n\n77\na   \xAA";
     p_context_init(&context, input);
     assert(p_parse(&context) == P_DECODE_ERROR);
-    assert(p_position(&context) == p_position_t(5, 4));
+    assert(p_position(&context) == p_position_t(6, 5));
+
+    assert(p_token_names[TOKEN_a] == "a");
+    assert(p_token_names[TOKEN_num] == "num");
 }

spec/test_field_aliases.c

@@ -0,0 +1,13 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "foo1\nbar2";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    return 0;
+}

spec/test_field_aliases.d

@@ -0,0 +1,15 @@
+import testparser;
+import std.stdio;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "foo1\nbar2";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+}

spec/test_lexer.c

@@ -41,50 +41,68 @@ int main()
     p_context_t context;
     p_context_init(&context, (uint8_t const *)input, strlen(input));
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 0u);
-    assert(token_info.position.col == 0u);
+    assert(token_info.position.row == 1u);
+    assert(token_info.position.col == 1u);
+    assert(token_info.end_position.row == 1u);
+    assert(token_info.end_position.col == 1u);
     assert(token_info.length == 1u);
     assert(token_info.token == TOKEN_int);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 0u);
-    assert(token_info.position.col == 2u);
+    assert(token_info.position.row == 1u);
+    assert(token_info.position.col == 3u);
+    assert(token_info.end_position.row == 1u);
+    assert(token_info.end_position.col == 3u);
     assert(token_info.length == 1u);
     assert(token_info.token == TOKEN_plus);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 0u);
-    assert(token_info.position.col == 4u);
+    assert(token_info.position.row == 1u);
+    assert(token_info.position.col == 5u);
+    assert(token_info.end_position.row == 1u);
+    assert(token_info.end_position.col == 5u);
     assert(token_info.length == 1u);
     assert(token_info.token == TOKEN_int);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 0u);
-    assert(token_info.position.col == 6u);
+    assert(token_info.position.row == 1u);
+    assert(token_info.position.col == 7u);
+    assert(token_info.end_position.row == 1u);
+    assert(token_info.end_position.col == 7u);
     assert(token_info.length == 1u);
     assert(token_info.token == TOKEN_times);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 1u);
-    assert(token_info.position.col == 0u);
+    assert(token_info.position.row == 2u);
+    assert(token_info.position.col == 1u);
+    assert(token_info.end_position.row == 2u);
+    assert(token_info.end_position.col == 3u);
     assert(token_info.length == 3u);
     assert(token_info.token == TOKEN_int);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 1u);
-    assert(token_info.position.col == 4u);
+    assert(token_info.position.row == 2u);
+    assert(token_info.position.col == 5u);
+    assert(token_info.end_position.row == 2u);
+    assert(token_info.end_position.col == 5u);
     assert(token_info.length == 1u);
     assert(token_info.token == TOKEN_plus);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 1u);
-    assert(token_info.position.col == 6u);
+    assert(token_info.position.row == 2u);
+    assert(token_info.position.col == 7u);
+    assert(token_info.end_position.row == 2u);
+    assert(token_info.end_position.col == 9u);
     assert(token_info.length == 3u);
     assert(token_info.token == TOKEN_int);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 1u);
-    assert(token_info.position.col == 9u);
+    assert(token_info.position.row == 2u);
+    assert(token_info.position.col == 10u);
+    assert(token_info.end_position.row == 2u);
+    assert(token_info.end_position.col == 10u);
     assert(token_info.length == 0u);
     assert(token_info.token == TOKEN___EOF);
 
     p_context_init(&context, (uint8_t const *)"", 0u);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info.position.row == 0u);
-    assert(token_info.position.col == 0u);
+    assert(token_info.position.row == 1u);
+    assert(token_info.position.col == 1u);
+    assert(token_info.end_position.row == 1u);
+    assert(token_info.end_position.col == 1u);
     assert(token_info.length == 0u);
     assert(token_info.token == TOKEN___EOF);
 

spec/test_lexer.d

@@ -47,23 +47,23 @@ unittest
     p_context_t context;
     p_context_init(&context, input);
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(0, 0), 1, TOKEN_int));
+    assert(token_info == p_token_info_t(p_position_t(1, 1), p_position_t(1, 1), 1, TOKEN_int));
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(0, 2), 1, TOKEN_plus));
+    assert(token_info == p_token_info_t(p_position_t(1, 3), p_position_t(1, 3), 1, TOKEN_plus));
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(0, 4), 1, TOKEN_int));
+    assert(token_info == p_token_info_t(p_position_t(1, 5), p_position_t(1, 5), 1, TOKEN_int));
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(0, 6), 1, TOKEN_times));
+    assert(token_info == p_token_info_t(p_position_t(1, 7), p_position_t(1, 7), 1, TOKEN_times));
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(1, 0), 3, TOKEN_int));
+    assert(token_info == p_token_info_t(p_position_t(2, 1), p_position_t(2, 3), 3, TOKEN_int));
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(1, 4), 1, TOKEN_plus));
+    assert(token_info == p_token_info_t(p_position_t(2, 5), p_position_t(2, 5), 1, TOKEN_plus));
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(1, 6), 3, TOKEN_int));
+    assert(token_info == p_token_info_t(p_position_t(2, 7), p_position_t(2, 9), 3, TOKEN_int));
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(1, 9), 0, TOKEN___EOF));
+    assert(token_info == p_token_info_t(p_position_t(2, 10), p_position_t(2, 10), 0, TOKEN___EOF));
 
     p_context_init(&context, "");
     assert(p_lex(&context, &token_info) == P_SUCCESS);
-    assert(token_info == p_token_info_t(p_position_t(0, 0), 0, TOKEN___EOF));
+    assert(token_info == p_token_info_t(p_position_t(1, 1), p_position_t(1, 1), 0, TOKEN___EOF));
 }

spec/test_lexer_multiple_modes.c

@@ -0,0 +1,20 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include <stdio.h>
+
+int main()
+{
+    char const * input = "abc.def";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    printf("pass1\n");
+
+    input = "abc .  abc";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    printf("pass2\n");
+
+    return 0;
+}

spec/test_lexer_multiple_modes.d

@@ -0,0 +1,21 @@
+import testparser;
+import std.stdio;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = `abc.def`;
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    writeln("pass1");
+
+    input = `abc .  abc`;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    writeln("pass2");
+}

spec/test_match_backslashes.c

@@ -0,0 +1,13 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+
+int main()
+{
+    char const * input = "\a\b\t\n\v\f\rt";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+
+    return 0;
+}

spec/test_match_backslashes.d

@@ -0,0 +1,15 @@
+import testparser;
+import std.stdio;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "\a\b\t\n\v\f\rt";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+}

spec/test_named_optional_rule_component_ast.c

@@ -0,0 +1,45 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "b";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+    assert(start->a == NULL);
+    assert(start->pToken2 != NULL);
+    assert_eq(TOKEN_b, start->pToken2->token);
+    assert(start->pR3 == NULL);
+    assert(start->pR == NULL);
+    assert(start->r == NULL);
+
+    input = "abcd";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+    assert(start->a != NULL);
+    assert_eq(TOKEN_a, start->pToken1->token);
+    assert(start->pToken2 != NULL);
+    assert(start->pR3 != NULL);
+    assert(start->pR != NULL);
+    assert(start->r != NULL);
+    assert(start->pR == start->pR3);
+    assert(start->pR == start->r);
+    assert_eq(TOKEN_c, start->pR->pToken1->token);
+
+    input = "bdc";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+    assert(start->a == NULL);
+    assert(start->pToken2 != NULL);
+    assert(start->r != NULL);
+    assert_eq(TOKEN_d, start->pR->pToken1->token);
+
+    return 0;
+}
+

spec/test_named_optional_rule_component_ast.d

@@ -0,0 +1,46 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "b";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+    assert(start.pToken1 is null);
+    assert(start.pToken2 !is null);
+    assert_eq(TOKEN_b, start.pToken2.token);
+    assert(start.pR3 is null);
+    assert(start.pR is null);
+    assert(start.r is null);
+
+    input = "abcd";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+    assert(start.pToken1 != null);
+    assert_eq(TOKEN_a, start.pToken1.token);
+    assert(start.pToken2 != null);
+    assert(start.pR3 != null);
+    assert(start.pR != null);
+    assert(start.r != null);
+    assert(start.pR == start.pR3);
+    assert(start.pR == start.r);
+    assert_eq(TOKEN_c, start.pR.pToken1.token);
+
+    input = "bdc";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+    assert(start.pToken1 is null);
+    assert(start.pToken2 !is null);
+    assert(start.pR !is null);
+    assert_eq(TOKEN_d, start.pR.pToken1.token);
+}

spec/test_optional_rule_component.c

@@ -0,0 +1,22 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+
+int main()
+{
+    char const * input = "b";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+
+    input = "abcd";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+
+    input = "abdc";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+
+    return 0;
+}
+

spec/test_optional_rule_component.d

@@ -0,0 +1,23 @@
+import testparser;
+import std.stdio;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "b";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+
+    input = "abcd";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+
+    input = "abdc";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+}

spec/test_optional_rule_component_ast.c

@@ -0,0 +1,42 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "b";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+    assert(start->pToken1 == NULL);
+    assert(start->pToken2 != NULL);
+    assert_eq(TOKEN_b, start->pToken2->token);
+    assert(start->pR3 == NULL);
+    assert(start->pR == NULL);
+
+    input = "abcd";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+    assert(start->pToken1 != NULL);
+    assert_eq(TOKEN_a, start->pToken1->token);
+    assert(start->pToken2 != NULL);
+    assert(start->pR3 != NULL);
+    assert(start->pR != NULL);
+    assert(start->pR == start->pR3);
+    assert_eq(TOKEN_c, start->pR->pToken1->token);
+
+    input = "bdc";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+    assert(start->pToken1 == NULL);
+    assert(start->pToken2 != NULL);
+    assert(start->pR != NULL);
+    assert_eq(TOKEN_d, start->pR->pToken1->token);
+
+    return 0;
+}
+

spec/test_optional_rule_component_ast.d

@@ -0,0 +1,43 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "b";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    Start * start = p_result(&context);
+    assert(start.pToken1 is null);
+    assert(start.pToken2 !is null);
+    assert_eq(TOKEN_b, start.pToken2.token);
+    assert(start.pR3 is null);
+    assert(start.pR is null);
+
+    input = "abcd";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+    assert(start.pToken1 != null);
+    assert_eq(TOKEN_a, start.pToken1.token);
+    assert(start.pToken2 != null);
+    assert(start.pR3 != null);
+    assert(start.pR != null);
+    assert(start.pR == start.pR3);
+    assert_eq(TOKEN_c, start.pR.pToken1.token);
+
+    input = "bdc";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+    start = p_result(&context);
+    assert(start.pToken1 is null);
+    assert(start.pToken2 !is null);
+    assert(start.pR !is null);
+    assert_eq(TOKEN_d, start.pR.pToken1.token);
+}

spec/test_parser_rule_from_multiple_states.c

@@ -8,8 +8,8 @@ int main()
     p_context_t context;
     p_context_init(&context, (uint8_t const *)input, strlen(input));
     assert(p_parse(&context) == P_UNEXPECTED_TOKEN);
-    assert(p_position(&context).row == 0);
-    assert(p_position(&context).col == 1);
+    assert(p_position(&context).row == 1);
+    assert(p_position(&context).col == 2);
     assert(context.token == TOKEN___EOF);
 
     input = "a b";

spec/test_parser_rule_from_multiple_states.d

@@ -12,7 +12,7 @@ unittest
     p_context_t context;
     p_context_init(&context, input);
     assert(p_parse(&context) == P_UNEXPECTED_TOKEN);
-    assert(p_position(&context) == p_position_t(0, 1));
+    assert(p_position(&context) == p_position_t(1, 2));
     assert(context.token == TOKEN___EOF);
 
     input = "a b";

spec/test_start_rule.c

@@ -0,0 +1,9 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    return 0;
+}

spec/test_start_rule.d

@@ -0,0 +1,8 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}

spec/test_start_rule_ast.c

@@ -0,0 +1,17 @@
+#include "testparser.h"
+#include <assert.h>
+#include <string.h>
+#include "testutils.h"
+
+int main()
+{
+    char const * input = "hi";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert_eq(P_SUCCESS, p_parse(&context));
+    Top * top = p_result(&context);
+    assert(top->pToken != NULL);
+    assert_eq(TOKEN_hi, top->pToken->token);
+
+    return 0;
+}

spec/test_start_rule_ast.d

@@ -0,0 +1,19 @@
+import testparser;
+import std.stdio;
+import testutils;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "hi";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert_eq(P_SUCCESS, p_parse(&context));
+    Top * top = p_result(&context);
+    assert(top.pToken !is null);
+    assert_eq(TOKEN_hi, top.pToken.token);
+}

spec/test_user_terminate_lexer.c

@@ -0,0 +1,19 @@
+#include "testparser.h"
+#include <assert.h>
+#include <stdio.h>
+#include <string.h>
+
+int main()
+{
+    char const * input = "a";
+    p_context_t context;
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_SUCCESS);
+
+    input = "b";
+    p_context_init(&context, (uint8_t const *)input, strlen(input));
+    assert(p_parse(&context) == P_USER_TERMINATED);
+    assert(p_user_terminate_code(&context) == 8675309);
+
+    return 0;
+}

spec/test_user_terminate_lexer.d

@@ -0,0 +1,20 @@
+import testparser;
+import std.stdio;
+
+int main()
+{
+    return 0;
+}
+
+unittest
+{
+    string input = "a";
+    p_context_t context;
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_SUCCESS);
+
+    input = "b";
+    p_context_init(&context, input);
+    assert(p_parse(&context) == P_USER_TERMINATED);
+    assert(p_user_terminate_code(&context) == 8675309);
+}

spec/testutils.c

@@ -17,7 +17,7 @@ void assert_eq_size_t_i(size_t expected, size_t actual, char const * file, size_
 void str_init(str_t * str, char const * cs)
 {
     size_t length = strlen(cs);
-    str->cs = malloc(length + 1u);
+    str->cs = (char *)malloc(length + 1u);
     strcpy(str->cs, cs);
 }
 
@@ -25,7 +25,7 @@ void str_append(str_t * str, char const * cs)
 {
     size_t length = strlen(str->cs);
     size_t length2 = strlen(cs);
-    char * new_cs = malloc(length + length2 + 1u);
+    char * new_cs = (char *)malloc(length + length2 + 1u);
     memcpy(new_cs, str->cs, length);
     strcpy(&new_cs[length], cs);
     free(str->cs);