In order to construct tt(mfcalc), the following steps are suggested: itemization( it() Construct a program tt(mfcalc.cc). Actually, it is already available, since all implementations of tt(main()) used so far are identical to each other. it() Construct the parser in a subdirectory tt(parser): itemization( it() First, construct b()'s input file as indicated above. Name this file tt(grammar); it() Run tt(bisonc++ grammar) to produce the files tt(Parser.h), tt(Parserbase.h), tt(Parser.ih) and tt(parse.cc); it() Modify tt(Parser.h) so as to include tt(FunctionPair, s_functions, s_funTab) and tt(d_symbols); it() Modify tt(Parser.ih) so as to include tt(cmath) and optionally `tt(using namespace std)', which is commented out by default; it() Implement tt(data.cc) and tt(lex.cc) to initialize the static data and to contain the lexical scanner, respectively. ) it() Now construct tt(mfcalc) in tt(mfcalc.cc)'s directory using the following command: verb( g++ -o mfcalc *.cc parser/*.cc ) )