Resolve #137

README.md

@@ -408,7 +408,12 @@ AST generation
 *cpp-peglib* is able to generate an AST (Abstract Syntax Tree) when parsing. `enable_ast` method on `peg::parser` class enables the feature.
 
 ```
-peg::parser parser("...");
+peg::parser parser(R"(
+  ...
+  defenition1 <- ... { no_ast_opt }
+  defenition2 <- ... { no_ast_opt }
+  ...
+)");
 
 parser.enable_ast();
 
@@ -416,14 +421,14 @@ shared_ptr<peg::Ast> ast;
 if (parser.parse("...", ast)) {
   cout << peg::ast_to_s(ast);
 
-  std::vector<std::string> exceptions = { "defenition1", "defenition2 };
-  ast = peg::AstOptimizer(true, exceptions).optimize(ast);
-
+  ast = parser.optimize_ast(ast);
   cout << peg::ast_to_s(ast);
 }
 ```
 
-`peg::AstOptimizer` removes redundant nodes to make a AST simpler. You can make your own AST optimizers to fit your needs.
+`optimize_ast` removes redundant nodes to make a AST simpler. If you want to disable this behavior from particular rules, `no_ast_opt` instruction can be used.
+
+It internally calls `peg::AstOptimizer` to do the job. You can make your own AST optimizers to fit your needs.
 
 See actual usages in the [AST calculator example](https://github.com/yhirose/cpp-peglib/blob/master/example/calc3.cc) and [PL/0 language example](https://github.com/yhirose/cpp-peglib/blob/master/pl0/pl0.cc).
 

docs/index.html

@@ -23,8 +23,7 @@
     <div class="editor-sub-header">AST</div>
     <pre id="code-ast" class="editor-area"></pre>
     <div class="editor-sub-header">Optimized AST&nbsp;&nbsp;&nbsp;&nbsp;
-      mode:&nbsp;<select id="opt_mode" type="checkbox"><option value="all">All</option><option value="only">Only</option></select>&nbsp;&nbsp;
-      rules:&nbsp;<input id="opt_rules" type="text" size="40" placeholder="Enter definition rules separated by comma."></input>
+      mode:&nbsp;<select id="opt_mode" type="checkbox"><option value="all">All</option><option value="only">Only</option></select>
     </div>
     <pre id="code-ast-optimized" class="editor-area"></pre>
     <div id="code-info" class="editor-info"></div>

docs/index.js

@@ -28,7 +28,6 @@ codeAstOptimized.setOptions({
 codeAstOptimized.renderer.$cursorLayer.element.style.opacity=0;
 
 $('#opt_mode').val(localStorage.getItem('optimazationMode') || 'all');
-$('#opt_rules').val(localStorage.getItem('optimazationRules'));
 
 function escapeHtml(unsafe) {
   return unsafe
@@ -61,12 +60,10 @@ function parse() {
   const codeText = code.getValue();
 
   const optimazationMode = $('#opt_mode').val();
-  const optimazationRules = $('#opt_rules').val();
 
   localStorage.setItem('grammarText', grammarText);
   localStorage.setItem('codeText', codeText);
   localStorage.setItem('optimazationMode', optimazationMode);
-  localStorage.setItem('optimazationRules', optimazationRules);
 
   $grammarInfo.html('');
   $grammarValidation.hide();
@@ -80,7 +77,7 @@ function parse() {
   }
 
   const mode = optimazationMode == 'all';
-  const data = JSON.parse(Module.lint(grammarText, codeText, mode, optimazationRules));
+  const data = JSON.parse(Module.lint(grammarText, codeText, mode));
 
   const isValid = data.grammar.length === 0;
   if (isValid) {
@@ -126,7 +123,6 @@ $('#code-info').on('click', 'li', makeOnClickInInfo(code));
 
 // Event handing in the AST optimazation
 $('#opt_mode').on('change', setupTimer);
-$('#opt_rules').on('keyup', setupTimer);
 
 // Show page
 $('#main').css({

docs/native.cpp

@@ -53,20 +53,7 @@ void parse_code(const std::string &text, peg::parser &peg, std::string &json,
   json += "]";
 }
 
-inline std::vector<std::string> splitRulesText(const std::string &s) {
-  std::vector<std::string> elems;
-  std::stringstream ss(s);
-  std::string elem;
-  while (getline(ss, elem, ',')) {
-    if (!elem.empty()) {
-      elems.push_back(elem);
-    }
-  }
-  return elems;
-}
-
-std::string lint(const std::string &grammarText, const std::string &codeText,
-                 bool opt_mode, const std::string &opt_rules_text) {
+std::string lint(const std::string &grammarText, const std::string &codeText, bool opt_mode) {
   std::string grammarResult;
   std::string codeResult;
   std::string astResult;
@@ -80,9 +67,8 @@ std::string lint(const std::string &grammarText, const std::string &codeText,
     parse_code(codeText, peg, codeResult, ast);
     if (ast) {
       astResult = escape_json(peg::ast_to_s(ast));
-      auto rules = splitRulesText(opt_rules_text);
       astResultOptimized = escape_json(
-          peg::ast_to_s(peg::AstOptimizer(opt_mode, rules).optimize(ast)));
+          peg::ast_to_s(peg.optimize_ast(ast, opt_mode)));
     }
   }
 

example/calc3.cc

@@ -54,7 +54,7 @@ int main(int argc, const char **argv) {
   auto expr = argv[1];
   std::shared_ptr<Ast> ast;
   if (parser.parse(expr, ast)) {
-    ast = AstOptimizer(true).optimize(ast);
+    ast = parser.optimize_ast(ast);
     std::cout << ast_to_s(ast);
     std::cout << expr << " = " << eval(*ast) << std::endl;
     return 0;

example/calc5.cc

@@ -54,7 +54,7 @@ int main(int argc, const char **argv) {
   auto expr = argv[1];
   std::shared_ptr<Ast> ast;
   if (parser.parse(expr, ast)) {
-    ast = AstOptimizer(true).optimize(ast);
+    ast = parser.optimize_ast(ast);
     std::cout << ast_to_s(ast);
     std::cout << expr << " = " << eval(*ast) << std::endl;
     return 0;

lint/peglint.cc

@@ -35,7 +35,6 @@ inline vector<string> split(const string &s, char delim) {
 int main(int argc, const char **argv) {
   auto opt_ast = false;
   auto opt_optimize = false;
-  auto opt_mode = true;
   vector<string> opt_rules;
   auto opt_help = false;
   auto opt_source = false;
@@ -50,17 +49,8 @@ int main(int argc, const char **argv) {
       opt_help = true;
     } else if (string("--ast") == arg) {
       opt_ast = true;
-    } else if (string("--opt") == arg || string("--opt-all") == arg) {
+    } else if (string("--opt") == arg) {
       opt_optimize = true;
-      opt_mode = true;
-    } else if (string("--opt-only") == arg) {
-      opt_optimize = true;
-      opt_mode = false;
-    } else if (string("--opt-rules") == arg) {
-      if (argi < argc) {
-        std::string s = argv[argi++];
-        opt_rules = split(s, ',');
-      }
     } else if (string("--source") == arg) {
       opt_source = true;
       if (argi < argc) {
@@ -190,7 +180,7 @@ int main(int argc, const char **argv) {
 
     if (ast) {
       if (opt_optimize) {
-        ast = peg::AstOptimizer(opt_mode, opt_rules).optimize(ast);
+        ast = parser.optimize_ast(ast);
       }
       std::cout << peg::ast_to_s(ast);
     }

peglib.h

@@ -2287,6 +2287,7 @@ public:
   bool disable_action = false;
 
   std::string error_message;
+  bool no_ast_opt = false;
 
 private:
   friend class Reference;
@@ -3058,9 +3059,9 @@ private:
     ~g["COMMA"] <= seq(chr(','), g["Spacing"]);
 
     // Instruction grammars
-    g["Instruction"] <=
-        seq(g["BeginBlacket"],
-            cho(cho(g["PrecedenceClimbing"]), cho(g["ErrorMessage"])),
+    g["Instruction"] <= seq(g["BeginBlacket"],
+                            cho(cho(g["PrecedenceClimbing"]),
+                                cho(g["ErrorMessage"]), cho(g["NoAstOpt"])),
                             g["EndBlacket"]);
 
     ~g["SpacesZom"] <= zom(g["Space"]);
@@ -3084,6 +3085,9 @@ private:
     g["ErrorMessage"] <=
         seq(lit("message"), g["SpacesOom"], g["LiteralD"], g["SpacesZom"]);
 
+    // No Ast node optimazation instruction
+    g["NoAstOpt"] <= seq(lit("no_ast_opt"), g["SpacesZom"]);
+
     // Set definition names
     for (auto &x : g) {
       x.second.name = x.first;
@@ -3403,6 +3407,12 @@ private:
       instruction.data = std::any_cast<std::string>(vs[0]);
       return instruction;
     };
+
+    g["NoAstOpt"] = [](const SemanticValues & /*vs*/) {
+      Instruction instruction;
+      instruction.type = "no_ast_opt";
+      return instruction;
+    };
   }
 
   bool apply_precedence_instruction(Definition &rule,
@@ -3618,6 +3628,8 @@ private:
         }
       } else if (instruction.type == "message") {
         rule.error_message = std::any_cast<std::string>(instruction.data);
+      } else if (instruction.type == "no_ast_opt") {
+        rule.no_ast_opt = true;
       }
     }
 
@@ -4035,11 +4047,10 @@ public:
 
   const Definition &operator[](const char *s) const { return (*grammar_)[s]; }
 
-  std::vector<std::string> get_rule_names() {
+  std::vector<std::string> get_rule_names() const {
     std::vector<std::string> rules;
-    rules.reserve(grammar_->size());
-    for (auto const &r : *grammar_) {
-      rules.emplace_back(r.first);
+    for (auto &[name, _] : *grammar_) {
+      rules.push_back(name);
     }
     return rules;
   }
@@ -4051,14 +4062,6 @@ public:
     }
   }
 
-  template <typename T = Ast> parser &enable_ast() {
-    for (auto &x : *grammar_) {
-      auto &rule = x.second;
-      if (!rule.action) { add_ast_action<T>(rule); }
-    }
-    return *this;
-  }
-
   void enable_trace(TracerEnter tracer_enter, TracerLeave tracer_leave) {
     if (grammar_ != nullptr) {
       auto &rule = (*grammar_)[start_];
@@ -4067,6 +4070,17 @@ public:
     }
   }
 
+  template <typename T = Ast> parser &enable_ast() {
+    for (auto &[_, rule] : *grammar_) {
+      if (!rule.action) { add_ast_action<T>(rule); }
+    }
+    return *this;
+  }
+
+  template <typename T> std::shared_ptr<T> optimize_ast(std::shared_ptr<T> ast, bool opt_mode = true) const {
+    return AstOptimizer(opt_mode, get_no_ast_opt_rules()).optimize(ast);
+  }
+
   Log log;
 
 private:
@@ -4077,6 +4091,14 @@ private:
     return ret && !r.recovered;
   }
 
+  std::vector<std::string> get_no_ast_opt_rules() const {
+    std::vector<std::string> rules;
+    for (auto &[name, rule] : *grammar_) {
+      if (rule.no_ast_opt) { rules.push_back(name); }
+    }
+    return rules;
+  }
+
   std::shared_ptr<Grammar> grammar_;
   std::string start_;
 };

test/test1.cc

@@ -657,7 +657,7 @@ TEST_CASE("Calculator test with AST", "[general]") {
 
   std::shared_ptr<Ast> ast;
   auto ret = parser.parse("1+2*3*(4-5+6)/7-8", ast);
-  ast = AstOptimizer(true).optimize(ast);
+  ast = parser.optimize_ast(ast);
   auto val = eval(*ast);
 
   REQUIRE(ret == true);

test/test2.cc

@@ -1124,7 +1124,7 @@ CATEGORY   <- < [-_a-zA-Z0-9\u0080-\uFFFF ]+ > _
 ATTRIBUTES <- ATTRIBUTE (',' _ ATTRIBUTE)*
 ATTRIBUTE  <- < [-_a-zA-Z0-9\u0080-\uFFFF]+ > _
 
-ENTRIES    <- (ENTRY (__ ENTRY)*)?
+ENTRIES    <- (ENTRY (__ ENTRY)*)? { no_ast_opt }
 
 ENTRY      <- ONE_WAY PHRASE ('|' _ PHRASE)* !'='
             / PHRASE ('|' _ PHRASE)+ !'='
@@ -1187,7 +1187,7 @@ rrr | sss
 
   )", ast));
 
-  ast = peg::AstOptimizer(true, {"ENTRIES"}).optimize(ast);
+  ast = pg.optimize_ast(ast);
 
   REQUIRE(ast_to_s(ast) ==
 R"(+ START
@@ -1275,7 +1275,7 @@ TEST_CASE("Error recovery 2", "[error]") {
   REQUIRE_FALSE(pg.parse(R"([000]],[111],[222z,"aaa,"bbb",ccc"],[ddd",444,555,"eee],[
   )", ast));
 
-  ast = peg::AstOptimizer(true, {"ENTRIES"}).optimize(ast);
+  ast = pg.optimize_ast(ast);
 
   REQUIRE(ast_to_s(ast) ==
 R"(+ START
@@ -1386,7 +1386,7 @@ SkipToRCUR ← (!RCUR (LCUR SkipToRCUR / .))* RCUR
 }
   )", ast));
 
-  ast = peg::AstOptimizer(true, {"ENTRIES"}).optimize(ast);
+  ast = pg.optimize_ast(ast);
 
   REQUIRE(ast_to_s(ast) ==
 R"(+ Prog