X-Git-Url: http://git.tdb.fi/?a=blobdiff_plain;f=source%2Ftextparser.cpp;h=0f17d3f3cc79808a7659e291e406f6b7bb814137;hb=256f7238bc60d6dcc31a564988f5cc02a60c4537;hp=78d2f1df6334b32dbf67a7871975deda4dafb0a4;hpb=2f79370bffe0bac865dc97c5114dc87c1936fbb4;p=libs%2Fdatafile.git

diff --git a/source/textparser.cpp b/source/textparser.cpp
index 78d2f1d..0f17d3f 100644
--- a/source/textparser.cpp
+++ b/source/textparser.cpp
@@ -1,7 +1,7 @@
 /* $Id$
 
 This file is part of libmspdatafile
-Copyright © 2007  Mikko Rasa, Mikkosoft Productions
+Copyright © 2007-2008  Mikko Rasa, Mikkosoft Productions
 Distributed under the LGPL
 */
 
@@ -47,7 +47,7 @@ Statement TextParser::parse_statement(const Token *t)
 			if(token.str.empty())
 				break;
 			else if(token.type!=Token::IDENTIFIER)
-				throw ParseError(format("%s: Syntax error at token '%s' (expected an identifier)", get_location(), token.str), src, in.get_line_number());
+				throw_at(ParseError(format("Syntax error at token '%s' (expected an identifier)", token.str)), get_location());
 			result.keyword=token.str;
 			result.valid=true;
 			result.source=src;
@@ -69,7 +69,7 @@ Statement TextParser::parse_statement(const Token *t)
 		else if(finish)
 		{
 			if(token.str!=";")
-				throw ParseError(format("%s: Syntax error at token '%s' (Expected a ';')", get_location(), token.str), src, in.get_line_number());
+				throw_at(ParseError(format("Syntax error at token '%s' (Expected a ';')", token.str)), get_location());
 			break;
 		}
 		else if(token.str=="{")
@@ -93,9 +93,9 @@ Statement TextParser::parse_statement(const Token *t)
 			//result.args.push_back(resolve_identifiertoken.str);
 		}
 		else if(token.str=="")
-			throw ParseError(src+": Unexcepted end of input", src, in.get_line_number());
+			throw_at(ParseError("Unexcepted end of input"), get_location());
 		else
-			throw ParseError(get_location()+": Syntax error", src, in.get_line_number());
+			throw_at(ParseError("Syntax error"), get_location());
 	}
 
 	return result;
@@ -104,10 +104,10 @@ Statement TextParser::parse_statement(const Token *t)
 Token TextParser::parse_token()
 {
 	int c=0;
-	unsigned comment=0;
+	int comment=0;
 
 	// Skip over comments and whitespace
-	while(in)
+	while(in && comment>=0)
 	{
 		c=in.get();
 		int next=in.peek();
@@ -123,11 +123,13 @@ Token TextParser::parse_token()
 		else if(comment==3)   // Skip the second character of block comment end
 			comment=0;
 		else if(!isspace(c) && !comment)
-			break;
+			comment=-1;
 	}
 
-	if(comment)  // Didn't hit any non-whitespace
-		throw ParseError(src+": Unfinished comment at end of input", src, in.get_line_number());
+	if(comment>0)  // EOF while in comment
+		throw_at(ParseError("Unfinished comment at end of input"), get_location());
+	else if(comment==0)  // Didn't hit any non-whitespace
+		return Token(Token::SPECIAL, "");
 
 	enum ParseState
 	{
@@ -167,7 +169,7 @@ Token TextParser::parse_token()
 	string     buf;
 	bool       escape=false;
 
-	while(in)
+	while(in || state==INIT)
 	{
 		if(state!=INIT)
 			c=in.get();
@@ -193,7 +195,7 @@ Token TextParser::parse_token()
 			else if(isalpha(c) || c=='_')
 				state=IDENTIFIER;
 			else
-				parse_error(c, state);
+				parse_error(c, "0-9A-Za-z_.\"{};+-");
 			break;
 
 		case SIGN:
@@ -204,7 +206,7 @@ Token TextParser::parse_token()
 			else if(c=='.')
 				state=FLOAT;
 			else
-				parse_error(c, state);
+				parse_error(c, "0-9.");
 			break;
 
 		case ZERO:
@@ -215,31 +217,31 @@ Token TextParser::parse_token()
 			else if(c=='.')
 				state=FLOAT;
 			else
-				parse_error(c, state);
+				parse_error(c, "0-9A-Fa-f.");
 			break;
 
 		case DECIMAL:
 			if(c=='.')
 				state=FLOAT;
 			else if(!isdigit(c))
-				parse_error(c, state);
+				parse_error(c, "0-9.");
 			break;
 
 		case HEXADECIMAL:
 			if(!isxdigit(c))
-				parse_error(c, state);
+				parse_error(c, "0-9A-Fa-f");
 			break;
 
 		case OCTAL:
 			if(!isodigit(c))
-				parse_error(c, state);
+				parse_error(c, "0-7");
 			break;
 
 		case FLOAT:
 			if(c=='e' || c=='E')
 				state=FLOATEXPINIT;
 			else if(!isdigit(c))
-				parse_error(c, state);
+				parse_error(c, "0-9Ee");
 			break;
 
 		case FLOATEXPINIT:
@@ -248,19 +250,19 @@ Token TextParser::parse_token()
 			else if(isdigit(c))
 				state=FLOATEXP;
 			else
-				parse_error(c, state);
+				parse_error(c, "0-9+-");
 			break;
 
 		case FLOATEXPSIGN:
 			if(isdigit(c))
 				state=FLOATEXP;
 			else
-				parse_error(c, state);
+				parse_error(c, "0-9");
 			break;
 
 		case FLOATEXP:
 			if(!isdigit(c))
-				parse_error(c, state);
+				parse_error(c, "0-9");
 			break;
 
 		case STRING:
@@ -272,9 +274,10 @@ Token TextParser::parse_token()
 				{
 					return Token(Token::STRING, c_unescape(buf.substr(1, buf.size()-2)));
 				}
-				catch(const Exception &e)
+				catch(Exception &e)
 				{
-					throw ParseError(format("%s: %s", get_location(), e.what()), src, in.get_line_number());
+					e.at(get_location());
+					throw;
 				}
 			}
 			else
@@ -283,11 +286,11 @@ Token TextParser::parse_token()
 
 		case IDENTIFIER:
 			if(!isalpha(c) && !isdigit(c) && c!='_')
-				parse_error(c, state);
+				parse_error(c, "0-9A-Za-z_");
 			break;
 
 		default:
-			throw Exception(get_location()+": Internal error (bad state)");
+			throw_at(InvalidState("Internal error (bad state)"), get_location());
 		}
 
 		if(is_delimiter(next) && state>=ACCEPT)
@@ -314,9 +317,9 @@ string TextParser::get_location()
 	return ss.str();
 }
 
-void TextParser::parse_error(int c, int state)
+void TextParser::parse_error(int c, const char *e)
 {
-	throw ParseError(format("%s: Parse error at '%c' (state %d)", get_location(), static_cast<char>(c), state), src, in.get_line_number());
+	throw_at(ParseError(format("Parse error at '%c', expected one of \"%s\"", static_cast<char>(c), e)), get_location());
 }
 
 } // namespace DataFile