From 83314506d701a8271fe56bc999a8f6acc29fad9b Mon Sep 17 00:00:00 2001
From: kevin <kpl@larke.org>
Date: Mon, 6 Jan 2014 16:10:05 -0500
Subject: [PATCH] cmLex.h/c : Added optional quoted character token recognizer.

---
 cmLex.c | 19 +++++++++++++++++++
 cmLex.h | 30 ++++++++++++++++--------------
 2 files changed, 35 insertions(+), 14 deletions(-)

diff --git a/cmLex.c b/cmLex.c
index 0ea0f35..7267493 100644
--- a/cmLex.c
+++ b/cmLex.c
@@ -120,6 +120,8 @@ cmRC_t _cmLexError( cmLex* p, unsigned rc, const char* fmt, ... )
   return rc;
 }
 
+// Locate 'keyStr' in cp[cn] and return the index into cp[cn] of the character
+// following the last char in 'keyStr'.  If keyStr is not found return cmInvalidIdx.
 unsigned _cmLexScanTo( const cmChar_t* cp, unsigned cn, const cmChar_t* keyStr )
 {
   unsigned i = 0;
@@ -336,6 +338,20 @@ unsigned _cmLexQStrMatcher( cmLex* p, const cmChar_t* cp, unsigned cn, const cmC
   return 0;
 }
 
+unsigned _cmLexQCharMatcher( cmLex* p, const cmChar_t* cp, unsigned cn, const cmChar_t* keyStr )
+{
+  unsigned i = 0;
+  if( i >= cn || cp[i]!='\'' )
+    return 0;
+
+  i+=2;
+
+  if( i >= cn || cp[i]!='\'')
+    return 0;
+
+  return 3;
+}
+
 
 unsigned _cmLexBlockCmtMatcher( cmLex* p, const cmChar_t* cp, unsigned cn, const cmChar_t* keyStr )
 {  
@@ -474,6 +490,9 @@ cmLexH cmLexInit( const cmChar_t* cp, unsigned cn, unsigned flags, cmRpt_t* rpt
   _cmLexInstallMatcher( p, kBlockCmtLexTId, _cmLexBlockCmtMatcher, NULL, NULL  );
   _cmLexInstallMatcher( p, kLineCmtLexTId,  _cmLexLineCmtMatcher,  NULL, NULL  );
 
+  if( cmIsFlag(flags,kReturnQCharLexFl) )
+    _cmLexInstallMatcher( p, kQCharLexTId,    _cmLexQCharMatcher,    NULL, NULL  );
+
   h.h = p;
 
   _cmLexReset(p);
diff --git a/cmLex.h b/cmLex.h
index 5eed6aa..c213e0c 100644
--- a/cmLex.h
+++ b/cmLex.h
@@ -11,18 +11,19 @@
 // Predefined Lexer Id's
 enum
 {
-  kErrorLexTId,    // 0  the lexer was unable to identify the current token
-  kUnknownLexTId,  // 1  the token is of an unknown type (only used when kReturnUnknownLexFl is set)
-  kEofLexTId,      // 2  the lexer reached the end of input
-  kSpaceLexTId,    // 3  white space
-  kRealLexTId,     // 4  real number (contains a decimal point or is in scientific notation) 
-  kIntLexTId,      // 5  decimal integer
-  kHexLexTId,      // 6  hexidecimal integer
-  kIdentLexTId,    // 7  identifier
-  kQStrLexTId,     // 8  quoted string
-  kBlockCmtLexTId, // 9  block comment
-  kLineCmtLexTId,  // 10  line comment
-  kUserLexTId      // 11 user registered token (See cmLexRegisterToken().)
+  kErrorLexTId,    //  0  the lexer was unable to identify the current token
+  kUnknownLexTId,  //  1  the token is of an unknown type (only used when kReturnUnknownLexFl is set)
+  kEofLexTId,      //  2  the lexer reached the end of input
+  kSpaceLexTId,    //  3  white space
+  kRealLexTId,     //  4  real number (contains a decimal point or is in scientific notation) 
+  kIntLexTId,      //  5  decimal integer
+  kHexLexTId,      //  6  hexidecimal integer
+  kIdentLexTId,    //  7  identifier
+  kQStrLexTId,     //  8  quoted string
+  kQCharLexTId,    //  9  quoted char
+  kBlockCmtLexTId, // 10  block comment
+  kLineCmtLexTId,  // 11  line comment
+  kUserLexTId      // 12 user registered token (See cmLexRegisterToken().)
 };
 
 // Lexer control flags used with cmLexInit().
@@ -31,7 +32,8 @@ enum
   kReturnSpaceLexFl    = 0x01, //< Return space tokens
   kReturnCommentsLexFl = 0x02, //< Return comment tokens
   kReturnUnknownLexFl  = 0x04, //< Return unknown tokens
-  kUserDefPriorityLexFl= 0x08  //< User defined tokens take priority even if a kIdentLexTId token has a longer match
+  kReturnQCharLexFl    = 0x08, //< Return quoted characters
+  kUserDefPriorityLexFl= 0x10  //< User defined tokens take priority even if a kIdentLexTId token has a longer match
 };
 
 // cmLex result codes.
@@ -84,7 +86,7 @@ cmRC_t             cmLexSetFile( cmLexH h, const cmChar_t* fn );
 cmRC_t             cmLexRegisterToken( cmLexH h, unsigned id, const cmChar_t* token );
 
 // Register a user defined token recognition function.  This function should return the count
-// of initial, consecutive, characters in 'cp' which match its token pattern.
+// of initial, consecutive, characters in 'cp[cn]' which match its token pattern.
 typedef unsigned (*cmLexUserMatcherPtr_t)( const cmChar_t* cp, unsigned cn );
 
 cmRC_t             cmLexRegisterMatcher( cmLexH h, unsigned id, cmLexUserMatcherPtr_t funcPtr );