Sped up PRISM parser by letting it skip the actual command definitions in the first run (because only gathering constants, variables and formulas is important in this particular run).

Former-commit-id: 0b25c73fa4

src/parser/PrismParser.cpp

@@ -139,7 +139,7 @@ namespace storm {
             updateListDefinition %= +updateDefinition(qi::_r1) % "+";
             updateListDefinition.name("update list");
             
-            commandDefinition = (qi::lit("[") > -(identifier[qi::_a = qi::_1]) > qi::lit("]") > expressionParser > qi::lit("->") > updateListDefinition(qi::_r1) > qi::lit(";"))[qi::_val = phoenix::bind(&PrismParser::createCommand, phoenix::ref(*this), qi::_a, qi::_2, qi::_3, qi::_r1)];
+            commandDefinition = qi::lit("[") > +(qi::char_ - qi::lit(";")) > qi::lit(";")[qi::_val = phoenix::construct<storm::prism::Command>()];
             commandDefinition.name("command definition");
             
             moduleDefinition = ((qi::lit("module") >> identifier >> *(variableDefinition(qi::_a, qi::_b))) > +commandDefinition(qi::_r1) > qi::lit("endmodule"))[qi::_val = phoenix::bind(&PrismParser::createModule, phoenix::ref(*this), qi::_1, qi::_a, qi::_b, qi::_2, qi::_r1)];
@@ -189,6 +189,10 @@ namespace storm {
         }
         
         void PrismParser::moveToSecondRun() {
+            // In the second run, we actually need to parse the commands instead of just skipping them,
+            // so we adapt the rule for parsing commands.
+            commandDefinition = (qi::lit("[") > -(identifier[qi::_a = qi::_1]) > qi::lit("]") > expressionParser > qi::lit("->") > updateListDefinition(qi::_r1) > qi::lit(";"))[qi::_val = phoenix::bind(&PrismParser::createCommand, phoenix::ref(*this), qi::_a, qi::_2, qi::_3, qi::_r1)];
+            
             this->secondRun = true;
             this->expressionParser.setIdentifierMapping(&this->identifiers_);
         }