[parser] Refactor of Parse*Statement*, part 1

src/parsing/preparser.cc

Index: src/parsing/preparser.cc
diff --git a/src/parsing/preparser.cc b/src/parsing/preparser.cc
index 369ff2415efb8dbe3351b73dfb3c818147460c64..4a3f680f035d1384bb8d445bf5b40ec5b4ee2910 100644
--- a/src/parsing/preparser.cc
+++ b/src/parsing/preparser.cc
@@ -122,223 +122,19 @@ PreParser::PreParseResult PreParser::PreParseLazyFunction(
 // it is used) are generally omitted.
 
 
-PreParser::Statement PreParser::ParseStatementListItem(bool* ok) {
-  // ECMA 262 6th Edition
-  // StatementListItem[Yield, Return] :
-  //   Statement[?Yield, ?Return]
-  //   Declaration[?Yield]
-  //
-  // Declaration[Yield] :
-  //   HoistableDeclaration[?Yield]
-  //   ClassDeclaration[?Yield]
-  //   LexicalDeclaration[In, ?Yield]
-  //
-  // HoistableDeclaration[Yield, Default] :
-  //   FunctionDeclaration[?Yield, ?Default]
-  //   GeneratorDeclaration[?Yield, ?Default]
-  //
-  // LexicalDeclaration[In, Yield] :
-  //   LetOrConst BindingList[?In, ?Yield] ;
-
-  switch (peek()) {
-    case Token::FUNCTION:
-      return ParseHoistableDeclaration(ok);
-    case Token::CLASS:
-      return ParseClassDeclaration(ok);
-    case Token::CONST:
-      return ParseVariableStatement(kStatementListItem, ok);
-    case Token::LET:
-      if (IsNextLetKeyword()) {
-        return ParseVariableStatement(kStatementListItem, ok);
-      }
-      break;
-    case Token::ASYNC:
-      if (allow_harmony_async_await() && PeekAhead() == Token::FUNCTION &&
-          !scanner()->HasAnyLineTerminatorAfterNext()) {
-        Consume(Token::ASYNC);
-        return ParseAsyncFunctionDeclaration(ok);
-      }
-    /* falls through */
-    default:
-      break;
-  }
-  return ParseStatement(kAllowLabelledFunctionStatement, ok);
-}
-
-PreParser::LazyParsingResult PreParser::ParseStatementList(int end_token,
-                                                           bool may_abort,
-                                                           bool* ok) {
-  // SourceElements ::
-  //   (Statement)* <end_token>
-
-  int count_statements = 0;
-
-  bool directive_prologue = true;
-  while (peek() != end_token) {
-    if (directive_prologue && peek() != Token::STRING) {
-      directive_prologue = false;
-    }
-    bool starts_with_identifier = peek() == Token::IDENTIFIER;
-    Scanner::Location token_loc = scanner()->peek_location();
-    Statement statement =
-        ParseStatementListItem(CHECK_OK_VALUE(kLazyParsingComplete));
-
-    if (directive_prologue) {
-      bool use_strict_found = statement.IsUseStrictLiteral();
-
-      if (use_strict_found) {
-        scope()->SetLanguageMode(
-            static_cast<LanguageMode>(scope()->language_mode() | STRICT));
-      } else if (!statement.IsStringLiteral()) {
-        directive_prologue = false;
-      }
-
-      if (use_strict_found && !scope()->HasSimpleParameters()) {
-        // TC39 deemed "use strict" directives to be an error when occurring
-        // in the body of a function with non-simple parameter list, on
-        // 29/7/2015. https://goo.gl/ueA7Ln
-        ReportMessageAt(token_loc,
-                        MessageTemplate::kIllegalLanguageModeDirective,
-                        "use strict");
-        *ok = false;
-        return kLazyParsingComplete;
-      }
-    }
-
-    // If we're allowed to reset to a bookmark, we will do so when we see a long
-    // and trivial function.
-    // Our current definition of 'long and trivial' is:
-    // - over 200 statements
-    // - all starting with an identifier (i.e., no if, for, while, etc.)
-    if (may_abort) {
-      if (!starts_with_identifier) {
-        may_abort = false;
-      } else if (++count_statements > kLazyParseTrialLimit) {
-        return kLazyParsingAborted;
-      }
-    }
-  }
-  return kLazyParsingComplete;
-}
-
-
-PreParser::Statement PreParser::ParseStatement(
-    AllowLabelledFunctionStatement allow_function, bool* ok) {
-  // Statement ::
-  //   EmptyStatement
-  //   ...
-
-  if (peek() == Token::SEMICOLON) {
-    Next();
-    return Statement::Default();
-  }
-  return ParseSubStatement(allow_function, ok);
-}
-
 PreParser::Statement PreParser::ParseScopedStatement(bool legacy, bool* ok) {
   if (is_strict(language_mode()) || peek() != Token::FUNCTION ||
       (legacy && allow_harmony_restrictive_declarations())) {
-    return ParseSubStatement(kDisallowLabelledFunctionStatement, ok);
+    return ParseStatement(nullptr, kDisallowLabelledFunctionStatement, ok);
   } else {
     BlockState block_state(&scope_state_);
     return ParseFunctionDeclaration(ok);
   }
 }
 
-PreParser::Statement PreParser::ParseSubStatement(
-    AllowLabelledFunctionStatement allow_function, bool* ok) {
-  // Statement ::
-  //   Block
-  //   VariableStatement
-  //   EmptyStatement
-  //   ExpressionStatement
-  //   IfStatement
-  //   IterationStatement
-  //   ContinueStatement
-  //   BreakStatement
-  //   ReturnStatement
-  //   WithStatement
-  //   LabelledStatement
-  //   SwitchStatement
-  //   ThrowStatement
-  //   TryStatement
-  //   DebuggerStatement
-
-  // Note: Since labels can only be used by 'break' and 'continue'
-  // statements, which themselves are only valid within blocks,
-  // iterations or 'switch' statements (i.e., BreakableStatements),
-  // labels can be simply ignored in all other cases; except for
-  // trivial labeled break statements 'label: break label' which is
-  // parsed into an empty statement.
-
-  // Keep the source position of the statement
-  switch (peek()) {
-    case Token::LBRACE:
-      return ParseBlock(ok);
-
-    case Token::SEMICOLON:
-      Next();
-      return Statement::Default();
-
-    case Token::IF:
-      return ParseIfStatement(ok);
-
-    case Token::DO:
-      return ParseDoWhileStatement(ok);
-
-    case Token::WHILE:
-      return ParseWhileStatement(ok);
-
-    case Token::FOR:
-      return ParseForStatement(ok);
-
-    case Token::CONTINUE:
-      return ParseContinueStatement(ok);
-
-    case Token::BREAK:
-      return ParseBreakStatement(ok);
-
-    case Token::RETURN:
-      return ParseReturnStatement(ok);
-
-    case Token::WITH:
-      return ParseWithStatement(ok);
-
-    case Token::SWITCH:
-      return ParseSwitchStatement(ok);
-
-    case Token::THROW:
-      return ParseThrowStatement(ok);
-
-    case Token::TRY:
-      return ParseTryStatement(ok);
-
-    case Token::FUNCTION:
-      // FunctionDeclaration only allowed as a StatementListItem, not in
-      // an arbitrary Statement position. Exceptions such as
-      // ES#sec-functiondeclarations-in-ifstatement-statement-clauses
-      // are handled by calling ParseScopedStatement rather than
-      // ParseSubStatement directly.
-      ReportMessageAt(scanner()->peek_location(),
-                      is_strict(language_mode())
-                          ? MessageTemplate::kStrictFunction
-                          : MessageTemplate::kSloppyFunction);
-      *ok = false;
-      return Statement::Default();
-
-    case Token::DEBUGGER:
-      return ParseDebuggerStatement(ok);
-
-    case Token::VAR:
-      return ParseVariableStatement(kStatement, ok);
-
-    default:
-      return ParseExpressionOrLabelledStatement(allow_function, ok);
-  }
-}
-
 PreParser::Statement PreParser::ParseHoistableDeclaration(
-    int pos, ParseFunctionFlags flags, bool* ok) {
+    int pos, ParseFunctionFlags flags, ZoneList<const AstRawString*>* names,
+    bool default_export, bool* ok) {
   const bool is_generator = flags & ParseFunctionFlags::kIsGenerator;
   const bool is_async = flags & ParseFunctionFlags::kIsAsync;
   DCHECK(!is_generator || !is_async);
@@ -358,7 +154,8 @@ PreParser::Statement PreParser::ParseHoistableDeclaration(
   return Statement::FunctionDeclaration();
 }
 
-PreParser::Statement PreParser::ParseAsyncFunctionDeclaration(bool* ok) {
+PreParser::Statement PreParser::ParseAsyncFunctionDeclaration(
+    ZoneList<const AstRawString*>* names, bool default_export, bool* ok) {
   // AsyncFunctionDeclaration ::
   //   async [no LineTerminator here] function BindingIdentifier[Await]
   //       ( FormalParameters[Await] ) { AsyncFunctionBody }
@@ -366,10 +163,11 @@ PreParser::Statement PreParser::ParseAsyncFunctionDeclaration(bool* ok) {
   int pos = position();
   Expect(Token::FUNCTION, CHECK_OK);
   ParseFunctionFlags flags = ParseFunctionFlags::kIsAsync;
-  return ParseHoistableDeclaration(pos, flags, ok);
+  return ParseHoistableDeclaration(pos, flags, names, default_export, ok);
 }
 
-PreParser::Statement PreParser::ParseHoistableDeclaration(bool* ok) {
+PreParser::Statement PreParser::ParseHoistableDeclaration(
+    ZoneList<const AstRawString*>* names, bool default_export, bool* ok) {
   // FunctionDeclaration ::
   //   'function' Identifier '(' FormalParameterListopt ')' '{' FunctionBody '}'
   // GeneratorDeclaration ::
@@ -382,13 +180,11 @@ PreParser::Statement PreParser::ParseHoistableDeclaration(bool* ok) {
   if (Check(Token::MUL)) {
     flags |= ParseFunctionFlags::kIsGenerator;
   }
-  return ParseHoistableDeclaration(pos, flags, ok);
+  return ParseHoistableDeclaration(pos, flags, names, default_export, ok);
 }
 
-
-PreParser::Statement PreParser::ParseClassDeclaration(bool* ok) {
-  Expect(Token::CLASS, CHECK_OK);
-
+PreParser::Statement PreParser::ParseClassDeclaration(
+    ZoneList<const AstRawString*>* names, bool default_export, bool* ok) {
   int pos = position();
   bool is_strict_reserved = false;
   Identifier name =
@@ -399,8 +195,8 @@ PreParser::Statement PreParser::ParseClassDeclaration(bool* ok) {
   return Statement::Default();
 }
 
-
-PreParser::Statement PreParser::ParseBlock(bool* ok) {
+PreParser::Statement PreParser::ParseBlock(
+    ZoneList<const AstRawString*>* labels, bool* ok) {
   // Block ::
   //   '{' StatementList '}'
 
@@ -416,15 +212,14 @@ PreParser::Statement PreParser::ParseBlock(bool* ok) {
   return final;
 }
 
-
 PreParser::Statement PreParser::ParseVariableStatement(
     VariableDeclarationContext var_context,
-    bool* ok) {
+    ZoneList<const AstRawString*>* names, bool* ok) {
   // VariableStatement ::
   //   VariableDeclarations ';'
 
   Statement result =
-      ParseVariableDeclarations(var_context, nullptr, nullptr, CHECK_OK);
+      ParseVariableDeclarations(var_context, nullptr, names, CHECK_OK);
   ExpectSemicolon(CHECK_OK);
   return result;
 }
@@ -442,10 +237,11 @@ PreParser::Statement PreParser::ParseFunctionDeclaration(bool* ok) {
       return Statement::Default();
     }
   }
-  return ParseHoistableDeclaration(pos, flags, ok);
+  return ParseHoistableDeclaration(pos, flags, nullptr, false, ok);
 }
 
 PreParser::Statement PreParser::ParseExpressionOrLabelledStatement(
+    ZoneList<const AstRawString*>* names,
     AllowLabelledFunctionStatement allow_function, bool* ok) {
   // ExpressionStatement | LabelledStatement ::
   //   Expression ';'
@@ -489,7 +285,7 @@ PreParser::Statement PreParser::ParseExpressionOrLabelledStatement(
       }
     }
     Statement statement =
-        ParseStatement(kDisallowLabelledFunctionStatement, ok);
+        ParseStatement(nullptr, kDisallowLabelledFunctionStatement, ok);
     return statement.IsJumpStatement() ? Statement::Default() : statement;
     // Preparsing is disabled for extensions (because the extension details
     // aren't passed to lazily compiled functions), so we don't
@@ -500,8 +296,8 @@ PreParser::Statement PreParser::ParseExpressionOrLabelledStatement(
   return Statement::ExpressionStatement(expr);
 }
 
-
-PreParser::Statement PreParser::ParseIfStatement(bool* ok) {
+PreParser::Statement PreParser::ParseIfStatement(
+    ZoneList<const AstRawString*>* labels, bool* ok) {
   // IfStatement ::
   //   'if' '(' Expression ')' Statement ('else' Statement)?
 
@@ -539,8 +335,8 @@ PreParser::Statement PreParser::ParseContinueStatement(bool* ok) {
   return Statement::Jump();
 }
 
-
-PreParser::Statement PreParser::ParseBreakStatement(bool* ok) {
+PreParser::Statement PreParser::ParseBreakStatement(
+    ZoneList<const AstRawString*>* labels, bool* ok) {
   // BreakStatement ::
   //   'break' [no line terminator] Identifier? ';'
 
@@ -593,8 +389,8 @@ PreParser::Statement PreParser::ParseReturnStatement(bool* ok) {
   return Statement::Jump();
 }
 
-
-PreParser::Statement PreParser::ParseWithStatement(bool* ok) {
+PreParser::Statement PreParser::ParseWithStatement(
+    ZoneList<const AstRawString*>* labels, bool* ok) {
   // WithStatement ::
   //   'with' '(' Expression ')' Statement
   Expect(Token::WITH, CHECK_OK);
@@ -613,8 +409,8 @@ PreParser::Statement PreParser::ParseWithStatement(bool* ok) {
   return Statement::Default();
 }
 
-
-PreParser::Statement PreParser::ParseSwitchStatement(bool* ok) {
+PreParser::Statement PreParser::ParseSwitchStatement(
+    ZoneList<const AstRawString*>* labels, bool* ok) {
   // SwitchStatement ::
   //   'switch' '(' Expression ')' '{' CaseClause* '}'
 
@@ -649,8 +445,8 @@ PreParser::Statement PreParser::ParseSwitchStatement(bool* ok) {
   return Statement::Default();
 }
 
-
-PreParser::Statement PreParser::ParseDoWhileStatement(bool* ok) {
+PreParser::Statement PreParser::ParseDoWhileStatement(
+    ZoneList<const AstRawString*>* labels, bool* ok) {
   // DoStatement ::
   //   'do' Statement 'while' '(' Expression ')' ';'
 
@@ -664,8 +460,8 @@ PreParser::Statement PreParser::ParseDoWhileStatement(bool* ok) {
   return Statement::Default();
 }
 
-
-PreParser::Statement PreParser::ParseWhileStatement(bool* ok) {
+PreParser::Statement PreParser::ParseWhileStatement(
+    ZoneList<const AstRawString*>* labels, bool* ok) {
   // WhileStatement ::
   //   'while' '(' Expression ')' Statement
 
@@ -677,8 +473,8 @@ PreParser::Statement PreParser::ParseWhileStatement(bool* ok) {
   return Statement::Default();
 }
 
-
-PreParser::Statement PreParser::ParseForStatement(bool* ok) {
+PreParser::Statement PreParser::ParseForStatement(
+    ZoneList<const AstRawString*>* labels, bool* ok) {
   // ForStatement ::
   //   'for' '(' Expression? ';' Expression? ';' Expression? ')' Statement
 
@@ -844,7 +640,7 @@ PreParser::Statement PreParser::ParseTryStatement(bool* ok) {
   {
     ReturnExprScope no_tail_calls(function_state_,
                                   ReturnExprContext::kInsideTryBlock);
-    ParseBlock(CHECK_OK);
+    ParseBlock(nullptr, CHECK_OK);
   }
 
   Token::Value tok = peek();
@@ -870,7 +666,7 @@ PreParser::Statement PreParser::ParseTryStatement(bool* ok) {
       BlockState block_state(&scope_state_, catch_scope);
       {
         BlockState block_state(&scope_state_);
-        ParseBlock(CHECK_OK);
+        ParseBlock(nullptr, CHECK_OK);
       }
     }
     catch_block_exists = true;
@@ -878,7 +674,7 @@ PreParser::Statement PreParser::ParseTryStatement(bool* ok) {
   }
   if (tok == Token::FINALLY) {
     Consume(Token::FINALLY);
-    ParseBlock(CHECK_OK);
+    ParseBlock(nullptr, CHECK_OK);
     if (FLAG_harmony_explicit_tailcalls && catch_block_exists &&
         tail_call_expressions_in_catch_block.has_explicit_tail_calls()) {
       // TODO(ishell): update chapter number.
@@ -919,6 +715,7 @@ PreParser::Expression PreParser::ParseFunctionLiteral(
   //   '(' FormalParameterList? ')' '{' FunctionBody '}'
 
   // Parse function body.
+  PreParserStatementList body;
   bool outer_is_script_scope = scope()->is_script_scope();
   DeclarationScope* function_scope = NewFunctionScope(kind);
   function_scope->SetLanguageMode(language_mode);
@@ -947,7 +744,7 @@ PreParser::Expression PreParser::ParseFunctionLiteral(
   if (is_lazily_parsed) {
     ParseLazyFunctionLiteralBody(false, CHECK_OK);
   } else {
-    ParseStatementList(Token::RBRACE, CHECK_OK);
+    ParseStatementList(body, Token::RBRACE, CHECK_OK);

[marja, 2016/09/05 08:14:33]

Maybe something for a follow-up CL: I find it somewhat surprising that body is
passed by value here - in PreParser it's fine since we don't need to collect
statements, but still, it's a bit weird. It would be more intuitive if we passed
a pointer to it, so the parameter type would be StatementListT* in both cases.
That would also signal to the reader than it's the output parameter.

Would that be feasible or does it conflict too much with other preparser-isms
(such as how ExpressionLists are used now)?

At least some of the ExpressionList usages were like this:
ExpressionListT list = impl()->NewExpressionList();
list->Add(..); // NOOP for PreParser

which is just NOOP but not surprising in the context of PreParser, whereas this
pass-by-value construct is surprising.

What do you think?

[nickie, 2016/09/05 11:29:31]

Yes, you're right.  It starts from the fact that in ParserTypes<Parser> almost
everything is a pointer (and not a const one), whereas in ParserTypes<PreParser>
they're objects.

I'm adding a TODO comment near the declaration of ParseStatementList, so we can
discuss this later, but there may be other similar cases where we use these
pointers to modify the AST.

   }
   Expect(Token::RBRACE, CHECK_OK);
 
@@ -1002,8 +799,9 @@ PreParser::Expression PreParser::ParseAsyncFunctionExpression(bool* ok) {
 PreParser::LazyParsingResult PreParser::ParseLazyFunctionLiteralBody(
     bool may_abort, bool* ok) {
   int body_start = position();
+  PreParserStatementList body;
   LazyParsingResult result = ParseStatementList(
-      Token::RBRACE, may_abort, CHECK_OK_VALUE(kLazyParsingComplete));
+      body, Token::RBRACE, may_abort, CHECK_OK_VALUE(kLazyParsingComplete));
   if (result == kLazyParsingAborted) return result;
 
   // Position right after terminal '}'.