c-based: fix to handle edge case

Units/parser-java.r/java_enum.java.d/expected.tags

@@ -5,6 +5,12 @@ FIRST_VALUE	input.java	/^		FIRST_VALUE,$/;"	e	enum:C.TrivialEnum	file:	end:3
 FancyEnum	input.java	/^		FancyEnum(int i) {$/;"	m	class:C.FancyEnum	end:11
 FancyEnum	input.java	/^	public enum FancyEnum {$/;"	g	class:C	end:14
 SECOND_VALUE	input.java	/^		SECOND_VALUE$/;"	e	enum:C.TrivialEnum	file:	end:4
+StringEnum	input.java	/^	private StringEnum(String _s) {$/;"	m	class:StringEnum	file:	end:25
+StringEnum	input.java	/^public enum StringEnum {$/;"	g	end:28
 TrivialEnum	input.java	/^	public enum TrivialEnum {$/;"	g	class:C	end:5
+X	input.java	/^	X("X"),$/;"	e	enum:StringEnum	file:	end:19
+Y	input.java	/^	Y("Y"),$/;"	e	enum:StringEnum	file:	end:20
+Z	input.java	/^	Z("Z");$/;"	e	enum:StringEnum	file:	end:21
 i	input.java	/^		private int i;$/;"	f	class:C.FancyEnum	file:	end:8
 m	input.java	/^		void m() {$/;"	m	class:C.FancyEnum	end:13
+s	input.java	/^	private final String s;$/;"	f	class:StringEnum	file:	end:27

Units/parser-java.r/java_enum.java.d/input.java

@@ -13,3 +13,16 @@ void m() {
 		}
 	}
 }
+
+public enum StringEnum {
+
+	X("X"),
+	Y("Y"),
+	Z("Z");
+
+	private StringEnum(String _s) {
+		s = _s;
+	}
+
+	private final String s;
+}

parsers/c-based.c

@@ -1486,28 +1486,8 @@
 	while (matchLevel > 0  &&  (c = skipToNonWhite ()) != EOF)
 	{
 		if (CollectingSignature)
-		{
-			if (c <= 0xff)
-				vStringPut (Signature, c);
-			else
-			{
-				char marker = '"';
+			cStringPut (Signature, c);
 
-				switch (c)
-				{
-					case CHAR_SYMBOL:
-						marker = '\'';
-						/* Fall through */
-					case STRING_SYMBOL:
-						vStringPut (Signature, marker);
-						vStringCat (Signature, cppGetLastCharOrStringContents ());
-						vStringPut (Signature, marker);
-						break;
-					default:
-						AssertNotReached();
-				}
-			}
-		}
 		if (c == begin)
 		{
 			++matchLevel;
@@ -1592,11 +1572,11 @@
 
 	do
 	{
-		vStringPut (name, c);
+		cStringPut (name, c);
 		if (CollectingSignature)
 		{
 			if (!first)
-				vStringPut (Signature, c);
+				cStringPut (Signature, c);
 			first = false;
 		}
 		c = cppGetc ();
@@ -1725,7 +1705,7 @@
 					vStringPut (name, ' ');
 					whiteSpace = false;
 				}
-				vStringPut (name, c);
+				cStringPut (name, c);
 			}
 			c = cppGetc ();
 		} while (! isOneOf (c, "(;")  &&  c != EOF);
@@ -1735,7 +1715,7 @@
 		vStringPut (name, ' ');  /* always separate operator from keyword */
 		do
 		{
-			vStringPut (name, c);
+			vStringPut (name, c);	/* acceptable are all ascii */
 			c = cppGetc ();
 		} while (isOneOf (c, acceptable));
 	}
@@ -2316,8 +2296,8 @@
 	do
 	{
 		int c = skipToNonWhite ();
-		vStringPut (Signature, c);
 
+		cStringPut (Signature, c);
 		switch (c)
 		{
 			case '^':

parsers/cpreprocessor.c

@@ -461,6 +461,20 @@
 /*  This puts a character back into the input queue for the input File. */
 extern void cppUngetc (const int c)
 {
+	if (c == STRING_SYMBOL || c == CHAR_SYMBOL)
+	{
+		Assert(Cpp.charOrStringContents != NULL);
+		cppUngetc(c == STRING_SYMBOL ? '"' : '\'');
+		cppUngetString(vStringValue(Cpp.charOrStringContents), vStringLength(Cpp.charOrStringContents));
+		cppUngetc(c == STRING_SYMBOL ? '"' : '\'');
+		vStringClear(Cpp.charOrStringContents);
+		return;
+	}
+	else if (c == EOF)
+	{
+		return;
+	}
+
 	if(!Cpp.ungetPointer)
 	{
 		// no unget data
@@ -1520,6 +1534,29 @@
 		vStringPut(condition, c);
 }
 
+extern void cStringPut (vString* string, const int c)
+{
+	if (c <= 0xff)
+		vStringPut (string, c);
+	else
+	{
+		char marker = '"';
+		switch (c)
+		{
+			case CHAR_SYMBOL:
+				marker = '\'';
+				/* Fall through */
+			case STRING_SYMBOL:
+				vStringPut (string, marker);
+				vStringCat (string, cppGetLastCharOrStringContents ());
+				vStringPut (string, marker);
+				break;
+			default:
+				AssertNotReached();
+		}
+	}
+}
+
 /*  This function returns the next character, stripping out comments,
  *  C pre-processor directives, and the contents of single and double
  *  quoted strings. In short, strip anything which places a burden upon

parsers/cpreprocessor.h

@@ -100,6 +100,11 @@ extern void cppUngetString(const char * string,int len);
 extern int cppGetc (void);
 extern const vString * cppGetLastCharOrStringContents (void);
 
+/*
+ * Replacement for vStringPut that can handle c > 0xff
+ */
+extern void cStringPut (vString * string, const int c);
+
 /* Notify the external parser state for the purpose of conditional
  * branch choice. The CXX parser stores the block level here. */
 extern void cppPushExternalParserBlock(void);

parsers/cxx/cxx_parser_tokenizer.c

@@ -857,7 +857,7 @@ static CXXCharTypeData g_aCharTable[128] =
 		0 ,
 		0
 	},
-	// 127 (0x7f) ''
+	// 127 (0x7f)
 	{ 0, 0, 0 }
 };
 
@@ -1491,9 +1491,7 @@ bool cxxParserParseNextToken(void)
 	if(g_cxx.iChar == STRING_SYMBOL)
 	{
 		t->eType = CXXTokenTypeStringConstant;
-		vStringPut(t->pszWord,'"');
-		vStringCat(t->pszWord,cppGetLastCharOrStringContents());
-		vStringPut(t->pszWord,'"');
+		cStringPut(t->pszWord,g_cxx.iChar);
 		g_cxx.iChar = cppGetc();
 		t->bFollowedBySpace = cppIsspace(g_cxx.iChar);
 		return true;
@@ -1539,9 +1537,7 @@ bool cxxParserParseNextToken(void)
 	if(g_cxx.iChar == CHAR_SYMBOL)
 	{
 		t->eType = CXXTokenTypeCharacterConstant;
-		vStringPut(t->pszWord,'\'');
-		vStringCat(t->pszWord,cppGetLastCharOrStringContents());
-		vStringPut(t->pszWord,'\'');
+		cStringPut(t->pszWord,g_cxx.iChar);
 		g_cxx.iChar = cppGetc();
 		t->bFollowedBySpace = cppIsspace(g_cxx.iChar);
 		return true;
@@ -1683,7 +1679,7 @@ bool cxxParserParseNextToken(void)
 	}
 
 	t->eType = CXXTokenTypeUnknown;
-	vStringPut(t->pszWord,g_cxx.iChar);
+	cStringPut(t->pszWord,g_cxx.iChar);
 	g_cxx.iChar = cppGetc();
 	t->bFollowedBySpace = cppIsspace(g_cxx.iChar);