Implement basic token pasting (## operator). This implements

test/Preprocessor/macro_paste_simple.c and macro_paste_bad.c.  There are
several known bugs still.

llvm-svn: 38733

6 files changed, 126 insertions, 1 deletions

diff --git a/clang/Lex/MacroExpander.cpp b/clang/Lex/MacroExpander.cpp
index 0564cbe2a6e..7d28de546a1 100644
--- a/clang/Lex/MacroExpander.cpp
+++ b/clang/Lex/MacroExpander.cpp
@@ -16,6 +16,7 @@
 #include "clang/Lex/Preprocessor.h"
 #include "clang/Basic/SourceManager.h"
 #include "clang/Basic/Diagnostic.h"
+#include "llvm/Config/Alloca.h"
 using namespace llvm;
 using namespace clang;
 
@@ -373,6 +374,10 @@ void MacroExpander::Lex(LexerToken &Tok) {
   
   // Get the next token to return.
   Tok = (*MacroTokens)[CurToken++];
+  
+  // If this token is followed by a token paste (##) operator, paste the tokens!
+  if (!isAtEnd() && (*MacroTokens)[CurToken].getKind() == tok::hashhash)
+    PasteTokens(Tok);
 
   // The token's current location indicate where the token was lexed from.  We
   // need this information to compute the spelling of the token, but any
@@ -402,6 +407,100 @@ void MacroExpander::Lex(LexerToken &Tok) {
   // Otherwise, return a normal token.
 }
 
+/// PasteTokens - Tok is the LHS of a ## operator, and CurToken is the ##
+/// operator.  Read the ## and RHS, and paste the LHS/RHS together.  If there
+/// are is another ## after it, chomp it iteratively.  Return the result as Tok.
+void MacroExpander::PasteTokens(LexerToken &Tok) {
+  do {
+    // Consume the ## operator.
+    SourceLocation PasteOpLoc = (*MacroTokens)[CurToken].getLocation();
+    ++CurToken;
+    assert(!isAtEnd() && "No token on the RHS of a paste operator!");
+  
+    // Get the RHS token.
+    const LexerToken &RHS = (*MacroTokens)[CurToken];
+  
+    bool isInvalid = false;
+
+    // TODO: Avoid // and /*, as the lexer would think it is the start of a
+    // comment and emit warnings that don't make sense.
+    
+    // Allocate space for the result token.  This is guaranteed to be enough for
+    // the two tokens and a null terminator.
+    char *Buffer = (char*)alloca(Tok.getLength() + RHS.getLength() + 1);
+    
+    // Get the spelling of the LHS token in Buffer.
+    const char *BufPtr = Buffer;
+    unsigned LHSLen = PP.getSpelling(Tok, BufPtr);
+    if (BufPtr != Buffer)   // Really, we want the chars in Buffer!
+      memcpy(Buffer, BufPtr, LHSLen);
+    
+    BufPtr = Buffer+LHSLen;
+    unsigned RHSLen = PP.getSpelling(RHS, BufPtr);
+    if (BufPtr != Buffer+LHSLen)   // Really, we want the chars in Buffer!
+      memcpy(Buffer+LHSLen, BufPtr, RHSLen);
+    
+    // Add null terminator.
+    Buffer[LHSLen+RHSLen] = '\0';
+    
+    // Plop the pasted result (including the trailing newline and null) into a
+    // scratch buffer where we can lex it.
+    SourceLocation ResultTokLoc = PP.CreateString(Buffer, LHSLen+RHSLen+1);
+    
+    // Lex the resultant pasted token into Result.
+    LexerToken Result;
+    
+    // FIXME: Handle common cases: ident+ident, ident+simplenumber here.
+
+    // Make a lexer to lex this string from.
+    SourceManager &SourceMgr = PP.getSourceManager();
+    const char *ResultStrData = SourceMgr.getCharacterData(ResultTokLoc);
+    
+    unsigned FileID = ResultTokLoc.getFileID();
+    assert(FileID && "Could not get FileID for paste?");
+    
+    // Make and enter a lexer object so that we lex and expand the paste result.
+    Lexer *TL = new Lexer(SourceMgr.getBuffer(FileID), FileID, PP,
+                          ResultStrData, 
+                          ResultStrData+LHSLen+RHSLen /*don't include null*/);
+    
+    // Lex a token in raw mode.  This way it won't look up identifiers
+    // automatically, lexing off the end will return an eof token, and warnings
+    // are disabled.  This returns true if the result token is the entire
+    // buffer.
+    bool IsComplete = TL->LexRawToken(Result);
+    
+    // If we got an EOF token, we didn't form even ONE token.  For example, we
+    // did "/ ## /" to get "//".
+    IsComplete &= Result.getKind() != tok::eof;
+
+    // We're now done with the temporary lexer.
+    delete TL;
+    
+    // If pasting the two tokens didn't form a full new token, this is an error.
+    // This occurs with "x ## +"  and other stuff.
+    if (!IsComplete) {
+      // If not in assembler language mode.
+      PP.Diag(PasteOpLoc, diag::err_pp_bad_paste, 
+              std::string(Buffer, Buffer+LHSLen+RHSLen));
+      return;
+    }
+    
+    // Turn ## into 'other' to avoid # ## # from looking like a paste operator.
+    if (Result.getKind() == tok::hashhash)
+      Result.SetKind(tok::unknown);
+    // FIXME: Turn __VARRGS__ into "not a token"?
+    
+    // Transfer properties of the LHS over the the Result.
+    Result.SetFlagValue(LexerToken::StartOfLine , Tok.isAtStartOfLine());
+    Result.SetFlagValue(LexerToken::LeadingSpace, Tok.hasLeadingSpace());
+    
+    // Finally, replace LHS with the result, consume the RHS, and iterate.
+    ++CurToken;
+    Tok = Result;
+  } while (!isAtEnd() && (*MacroTokens)[CurToken].getKind() == tok::hashhash);
+}
+
 /// isNextTokenLParen - If the next token lexed will pop this macro off the
 /// expansion stack, return 2.  If the next unexpanded token is a '(', return
 /// 1, otherwise return 0.
diff --git a/clang/include/clang/Basic/DiagnosticKinds.def b/clang/include/clang/Basic/DiagnosticKinds.def
index 26d9f74095b..89c1089b80f 100644
--- a/clang/include/clang/Basic/DiagnosticKinds.def
+++ b/clang/include/clang/Basic/DiagnosticKinds.def
@@ -223,7 +223,8 @@ DIAG(err_too_many_args_in_macro_invoc, ERROR,
      "too many arguments provided to function-like macro invocation")
 DIAG(err_too_few_args_in_macro_invoc, ERROR,
      "too few arguments provided to function-like macro invocation")
-
+DIAG(err_pp_bad_paste, ERROR,
+     "pasting \"%s\" does not give a valid preprocessing token")
 
 // Should be a sorry?
 DIAG(err_pp_I_dash_not_supported, ERROR,
diff --git a/clang/include/clang/Lex/Lexer.h b/clang/include/clang/Lex/Lexer.h
index b7e8c53fa54..94133abacca 100644
--- a/clang/include/clang/Lex/Lexer.h
+++ b/clang/include/clang/Lex/Lexer.h
@@ -155,6 +155,17 @@ public:
     LexTokenInternal(Result);
   }
   
+  /// LexRawToken - Switch the lexer to raw mode, lex a token into Result and
+  /// switch it back.  Return true if the 'next character to read' pointer
+  /// points and the end of the lexer buffer, false otherwise.
+  bool LexRawToken(LexerToken &Result) {
+    assert(!LexingRawMode && "Already in raw mode!");
+    LexingRawMode = true;
+    Lex(Result);
+    LexingRawMode = false;
+    return BufferPtr == BufferEnd; 
+  }
+  
   /// ReadToEndOfLine - Read the rest of the current preprocessor line as an
   /// uninterpreted string.  This switches the lexer out of directive mode.
   std::string ReadToEndOfLine();
diff --git a/clang/include/clang/Lex/MacroExpander.h b/clang/include/clang/Lex/MacroExpander.h
index 1439a049c34..2835449038e 100644
--- a/clang/include/clang/Lex/MacroExpander.h
+++ b/clang/include/clang/Lex/MacroExpander.h
@@ -134,6 +134,12 @@ private:
     return CurToken == MacroTokens->size();
   }
   
+  /// PasteTokens - Tok is the LHS of a ## operator, and CurToken is the ##
+  /// operator.  Read the ## and RHS, and paste the LHS/RHS together.  If there
+  /// are is another ## after it, chomp it iteratively.  Return the result as
+  /// Tok.
+  void PasteTokens(LexerToken &Tok);
+  
   /// Expand the arguments of a function-like macro so that we can quickly
   /// return preexpanded tokens from MacroTokens.
   void ExpandFunctionArguments();
diff --git a/clang/test/Preprocessor/macro_paste_bad.c b/clang/test/Preprocessor/macro_paste_bad.c
new file mode 100644
index 00000000000..60caa427438
--- /dev/null
+++ b/clang/test/Preprocessor/macro_paste_bad.c
@@ -0,0 +1,5 @@
+// RUN: clang -Eonly %s 2>&1 | grep error
+// pasting ""x"" and ""+"" does not give a valid preprocessing token
+#define XYZ  x ## +
+XYZ
+
diff --git a/clang/test/Preprocessor/macro_paste_simple.c b/clang/test/Preprocessor/macro_paste_simple.c
new file mode 100644
index 00000000000..e8dc1e84045
--- /dev/null
+++ b/clang/test/Preprocessor/macro_paste_simple.c
@@ -0,0 +1,3 @@
+// clang %s -E | grep "barbaz123"
+
+#define FOO bar ## baz ## 123