X-Git-Url: http://git.tdb.fi/?a=blobdiff_plain;f=source%2Ftextparser.cpp;h=b8515a9b487fa2749d78c5c41a734c2f00c1b70f;hb=256b44a5009467171af53316141277027bcc0ba4;hp=225425b4e23166eea05fe0697f16974c327ab1c0;hpb=6653c7d83dbe1fe81a541a125be8bb808b234eb7;p=libs%2Fdatafile.git

diff --git a/source/textparser.cpp b/source/textparser.cpp
index 225425b..b8515a9 100644
--- a/source/textparser.cpp
+++ b/source/textparser.cpp
@@ -1,5 +1,6 @@
 #include <msp/strings/format.h>
 #include <msp/strings/utils.h>
+#include "except.h"
 #include "input.h"
 #include "textparser.h"
 #include "token.h"
@@ -9,33 +10,11 @@ using namespace std;
 namespace Msp {
 namespace DataFile {
 
-class parse_error: public runtime_error
-{
-public:
-	parse_error(const std::string &t):
-		runtime_error(t.empty() ? "at end of input" : format("after '%s'", t))
-	{ }
-
-	virtual ~parse_error() throw() { }
-};
-
-
-class syntax_error: public runtime_error
-{
-public:
-	syntax_error(const std::string &t):
-		runtime_error(t.empty() ? "at end of input" : format("at '%s'", t))
-	{ }
-
-	virtual ~syntax_error() throw() { }
-};
-
-
 TextParser::TextParser(Input &i, const string &s):
 	ParserMode(i, s)
 { }
 
-Statement TextParser::parse(bool)
+Statement TextParser::parse()
 {
 	return parse_statement(0);
 }
@@ -43,8 +22,7 @@ Statement TextParser::parse(bool)
 Statement TextParser::parse_statement(const Token *t)
 {
 	Statement result;
-	bool sub = false;
-	bool finish = false;
+	unsigned sub = 0;
 
 	while(in)
 	{
@@ -68,44 +46,28 @@ Statement TextParser::parse_statement(const Token *t)
 			result.source = src;
 			result.line = in.get_line_number();
 		}
-		else if(sub)
+		else if(sub==1)
 		{
 			if(token.str=="}")
-			{
-				sub = false;
-				finish = true;
-			}
+				sub = 2;
 			else
 			{
 				Statement ss = parse_statement(&token);
 				result.sub.push_back(ss);
 			}
 		}
-		else if(finish)
+		else if(sub==2)
 		{
 			if(token.str!=";")
 				throw syntax_error(token.str);
 			break;
 		}
 		else if(token.str=="{")
-			sub = true;
+			sub = 1;
 		else if(token.str==";")
 			break;
-		else if(token.type==Token::INTEGER)
-			result.append(lexical_cast<IntType::Store>(token.str));
-		else if(token.type==Token::FLOAT)
-			result.append(lexical_cast<FloatType::Store>(token.str));
-		else if(token.type==Token::STRING)
-			result.append(token.str);
-		else if(token.type==Token::IDENTIFIER)
-		{
-			if(token.str=="true")
-				result.append(true);
-			else if(token.str=="false")
-				result.append(false);
-			else
-				result.append(Symbol(token.str));
-		}
+		else if(token.type!=Token::SPECIAL)
+			result.append_from_token(token);
 		else
 			throw syntax_error(token.str);
 	}
@@ -124,9 +86,9 @@ Token TextParser::parse_token()
 		c = in.get();
 		int next = in.peek();
 
-		if(c=='/' && next=='/')
+		if(c=='/' && next=='/' && !comment)
 			comment = 1;
-		else if(c=='/' && next=='*')
+		else if(c=='/' && next=='*' && !comment)
 			comment = 2;
 		else if(c=='\n' && comment==1)
 			comment = 0;
@@ -150,6 +112,8 @@ Token TextParser::parse_token()
 		FLOATEXPINIT,
 		FLOATEXPSIGN,
 		STRING,
+		STRING_ESCAPE,
+		STRING_BASE64,
 		ACCEPT,
 		ZERO,
 		DECIMAL,
@@ -158,6 +122,7 @@ Token TextParser::parse_token()
 		FLOAT,
 		FLOATEXP,
 		STRING_END,
+		STRING_BASE64_PAD,
 		IDENTIFIER
 	};
 
@@ -169,6 +134,8 @@ Token TextParser::parse_token()
 		Token::SPECIAL,
 		Token::SPECIAL,
 		Token::SPECIAL,
+		Token::SPECIAL,
+		Token::SPECIAL,
 		Token::INTEGER,
 		Token::INTEGER,
 		Token::INTEGER,
@@ -176,12 +143,12 @@ Token TextParser::parse_token()
 		Token::FLOAT,
 		Token::FLOAT,
 		Token::STRING,
+		Token::STRING,
 		Token::IDENTIFIER
 	};
 
 	ParseState state = INIT;
 	string buf;
-	bool escape = false;
 
 	while(in || state==INIT)
 	{
@@ -202,6 +169,8 @@ Token TextParser::parse_token()
 				state = FLOAT;
 			else if(c=='"')
 				state = STRING;
+			else if(c=='=')
+				state = STRING_BASE64;
 			else if(c=='{' || c=='}' || c==';')
 				return Token(Token::SPECIAL, string(1, c));
 			else if(isdigit(c))
@@ -283,11 +252,25 @@ Token TextParser::parse_token()
 
 		case STRING:
 			if(c=='\\')
-				escape = !escape;
-			else if(c=='"' && !escape)
+				state = STRING_ESCAPE;
+			else if(c=='"')
 				state = STRING_END;
-			else
-				escape = false;
+			break;
+
+		case STRING_ESCAPE:
+			state = STRING;
+			break;
+
+		case STRING_BASE64:
+			if(c=='=')
+				state = STRING_BASE64_PAD;
+			else if(!isalnum(c) && c!='+' && c!='/')
+				throw parse_error(buf);
+			break;
+
+		case STRING_BASE64_PAD:
+			if(c!='=')
+				throw parse_error(buf);
 			break;
 
 		case IDENTIFIER:
@@ -308,6 +291,8 @@ Token TextParser::parse_token()
 				return Token(Token::IDENTIFIER, buf.substr(1));
 			else if(state==STRING_END)
 				return Token(Token::STRING, c_unescape(buf.substr(1, buf.size()-2)));
+			else if(state==STRING_BASE64 || state==STRING_BASE64_PAD)
+				return Token(Token::STRING, base64_decode(buf));
 			else
 				return Token(token_type[state], buf);
 		}
@@ -326,5 +311,42 @@ bool TextParser::isodigit(int c)
 	return (c>='0' && c<='7');
 }
 
+string TextParser::base64_decode(const string &data)
+{
+	string bin;
+	bin.reserve(data.size()*3/4);
+	unsigned accum = 0;
+	unsigned a_bits = 0;
+	for(char c: data)
+	{
+		unsigned d;
+		if(c>='A' && c<='Z')
+			d = c-'A';
+		else if(c>='a' && c<='z')
+			d = 26+(c-'a');
+		else if(c>='0' && c<='9')
+			d = 52+(c-'0');
+		else if(c=='+')
+			d = 62;
+		else if(c=='/')
+			d = 63;
+		else if(c=='=')
+			continue;
+		else
+			throw invalid_argument("TextParser::base64_decode");
+
+		accum = (accum<<6)|d;
+		a_bits += 6;
+
+		if(a_bits>=8)
+		{
+			bin += (accum>>(a_bits-8))&0xFF;
+			a_bits -= 8;
+		}
+	}
+
+	return bin;
+}
+
 } // namespace DataFile
 } // namespace Msp