implement Feature.CSYNTAX, Feature.KEEPCOMMENTS, etc

8 files changed, 55 insertions, 47 deletions

diff --git a/src/java/org/anarres/cpp/CppReader.java b/src/java/org/anarres/cpp/CppReader.java
index 92778e0..a429fd4 100644
--- a/src/java/org/anarres/cpp/CppReader.java
+++ b/src/java/org/anarres/cpp/CppReader.java
@@ -95,8 +95,9 @@ public class CppReader extends Reader {
 					case EOF:
 						token = null;
 						return false;
-					case COMMENT:
-						if (false) {
+					case CCOMMENT:
+					case CPPCOMMENT:
+						if (!cpp.getFeature(Feature.KEEPCOMMENTS)) {
 							token = " ";
 							break;
 						}
diff --git a/src/java/org/anarres/cpp/Feature.java b/src/java/org/anarres/cpp/Feature.java
index 629b0ec..2fc7a61 100644
--- a/src/java/org/anarres/cpp/Feature.java
+++ b/src/java/org/anarres/cpp/Feature.java
@@ -24,4 +24,6 @@ public enum Feature {
 	DIGRAPHS,
 	TRIGRAPHS,
 	LINEMARKERS,
+	CSYNTAX,
+	KEEPCOMMENTS,
 }
diff --git a/src/java/org/anarres/cpp/LexerSource.java b/src/java/org/anarres/cpp/LexerSource.java
index 0735e28..44c6224 100644
--- a/src/java/org/anarres/cpp/LexerSource.java
+++ b/src/java/org/anarres/cpp/LexerSource.java
@@ -250,7 +250,7 @@ public class LexerSource extends Source {
 				text.append((char)d);
 			} while (d == '*');
 		} while (d != '/');
-		return new Token(COMMENT, text.toString());
+		return new Token(CCOMMENT, text.toString());
 	}
 
 	private Token cppcomment()
@@ -263,7 +263,7 @@ public class LexerSource extends Source {
 			d = read();
 		}
 		unread(d);
-		return new Token(COMMENT, text.toString());
+		return new Token(CPPCOMMENT, text.toString());
 	}
 
 	private int escape(StringBuilder text)
@@ -326,13 +326,13 @@ public class LexerSource extends Source {
 		else if (isLineSeparator(d)) {
 			unread(d);
 			// error("Unterminated character literal");
-			return new Token(ERROR, text.toString(),
+			return new Token(INVALID, text.toString(),
 							"Unterminated character literal");
 		}
 		else if (d == '\'') {
 			text.append('\'');
 			// error("Empty character literal");
-			return new Token(ERROR, text.toString(),
+			return new Token(INVALID, text.toString(),
 							"Empty character literal");
 		}
 		else if (!Character.isDefined(d)) {
@@ -357,7 +357,7 @@ public class LexerSource extends Source {
 				text.append((char)e);
 				e = read();
 			}
-			return new Token(ERROR, text.toString(),
+			return new Token(INVALID, text.toString(),
 							"Illegal character constant");
 		}
 		text.append('\'');
@@ -389,13 +389,13 @@ public class LexerSource extends Source {
 			else if (c == -1) {
 				unread(c);
 				// error("End of file in string literal after " + buf);
-				return new Token(ERROR, text.toString(),
+				return new Token(INVALID, text.toString(),
 						"End of file in string literal after " + buf);
 			}
 			else if (isLineSeparator(c)) {
 				unread(c);
 				// error("Unterminated string literal after " + buf);
-				return new Token(ERROR, text.toString(),
+				return new Token(INVALID, text.toString(),
 						"Unterminated string literal after " + buf);
 			}
 			else {
@@ -435,7 +435,7 @@ public class LexerSource extends Source {
 			}
 			else if (Character.isLetter(d)) {
 				unread(d);
-				return new Token(ERROR, text.toString(),
+				return new Token(INVALID, text.toString(),
 						"Invalid suffix \"" + (char)d +
 						"\" on numeric constant");
 			}
@@ -472,7 +472,7 @@ public class LexerSource extends Source {
 		if (Character.digit(d, 16) == -1) {
 			unread(d);
 			// error("Illegal hexadecimal constant " + (char)d);
-			return new Token(ERROR, text.toString(),
+			return new Token(INVALID, text.toString(),
 					"Illegal hexadecimal digit " + (char)d +
 					" after "+ text);
 		}
diff --git a/src/java/org/anarres/cpp/MacroTokenSource.java b/src/java/org/anarres/cpp/MacroTokenSource.java
index b6500ff..264a1a6 100644
--- a/src/java/org/anarres/cpp/MacroTokenSource.java
+++ b/src/java/org/anarres/cpp/MacroTokenSource.java
@@ -111,9 +111,10 @@ import static org.anarres.cpp.Token.*;
 		int	count = 2;
 		for (int i = 0; i < count; i++) {
 			if (!tokens.hasNext()) {
-				err = new Token(ERROR,
-						ptok.getLine(), ptok.getColumn(),
-						ptok.getText(), "Paste at end of expansion");
+				/* XXX This one really should throw. */
+				error(ptok.getLine(), ptok.getColumn(),
+						"Paste at end of expansion");
+				buf.append(' ').append(ptok.getText());
 				break;
 			}
 			Token	tok = tokens.next();
@@ -129,7 +130,8 @@ import static org.anarres.cpp.Token.*;
 					concat(buf, args.get(idx));
 					break;
 				/* XXX Test this. */
-				case COMMENT:
+				case CCOMMENT:
+				case CPPCOMMENT:
 					break;
 				default:
 					buf.append(tok.getText());
@@ -156,10 +158,9 @@ import static org.anarres.cpp.Token.*;
 			if (arg != null) {
 				if (arg.hasNext()) {
 					Token	tok = arg.next();
-					if (tok.getType() == M_PASTE)
-						tok = new Token(ERROR,
-								tok.getLine(), tok.getColumn(),
-								tok.getText(), "Unexpected paste token");
+					/* XXX PASTE -> INVALID. */
+					assert tok.getType() != M_PASTE :
+								"Unexpected paste token";
 					return tok;
 				}
 				arg = null;
diff --git a/src/java/org/anarres/cpp/Main.java b/src/java/org/anarres/cpp/Main.java
index b7e11df..0b97f4e 100644
--- a/src/java/org/anarres/cpp/Main.java
+++ b/src/java/org/anarres/cpp/Main.java
@@ -101,8 +101,9 @@ public class Main {
 		pp.addFeature(Feature.LINEMARKERS);
 		pp.addWarning(Warning.IMPORT);
 		pp.setListener(new PreprocessorListener());
-
 		pp.addMacro("__JCPP__");
+		pp.getSystemIncludePath().add("/usr/local/include");
+		pp.getSystemIncludePath().add("/usr/include");
 
         GETOPT: while ((c = g.getopt()) != -1) {
             switch (c) {
@@ -157,11 +158,6 @@ public class Main {
 			}
 		}
 
-		List<String>	path = pp.getSystemIncludePath();
-		path.add("/usr/local/include");
-		path.add("/usr/include");
-		// path.add("/usr/lib/gcc/i686-pc-linux-gnu/4.1.2/include");
-
 		for (int i = g.getOptind(); i < args.length; i++)
 			pp.addInput(new FileLexerSource(new File(args[i])));
 
diff --git a/src/java/org/anarres/cpp/Preprocessor.java b/src/java/org/anarres/cpp/Preprocessor.java
index 4547966..a1a72d9 100644
--- a/src/java/org/anarres/cpp/Preprocessor.java
+++ b/src/java/org/anarres/cpp/Preprocessor.java
@@ -226,9 +226,6 @@ public class Preprocessor {
 	/**
 	 * Adds input for the Preprocessor.
 	 *
-	 * XXX Inputs should be maintained off the source stack.
-	 * <p>
-	 *
 	 * Inputs are processed in the order in which they are added.
 	 */
 	public void addInput(Source source) {
@@ -530,7 +527,9 @@ public class Preprocessor {
 
 	private boolean isWhite(Token tok) {
 		int	type = tok.getType();
-		return (type == WHITESPACE) || (type == COMMENT);
+		return (type == WHITESPACE)
+			|| (type == CCOMMENT)
+			|| (type == CPPCOMMENT);
 	}
 
 	private Token source_token_nonwhite()
@@ -571,7 +570,8 @@ public class Preprocessor {
 				// System.out.println("pp: open: token is " + tok);
 				switch (tok.getType()) {
 					case WHITESPACE:	/* XXX Really? */
-					case COMMENT:
+					case CCOMMENT:
+					case CPPCOMMENT:
 					case NL:
 						break;	/* continue */
 					case '(':
@@ -638,7 +638,8 @@ public class Preprocessor {
 							break;
 
 						case WHITESPACE:
-						case COMMENT:
+						case CCOMMENT:
+						case CPPCOMMENT:
 							/* Avoid duplicating spaces. */
 							space = true;
 							break;
@@ -671,6 +672,11 @@ public class Preprocessor {
 					return false;
 				}
 
+				/*
+				for (Argument a : args)
+					a.expand(this);
+				*/
+
 				for (int i = 0; i < args.size(); i++) {
 					args.get(i).expand(this);
 				}
@@ -748,7 +754,8 @@ public class Preprocessor {
 					break EXPANSION;
 
 				case WHITESPACE:
-				case COMMENT:
+				case CCOMMENT:
+				case CPPCOMMENT:
 					space = true; 
 					break;
 
@@ -863,7 +870,8 @@ public class Preprocessor {
 				case NL:
 					break EXPANSION;
 
-				case COMMENT:
+				case CCOMMENT:
+				case CPPCOMMENT:
 					// break;
 				case WHITESPACE:
 					if (!paste)
@@ -1109,7 +1117,8 @@ public class Preprocessor {
 					warning(tok,
 						"Empty #" + "pragma");
 					return tok;
-				case COMMENT:
+				case CCOMMENT:
+				case CPPCOMMENT:
 				case WHITESPACE:
 					continue NAME;
 				case IDENTIFIER:
@@ -1135,7 +1144,8 @@ public class Preprocessor {
 				case NL:
 					/* This may contain one or more newlines. */
 					break VALUE;
-				case COMMENT:
+				case CCOMMENT:
+				case CPPCOMMENT:
 					break;
 				case WHITESPACE:
 					value.add(tok);
@@ -1488,7 +1498,8 @@ public class Preprocessor {
 						/* The preprocessor has to take action here. */
 						break;
 					case WHITESPACE:
-					case COMMENT:
+					case CCOMMENT:
+					case CPPCOMMENT:
 						// Patch up to preserve whitespace.
 						/* XXX We might want to return tok here in C */
 						return toWhitespace(tok);
@@ -1510,7 +1521,8 @@ public class Preprocessor {
 				case NL:
 					return tok;
 
-				case COMMENT:
+				case CCOMMENT:
+				case CPPCOMMENT:
 					return tok;
 
 				case '!': case '%': case '&':
@@ -1570,14 +1582,9 @@ public class Preprocessor {
 						return tok;
 					break;
 
-				case ERROR:
-					PreprocessorListener	l = getListener();
-					if (l != null) {
-						l.handleError(getSource(),
-								tok.getLine(), tok.getColumn(),
-								String.valueOf(tok.getValue()));
-						break;
-					}
+				case INVALID:
+					if (features.contains(Feature.CSYNTAX))
+						error(tok, String.valueOf(tok.getValue()));
 					return tok;
 
 				default:
diff --git a/src/java/org/anarres/cpp/Source.java b/src/java/org/anarres/cpp/Source.java
index 1ce9f47..f6cd3b8 100644
--- a/src/java/org/anarres/cpp/Source.java
+++ b/src/java/org/anarres/cpp/Source.java
@@ -255,7 +255,8 @@ public abstract class Source implements Iterable<Token> {
 				case NL:
 					/* This may contain one or more newlines. */
 					return tok;
-				case COMMENT:
+				case CCOMMENT:
+				case CPPCOMMENT:
 				case WHITESPACE:
 					break;
 				default:
diff --git a/src/java/org/anarres/cpp/Token.java b/src/java/org/anarres/cpp/Token.java
index 537704b..bb56531 100644
--- a/src/java/org/anarres/cpp/Token.java
+++ b/src/java/org/anarres/cpp/Token.java
@@ -151,7 +151,7 @@ public final class Token {
 	}
 
 #set ($i = 257)
-#set ($tokens = [ "AND_EQ", "ARROW", "CHARACTER", "COMMENT", "DEC", "DIV_EQ", "ELLIPSIS", "EOF", "EQ", "GE", "HASH", "HEADER", "IDENTIFIER", "INC", "INTEGER", "LAND", "LAND_EQ", "LE", "LITERAL", "LOR", "LOR_EQ", "LSH", "LSH_EQ", "MOD_EQ", "MULT_EQ", "NE", "NL", "OR_EQ", "PASTE", "PLUS_EQ", "RANGE", "RSH", "RSH_EQ", "STRING", "SUB_EQ", "WHITESPACE", "XOR_EQ", "M_ARG", "M_PASTE", "M_STRING", "P_LINE", "ERROR" ])
+#set ($tokens = [ "AND_EQ", "ARROW", "CHARACTER", "CCOMMENT", "CPPCOMMENT", "DEC", "DIV_EQ", "ELLIPSIS", "EOF", "EQ", "GE", "HASH", "HEADER", "IDENTIFIER", "INC", "INTEGER", "LAND", "LAND_EQ", "LE", "LITERAL", "LOR", "LOR_EQ", "LSH", "LSH_EQ", "MOD_EQ", "MULT_EQ", "NE", "NL", "OR_EQ", "PASTE", "PLUS_EQ", "RANGE", "RSH", "RSH_EQ", "STRING", "SUB_EQ", "WHITESPACE", "XOR_EQ", "M_ARG", "M_PASTE", "M_STRING", "P_LINE", "INVALID" ])
 #foreach ($token in $tokens)
 	/** The token type $token. */
 	public static final int $token = $i;