C++1y literal suffix support:

* Allow ns, us, ms, s, min, h as numeric ud-suffixes
 * Allow s as string ud-suffix

llvm-svn: 186933

clang/include/clang/Basic/DiagnosticSemaKinds.td

@@ -5804,8 +5804,8 @@ def err_literal_operator_params : Error<
 def err_literal_operator_extern_c : Error<
   "literal operator must have C++ linkage">;
 def warn_user_literal_reserved : Warning<
-  "user-defined literal suffixes not starting with '_' are reserved; "
-  "no literal will invoke this operator">,
+  "user-defined literal suffixes not starting with '_' are reserved"
+  "%select{; no literal will invoke this operator|}0">,
   InGroup<UserDefinedLiterals>;
 
 // C++ conversion functions

clang/include/clang/Lex/Lexer.h

@@ -569,8 +569,9 @@ private:
 
   void SkipBytes(unsigned Bytes, bool StartOfLine);
 
-  const char *LexUDSuffix(Token &Result, const char *CurPtr);
-  
+  const char *LexUDSuffix(Token &Result, const char *CurPtr,
+                          bool IsStringLiteral);
+
   // Helper functions to lex the remainder of a token of the specific type.
   void LexIdentifier         (Token &Result, const char *CurPtr);
   void LexNumericConstant    (Token &Result, const char *CurPtr);

clang/include/clang/Lex/LiteralSupport.h

@@ -79,6 +79,8 @@ public:
     return SuffixBegin - ThisTokBegin;
   }
 
+  static bool isValidUDSuffix(const LangOptions &LangOpts, StringRef Suffix);
+
   unsigned getRadix() const { return radix; }
 
   /// GetIntegerValue - Convert this numeric literal value to an APInt that

clang/lib/Lex/Lexer.cpp

@@ -1586,7 +1586,8 @@ void Lexer::LexNumericConstant(Token &Result, const char *CurPtr) {
 
 /// LexUDSuffix - Lex the ud-suffix production for user-defined literal suffixes
 /// in C++11, or warn on a ud-suffix in C++98.
-const char *Lexer::LexUDSuffix(Token &Result, const char *CurPtr) {
+const char *Lexer::LexUDSuffix(Token &Result, const char *CurPtr,
+                               bool IsStringLiteral) {
   assert(getLangOpts().CPlusPlus);
 
   // Maximally munch an identifier. FIXME: UCNs.
@@ -1606,9 +1607,20 @@ const char *Lexer::LexUDSuffix(Token &Result, const char *CurPtr) {
     // that does not start with an underscore is ill-formed. As a conforming
     // extension, we treat all such suffixes as if they had whitespace before
     // them.
-    if (C != '_') {
+    bool IsUDSuffix = false;
+    if (C == '_')
+      IsUDSuffix = true;
+    else if (IsStringLiteral && C == 's' && getLangOpts().CPlusPlus1y) {
+      // In C++1y, "s" is a valid ud-suffix for a string literal.
+      unsigned NextSize;
+      if (!isIdentifierBody(getCharAndSizeNoWarn(CurPtr + Size, NextSize,
+                                                 getLangOpts())))
+        IsUDSuffix = true;
+    }
+
+    if (!IsUDSuffix) {
       if (!isLexingRawMode())
-        Diag(CurPtr, getLangOpts().MicrosoftMode ? 
+        Diag(CurPtr, getLangOpts().MicrosoftMode ?
             diag::ext_ms_reserved_user_defined_literal :
             diag::ext_reserved_user_defined_literal)
           << FixItHint::CreateInsertion(getSourceLocation(CurPtr), " ");
@@ -1667,7 +1679,7 @@ void Lexer::LexStringLiteral(Token &Result, const char *CurPtr,
 
   // If we are in C++11, lex the optional ud-suffix.
   if (getLangOpts().CPlusPlus)
-    CurPtr = LexUDSuffix(Result, CurPtr);
+    CurPtr = LexUDSuffix(Result, CurPtr, true);
 
   // If a nul character existed in the string, warn about it.
   if (NulCharacter && !isLexingRawMode())
@@ -1750,7 +1762,7 @@ void Lexer::LexRawStringLiteral(Token &Result, const char *CurPtr,
 
   // If we are in C++11, lex the optional ud-suffix.
   if (getLangOpts().CPlusPlus)
-    CurPtr = LexUDSuffix(Result, CurPtr);
+    CurPtr = LexUDSuffix(Result, CurPtr, true);
 
   // Update the location of token as well as BufferPtr.
   const char *TokStart = BufferPtr;
@@ -1840,7 +1852,7 @@ void Lexer::LexCharConstant(Token &Result, const char *CurPtr,
 
   // If we are in C++11, lex the optional ud-suffix.
   if (getLangOpts().CPlusPlus)
-    CurPtr = LexUDSuffix(Result, CurPtr);
+    CurPtr = LexUDSuffix(Result, CurPtr, false);
 
   // If a nul character existed in the character, warn about it.
   if (NulCharacter && !isLexingRawMode())

clang/lib/Lex/LiteralSupport.cpp

@@ -413,10 +413,12 @@ static void EncodeUCNEscape(const char *ThisTokBegin, const char *&ThisTokBuf,
 ///         decimal-constant integer-suffix
 ///         octal-constant integer-suffix
 ///         hexadecimal-constant integer-suffix
+///         binary-literal integer-suffix [GNU, C++1y]
 ///       user-defined-integer-literal: [C++11 lex.ext]
 ///         decimal-literal ud-suffix
 ///         octal-literal ud-suffix
 ///         hexadecimal-literal ud-suffix
+///         binary-literal ud-suffix [GNU, C++1y]
 ///       decimal-constant:
 ///         nonzero-digit
 ///         decimal-constant digit
@@ -428,6 +430,10 @@ static void EncodeUCNEscape(const char *ThisTokBegin, const char *&ThisTokBuf,
 ///         hexadecimal-constant hexadecimal-digit
 ///       hexadecimal-prefix: one of
 ///         0x 0X
+///       binary-literal:
+///         0b binary-digit
+///         0B binary-digit
+///         binary-literal binary-digit
 ///       integer-suffix:
 ///         unsigned-suffix [long-suffix]
 ///         unsigned-suffix [long-long-suffix]
@@ -441,6 +447,9 @@ static void EncodeUCNEscape(const char *ThisTokBegin, const char *&ThisTokBuf,
 ///         0 1 2 3 4 5 6 7 8 9
 ///         a b c d e f
 ///         A B C D E F
+///       binary-digit:
+///         0
+///         1
 ///       unsigned-suffix: one of
 ///         u U
 ///       long-suffix: one of
@@ -515,6 +524,7 @@ NumericLiteralParser::NumericLiteralParser(StringRef TokSpelling,
   // Parse the suffix.  At this point we can classify whether we have an FP or
   // integer constant.
   bool isFPConstant = isFloatingLiteral();
+  const char *ImaginarySuffixLoc = 0;
 
   // Loop over all of the characters of the suffix.  If we see something bad,
   // we break out of the loop.
@@ -598,9 +608,8 @@ NumericLiteralParser::NumericLiteralParser(StringRef TokSpelling,
     case 'j':
     case 'J':
       if (isImaginary) break;   // Cannot be repeated.
-      PP.Diag(PP.AdvanceToTokenCharacter(TokLoc, s - ThisTokBegin),
-              diag::ext_imaginary_constant);
       isImaginary = true;
+      ImaginarySuffixLoc = s;
       continue;  // Success.
     }
     // If we reached here, there was an error or a ud-suffix.
@@ -608,9 +617,17 @@ NumericLiteralParser::NumericLiteralParser(StringRef TokSpelling,
   }
 
   if (s != ThisTokEnd) {
-    if (PP.getLangOpts().CPlusPlus11 && s == SuffixBegin && *s == '_') {
-      // We have a ud-suffix! By C++11 [lex.ext]p10, ud-suffixes not starting
-      // with an '_' are ill-formed.
+    if (isValidUDSuffix(PP.getLangOpts(),
+                        StringRef(SuffixBegin, ThisTokEnd - SuffixBegin))) {
+      // Any suffix pieces we might have parsed are actually part of the
+      // ud-suffix.
+      isLong = false;
+      isUnsigned = false;
+      isLongLong = false;
+      isFloat = false;
+      isImaginary = false;
+      isMicrosoftInteger = false;
+
       saw_ud_suffix = true;
       return;
     }
@@ -623,6 +640,35 @@ NumericLiteralParser::NumericLiteralParser(StringRef TokSpelling,
     hadError = true;
     return;
   }
+
+  if (isImaginary) {
+    PP.Diag(PP.AdvanceToTokenCharacter(TokLoc,
+                                       ImaginarySuffixLoc - ThisTokBegin),
+            diag::ext_imaginary_constant);
+  }
+}
+
+/// Determine whether a suffix is a valid ud-suffix. We avoid treating reserved
+/// suffixes as ud-suffixes, because the diagnostic experience is better if we
+/// treat it as an invalid suffix.
+bool NumericLiteralParser::isValidUDSuffix(const LangOptions &LangOpts,
+                                           StringRef Suffix) {
+  if (!LangOpts.CPlusPlus11 || Suffix.empty())
+    return false;
+
+  // By C++11 [lex.ext]p10, ud-suffixes starting with an '_' are always valid.
+  if (Suffix[0] == '_')
+    return true;
+
+  // In C++11, there are no library suffixes.
+  if (!LangOpts.CPlusPlus1y)
+    return false;
+
+  // In C++1y, "s", "h", "min", "ms", "us", and "ns" are used in the library.
+  return llvm::StringSwitch<bool>(Suffix)
+      .Cases("h", "min", "s", true)
+      .Cases("ms", "us", "ns", true)
+      .Default(false);
 }
 
 /// ParseNumberStartingWithZero - This method is called when the first character

clang/lib/Sema/SemaDeclCXX.cpp

@@ -27,6 +27,7 @@
 #include "clang/AST/TypeOrdering.h"
 #include "clang/Basic/PartialDiagnostic.h"
 #include "clang/Basic/TargetInfo.h"
+#include "clang/Lex/LiteralSupport.h"
 #include "clang/Lex/Preprocessor.h"
 #include "clang/Sema/CXXFieldCollector.h"
 #include "clang/Sema/DeclSpec.h"
@@ -10650,7 +10651,8 @@ FinishedParams:
     // C++11 [usrlit.suffix]p1:
     //   Literal suffix identifiers that do not start with an underscore
     //   are reserved for future standardization.
-    Diag(FnDecl->getLocation(), diag::warn_user_literal_reserved);
+    Diag(FnDecl->getLocation(), diag::warn_user_literal_reserved)
+      << NumericLiteralParser::isValidUDSuffix(getLangOpts(), LiteralName);
   }
 
   return false;

clang/test/SemaCXX/cxx1y-user-defined-literals.cpp

@@ -0,0 +1,32 @@
+// RUN: %clang_cc1 -std=c++1y %s -include %s -verify
+
+#ifndef INCLUDED
+#define INCLUDED
+
+#pragma clang system_header
+namespace std {
+  using size_t = decltype(sizeof(0));
+
+  struct duration {};
+  duration operator"" ns(unsigned long long);
+  duration operator"" us(unsigned long long);
+  duration operator"" ms(unsigned long long);
+  duration operator"" s(unsigned long long);
+  duration operator"" min(unsigned long long);
+  duration operator"" h(unsigned long long);
+
+  struct string {};
+  string operator"" s(const char*, size_t);
+}
+
+#else
+
+using namespace std;
+duration a = 1ns, b = 1us, c = 1ms, d = 1s, e = 1min, f = 1h;
+string s = "foo"s;
+char error = 'x's; // expected-error {{invalid suffix}} expected-error {{expected ';'}}
+
+int _1z = 1z; // expected-error {{invalid suffix}}
+int _1b = 1b; // expected-error {{invalid digit}}
+
+#endif