parser: add necessary checks to operator() to stop accepting nonsense.

Since the parser is recursive parse_* methods (in particular,
parse_binop_rhs()) does NOT check if the last unparsed token is valid.
Thus parse_expression() stops if it encounters an unknown token. That's
why parser::operator() needs to make sure nothing is left in the input
stream.

diff --git a/check/parser_bugs.cpp b/check/parser_bugs.cpp
index a83f668a4b620a7bcbbcb74ab75306aa4fd4127c..af32c5ef89c81548eea433f16d9bd7ab882ed046 100644 (file)
--- a/check/parser_bugs.cpp
+++ b/check/parser_bugs.cpp
@@ -58,6 +58,22 @@ static int check3(std::ostream& err_str)
        return 0;
 }
 
+/// parser happily accepted various junk like 'x^2()+1'
+static int check4(std::ostream& err_str)
+{
+       const std::string junk("x^2()+1");
+       parser reader;
+       ex e;
+       try {
+               e = reader(junk);
+               err_str << "parser accepts junk: \"" << junk << "\"" << std::endl;
+               return 1;
+       } catch (parse_error& err) {
+               // Ok, parser rejects the nonsense.
+               return 0;
+       }
+}
+
 int main(int argc, char** argv)
 {
        std::cout << "checking for parser bugs. " << std::flush;
@@ -66,6 +82,7 @@ int main(int argc, char** argv)
        errors += check1(err_str);
        errors += check2(err_str);
        errors += check3(err_str);
+       errors += check4(err_str);
        if (errors) {
                std::cout << "Yes, unfortunately:" << std::endl;
                std::cout << err_str.str();

diff --git a/ginac/parser/parser.cpp b/ginac/parser/parser.cpp
index 6cac523c38625d8bddad69dcbc648d4b629665c4..6411ba1416e740a85705ceee3153815b33e9364f 100644 (file)
--- a/ginac/parser/parser.cpp
+++ b/ginac/parser/parser.cpp
@@ -138,6 +138,13 @@ ex parser::operator()(std::istream& input)
        scanner->switch_input(&input);
        get_next_tok();
        ex ret = parse_expression();
+       // parse_expression() stops if it encounters an unknown token.
+       // This is not a bug: since the parser is recursive checking
+       // whether the next token is valid is responsibility of the caller.
+       // Hence make sure nothing is left in the stream:
+       if (token != lexer::token_type::eof)
+               Parse_error("expected EOF");
+
        return ret;
 }