Merge pull request #14066 from ethereum/update_grammar

Update grammar `functionDefinition` parser rule and rename `NumberUnit` lexer rule

Changelog.md

@@ -10,6 +10,9 @@ Compiler Features:
 
 
 Bugfixes:
+ * Antlr Grammar: Fix discrepancy with the parser, which allowed octal numbers.
+ * Antlr Grammar: Fix of a discrepancy with the parser, which allowed numbers followed by an identifier with no whitespace.
+ * Antlr Grammar: Stricter rules for function definitions. The grammar will no longer accept as valid free functions having specifiers which are exclusive to contract functions.
 
 
 ### 0.8.19 (2023-02-22)

docs/grammar/SolidityLexer.g4

@@ -62,7 +62,7 @@ New: 'new';
 /**
  * Unit denomination for numbers.
  */
-NumberUnit: 'wei' | 'gwei' | 'ether' | 'seconds' | 'minutes' | 'hours' | 'days' | 'weeks' | 'years';
+SubDenomination: 'wei' | 'gwei' | 'ether' | 'seconds' | 'minutes' | 'hours' | 'days' | 'weeks' | 'years';
 Override: 'override';
 Payable: 'payable';
 Pragma: 'pragma' -> pushMode(PragmaMode);
@@ -220,6 +220,14 @@ fragment EvenHexDigits: HexCharacter HexCharacter ('_'? HexCharacter HexCharacte
 //@doc:inline
 fragment HexCharacter: [0-9A-Fa-f];
 
+/**
+ * Scanned but not used by any rule, i.e, disallowed.
+ * solc parser considers number starting with '0', not immediately followed by '.' or 'x' as
+ * octal, even if non octal digits '8' and '9' are present.
+ */
+OctalNumber: '0' DecimalDigits ('.' DecimalDigits)?;
+
+
 /**
  * A decimal number literal consists of decimal digits that may be delimited by underscores and
  * an optional positive or negative exponent.
@@ -230,6 +238,12 @@ DecimalNumber: (DecimalDigits | (DecimalDigits? '.' DecimalDigits)) ([eE] '-'? D
 fragment DecimalDigits: [0-9] ('_'? [0-9])* ;
 
 
+/**
+ * This is needed to avoid successfully parsing a number followed by a string with no whitespace between.
+ */
+DecimalNumberFollowedByIdentifier: DecimalNumber Identifier;
+
+
 /**
  * An identifier in solidity has to start with a letter, a dollar-sign or an underscore and
  * may additionally contain numbers after the first symbol.

docs/grammar/SolidityParser.g4

@@ -16,7 +16,7 @@ sourceUnit: (
 	| contractDefinition
 	| interfaceDefinition
 	| libraryDefinition
-	| functionDefinition
+	| freeFunctionDefinition
 	| constantVariableDeclaration
 	| structDefinition
 	| enumDefinition
@@ -85,7 +85,7 @@ inheritanceSpecifier: name=identifierPath arguments=callArgumentList?;
  */
 contractBodyElement:
 	constructorDefinition
-	| functionDefinition
+	| contractFunctionDefinition
 	| modifierDefinition
 	| fallbackFunctionDefinition
 	| receiveFunctionDefinition
@@ -156,12 +156,12 @@ overrideSpecifier: Override (LParen overrides+=identifierPath (Comma overrides+=
  * Depending on the context in which the function is defined, further restrictions may apply,
  * e.g. functions in interfaces have to be unimplemented, i.e. may not contain a body block.
  */
-functionDefinition
+contractFunctionDefinition
 locals[
 	boolean visibilitySet = false,
 	boolean mutabilitySet = false,
 	boolean virtualSet = false,
-	boolean overrideSpecifierSet = false
+	boolean overrideSpecifierSet = false,
 ]
 :
 	Function (identifier | Fallback | Receive)
@@ -175,6 +175,17 @@ locals[
 	 )*
 	(Returns LParen returnParameters=parameterList RParen)?
 	(Semicolon | body=block);
+
+/**
+ * The definition of a free function.
+ */
+ freeFunctionDefinition:
+ 	Function (identifier | Fallback | Receive)
+ 	LParen (arguments=parameterList)? RParen
+ 	stateMutability?
+ 	(Returns LParen returnParameters=parameterList RParen)?
+ 	(Semicolon | body=block);
+
 /**
  * The definition of a modifier.
  * Note that within the body block of a modifier, the underscore cannot be used as identifier,
@@ -394,6 +405,7 @@ expression:
  	| (
 		identifier
 		| literal
+		| literalWithSubDenomination
 		| elementaryTypeName[false]
 	  ) # PrimaryExpression
 ;
@@ -412,6 +424,9 @@ inlineArrayExpression: LBrack (expression ( Comma expression)* ) RBrack;
 identifier: Identifier | From | Error | Revert | Global;
 
 literal: stringLiteral | numberLiteral | booleanLiteral | hexStringLiteral | unicodeStringLiteral;
+
+literalWithSubDenomination: numberLiteral SubDenomination;
+
 booleanLiteral: True | False;
 /**
  * A full string literal consists of either one or several consecutive quoted strings.
@@ -429,7 +444,8 @@ unicodeStringLiteral: UnicodeStringLiteral+;
 /**
  * Number literals can be decimal or hexadecimal numbers with an optional unit.
  */
-numberLiteral: (DecimalNumber | HexNumber) NumberUnit?;
+numberLiteral: DecimalNumber | HexNumber;
+
 /**
  * A curly-braced block of statements. Opens its own scope.
  */

test/libsolidity/syntaxTests/denominations/invalid_denomination_no_whitespace.sol

@@ -0,0 +1,5 @@
+contract C {
+	uint constant y = 1wei;
+}
+// ----
+// ParserError 8936: (32-33): Identifier-start is not allowed at end of a number.

test/libsolidity/syntaxTests/denominations/two_denominations_same_literal.sol

@@ -0,0 +1,5 @@
+contract C {
+	uint constant y = 8 gwei ether;
+}
+// ----
+// ParserError 2314: (39-44): Expected ';' but got 'ether'

test/libsolidity/syntaxTests/freeFunctions/named_fallback_warning.sol

@@ -0,0 +1,3 @@
+function fallback() {}
+// ----
+// Warning 3445: (9-17): This function is named "fallback" but is not the fallback function of the contract. If you intend this to be a fallback function, use "fallback(...) { ... }" without the "function" keyword to define it.

test/libsolidity/syntaxTests/freeFunctions/named_receive_warning.sol

@@ -0,0 +1,3 @@
+function receive() {}
+// ----
+// Warning 3445: (9-16): This function is named "receive" but is not the receive function of the contract. If you intend this to be a receive function, use "receive(...) { ... }" without the "function" keyword to define it.

test/libsolidity/syntaxTests/literals/invalid_hex_number.sol

@@ -0,0 +1,5 @@
+contract C {
+    uint x = 0x1000abcdefgh;
+}
+// ----
+// ParserError 8936: (26-38): Identifier-start is not allowed at end of a number.

test/libsolidity/syntaxTests/literals/invalid_octal_denomination_no_whitespace.sol

@@ -0,0 +1,5 @@
+contract C {
+    uint y = 01gwei;
+}
+// ----
+// ParserError 8936: (26-27): Octal numbers not allowed.

test/libsolidity/syntaxTests/literals/invalid_octal_digits.sol

@@ -0,0 +1,5 @@
+contract C {
+    uint y = 098;
+}
+// ----
+// ParserError 8936: (26-27): Octal numbers not allowed.

test/libsolidity/syntaxTests/literals/invalid_octal_number.sol

@@ -0,0 +1,5 @@
+contract C {
+    uint x = 0100;
+}
+// ----
+// ParserError 8936: (26-27): Octal numbers not allowed.