1 files changed, 28 insertions, 20 deletions

diff --git a/lib/Lex/Lexer.cpp b/lib/Lex/Lexer.cpp
index 6f6b50b..ca5252e 100644
--- a/lib/Lex/Lexer.cpp
+++ b/lib/Lex/Lexer.cpp
@@ -540,16 +540,16 @@ namespace {
   };
 }
 
-std::pair<unsigned, bool>
-Lexer::ComputePreamble(const llvm::MemoryBuffer *Buffer,
-                       const LangOptions &LangOpts, unsigned MaxLines) {
+std::pair<unsigned, bool> Lexer::ComputePreamble(StringRef Buffer,
+                                                 const LangOptions &LangOpts,
+                                                 unsigned MaxLines) {
   // Create a lexer starting at the beginning of the file. Note that we use a
   // "fake" file source location at offset 1 so that the lexer will track our
   // position within the file.
   const unsigned StartOffset = 1;
   SourceLocation FileLoc = SourceLocation::getFromRawEncoding(StartOffset);
-  Lexer TheLexer(FileLoc, LangOpts, Buffer->getBufferStart(),
-                 Buffer->getBufferStart(), Buffer->getBufferEnd());
+  Lexer TheLexer(FileLoc, LangOpts, Buffer.begin(), Buffer.begin(),
+                 Buffer.end());
   TheLexer.SetCommentRetentionState(true);
 
   // StartLoc will differ from FileLoc if there is a BOM that was skipped.
@@ -563,9 +563,9 @@ Lexer::ComputePreamble(const llvm::MemoryBuffer *Buffer,
 
   unsigned MaxLineOffset = 0;
   if (MaxLines) {
-    const char *CurPtr = Buffer->getBufferStart();
+    const char *CurPtr = Buffer.begin();
     unsigned CurLine = 0;
-    while (CurPtr != Buffer->getBufferEnd()) {
+    while (CurPtr != Buffer.end()) {
       char ch = *CurPtr++;
       if (ch == '\n') {
         ++CurLine;
@@ -573,8 +573,8 @@ Lexer::ComputePreamble(const llvm::MemoryBuffer *Buffer,
           break;
       }
     }
-    if (CurPtr != Buffer->getBufferEnd())
-      MaxLineOffset = CurPtr - Buffer->getBufferStart();
+    if (CurPtr != Buffer.end())
+      MaxLineOffset = CurPtr - Buffer.begin();
   }
 
   do {
@@ -1597,7 +1597,7 @@ bool Lexer::LexNumericConstant(Token &Result, const char *CurPtr) {
   }
 
   // If we have a digit separator, continue.
-  if (C == '\'' && getLangOpts().CPlusPlus1y) {
+  if (C == '\'' && getLangOpts().CPlusPlus14) {
     unsigned NextSize;
     char Next = getCharAndSizeNoWarn(CurPtr + Size, NextSize, getLangOpts());
     if (isIdentifierBody(Next)) {
@@ -1660,7 +1660,7 @@ const char *Lexer::LexUDSuffix(Token &Result, const char *CurPtr,
     bool IsUDSuffix = false;
     if (C == '_')
       IsUDSuffix = true;
-    else if (IsStringLiteral && getLangOpts().CPlusPlus1y) {
+    else if (IsStringLiteral && getLangOpts().CPlusPlus14) {
       // In C++1y, we need to look ahead a few characters to see if this is a
       // valid suffix for a string literal or a numeric literal (this could be
       // the 'operator""if' defining a numeric literal operator).
@@ -1889,17 +1889,20 @@ bool Lexer::LexAngledStringLiteral(Token &Result, const char *CurPtr) {
 
 
 /// LexCharConstant - Lex the remainder of a character constant, after having
-/// lexed either ' or L' or u' or U'.
+/// lexed either ' or L' or u8' or u' or U'.
 bool Lexer::LexCharConstant(Token &Result, const char *CurPtr,
                             tok::TokenKind Kind) {
   // Does this character contain the \0 character?
   const char *NulCharacter = nullptr;
 
-  if (!isLexingRawMode() &&
-      (Kind == tok::utf16_char_constant || Kind == tok::utf32_char_constant))
-    Diag(BufferPtr, getLangOpts().CPlusPlus
-           ? diag::warn_cxx98_compat_unicode_literal
-           : diag::warn_c99_compat_unicode_literal);
+  if (!isLexingRawMode()) {
+    if (Kind == tok::utf16_char_constant || Kind == tok::utf32_char_constant)
+      Diag(BufferPtr, getLangOpts().CPlusPlus
+                          ? diag::warn_cxx98_compat_unicode_literal
+                          : diag::warn_c99_compat_unicode_literal);
+    else if (Kind == tok::utf8_char_constant)
+      Diag(BufferPtr, diag::warn_cxx14_compat_u8_character_literal);
+  }
 
   char C = getAndAdvanceChar(CurPtr, Result);
   if (C == '\'') {
@@ -2319,7 +2322,7 @@ bool Lexer::SkipBlockComment(Token &Result, const char *CurPtr,
         '/', '/', '/', '/',  '/', '/', '/', '/'
       };
       while (CurPtr+16 <= BufferEnd &&
-             !vec_any_eq(*(vector unsigned char*)CurPtr, Slashes))
+             !vec_any_eq(*(const vector unsigned char*)CurPtr, Slashes))
         CurPtr += 16;
 #else
       // Scan for '/' quickly.  Many block comments are very large.
@@ -2585,8 +2588,8 @@ static const char *FindConflictEnd(const char *CurPtr, const char *BufferEnd,
   size_t Pos = RestOfBuffer.find(Terminator);
   while (Pos != StringRef::npos) {
     // Must occur at start of line.
-    if (RestOfBuffer[Pos-1] != '\r' &&
-        RestOfBuffer[Pos-1] != '\n') {
+    if (Pos == 0 ||
+        (RestOfBuffer[Pos - 1] != '\r' && RestOfBuffer[Pos - 1] != '\n')) {
       RestOfBuffer = RestOfBuffer.substr(Pos+TermLen);
       Pos = RestOfBuffer.find(Terminator);
       continue;
@@ -3068,6 +3071,11 @@ LexNextToken:
                                ConsumeChar(ConsumeChar(CurPtr, SizeTmp, Result),
                                            SizeTmp2, Result),
                                tok::utf8_string_literal);
+        if (Char2 == '\'' && LangOpts.CPlusPlus1z)
+          return LexCharConstant(
+              Result, ConsumeChar(ConsumeChar(CurPtr, SizeTmp, Result),
+                                  SizeTmp2, Result),
+              tok::utf8_char_constant);
 
         if (Char2 == 'R' && LangOpts.CPlusPlus11) {
           unsigned SizeTmp3;