v2.2.0

CHANGELOG.md

@@ -1,3 +1,22 @@
+## 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

Gemfile

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

Gemfile.lock

@@ -1,6 +1,7 @@
 GEM
   remote: https://rubygems.org/
   specs:
+    base64 (0.2.0)
     date (3.4.1)
     diff-lcs (1.5.1)
     docile (1.4.1)
@@ -37,6 +38,7 @@ PLATFORMS
   ruby
 
 DEPENDENCIES
+  base64
   rake
   rdoc
   redcarpet

doc/user_guide.md

@@ -536,6 +536,28 @@ 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).
@@ -762,10 +784,16 @@ 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 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.
@@ -999,6 +1027,26 @@ if (p_parse(&context) == P_UNEXPECTED_TOKEN)
 }
 ```
 
+### `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`

lib/propane/generator.rb

@@ -43,8 +43,8 @@ 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"

lib/propane/grammar.rb

@@ -25,7 +25,7 @@ class Propane
       @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_"
@@ -58,7 +58,7 @@ 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!
@@ -81,8 +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
 
@@ -117,7 +117,7 @@ class Propane
         md = consume!(/([\w.]+)\s*/, "expected module name")
         @modulename = md[1]
         consume!(/;/, "expected `;'")
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -153,9 +153,9 @@ class Propane
         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
@@ -173,7 +173,7 @@ class Propane
         consume!(/;/, "expected `;'");
         token = Token.new(name, ptypename, @line_number)
         @tokens << token
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -186,8 +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
@@ -208,7 +208,7 @@ class Propane
           end
         end
         @rules << Rule.new(rule_name, components, code, ptypename, @line_number)
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -225,8 +225,8 @@ class Propane
         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
@@ -247,7 +247,7 @@ class Propane
         else
           @code_blocks[name] = code
         end
-        @mode = nil
+        @modeline = nil
         true
       end
     end
@@ -315,6 +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/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.
@@ -46,14 +46,14 @@ 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.nfa.end_state.accepts = self

lib/propane/rule_set.rb

@@ -119,6 +119,8 @@ class Propane
     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|
@@ -136,6 +138,16 @@ class Propane
             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]
@@ -150,18 +162,6 @@ class Propane
             "#{grammar.ast_prefix}#{node_name}#{grammar.ast_suffix}"
         end
       end
-      # Now merge in the field aliases as given by the user in the
-      # grammar.
-      field_aliases = {}
-      @rules.each do |rule|
-        rule.aliases.each do |alias_name, index|
-          if field_aliases[alias_name] && field_aliases[alias_name] != 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] = index
-          @ast_fields[index][alias_name] = @ast_fields[index].first[1]
-        end
-      end
     end
 
   end

lib/propane/version.rb

@@ -1,3 +1,3 @@
 class Propane
-  VERSION = "2.1.0"
+  VERSION = "2.2.0"
 end

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_spec.rb

@@ -621,6 +621,62 @@ EOF
         ])
       end
 
+      it "multiple lexer modes may apply to a pattern" do
+        case language
+        when "c"
+          write_grammar <<EOF
+<<
+#include <stdio.h>
+>>
+ptype char;
+token abc;
+token def;
+default, identonly: token ident /[a-z]+/ <<
+  $$ = match[0];
+  $mode(default);
+  return $token(ident);
+>>
+token dot /\\./ <<
+  $mode(identonly);
+>>
+default, identonly: drop /\\s+/;
+Start -> abc dot ident <<
+  printf("ident: %c\\n", $3);
+>>
+EOF
+        when "d"
+          write_grammar <<EOF
+<<
+import std.stdio;
+>>
+ptype char;
+token abc;
+token def;
+default, identonly: token ident /[a-z]+/ <<
+  $$ = match[0];
+  $mode(default);
+>>
+token dot /\\./ <<
+  $mode(identonly);
+>>
+default, identonly: drop /\\s+/;
+Start -> abc dot ident <<
+  writeln("ident: ", $3);
+>>
+EOF
+        end
+        run_propane(language: language)
+        compile("spec/test_lexer_multiple_modes.#{language}", language: language)
+        results = run_test
+        expect(results.status).to eq 0
+        verify_lines(results.stdout, [
+          "ident: d",
+          "pass1",
+          "ident: a",
+          "pass2",
+        ])
+      end
+
       it "executes user code associated with a parser rule" do
         case language
         when "c"
@@ -1219,6 +1275,29 @@ EOF
         expect(results.status).to eq 0
       end
 
+      it "aliases the correct field when multiple rules are in a rule set in AST mode" do
+        write_grammar <<EOF
+ast;
+
+token a;
+token b;
+token c;
+drop /\\s+/;
+Start -> a;
+Start -> Foo;
+Start -> T:first T:second T:third;
+Foo -> b;
+T -> a;
+T -> b;
+T -> c;
+EOF
+        run_propane(language: language)
+        compile("spec/test_ast_field_aliases.#{language}", language: language)
+        results = run_test
+        expect(results.stderr).to eq ""
+        expect(results.status).to eq 0
+      end
+
       it "allows specifying field aliases when AST mode is not enabled" do
         if language == "d"
           write_grammar <<EOF
@@ -1253,6 +1332,56 @@ Start -> id:first id:second <<
   printf("first is %s\\n", ${first});
   printf("second is %s\\n", ${second});
 >>
+EOF
+        end
+        run_propane(language: language)
+        compile("spec/test_field_aliases.#{language}", language: language)
+        results = run_test
+        expect(results.stderr).to eq ""
+        expect(results.status).to eq 0
+        expect(results.stdout).to match /first is foo1.*second is bar2/m
+      end
+
+      it "aliases the correct field when multiple rules are in a rule set when AST mode is not enabled" do
+        if language == "d"
+          write_grammar <<EOF
+<<
+import std.stdio;
+>>
+ptype string;
+token id /[a-zA-Z_][a-zA-Z0-9_]*/ <<
+  $$ = match;
+>>
+drop /\\s+/;
+Start -> id;
+Start -> Foo;
+Start -> id:first id:second <<
+  writeln("first is ", ${first});
+  writeln("second is ", ${second});
+>>
+Foo -> ;
+EOF
+        else
+          write_grammar <<EOF
+<<
+#include <stdio.h>
+#include <string.h>
+>>
+ptype char const *;
+token id /[a-zA-Z_][a-zA-Z0-9_]*/ <<
+  char * s = malloc(match_length + 1);
+  strncpy(s, (char const *)match, match_length);
+  s[match_length] = 0;
+  $$ = s;
+>>
+drop /\\s+/;
+Start -> id;
+Start -> Foo;
+Start -> id:first id:second <<
+  printf("first is %s\\n", ${first});
+  printf("second is %s\\n", ${second});
+>>
+Foo -> ;
 EOF
         end
         run_propane(language: language)

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");
+}