Converting MINI-BASIC in MASM over to C++?

**Paul Panks** · 03-12-2009

Global variables are necessary, I think. I know they are hard to keep track of, but that is how programming can be sometimes. I don't know much C, so someone else would have to translate it to C.

Paul

**matsp** · 03-12-2009

Originally Posted by Paul Panks

Global variables are necessary, I think. I know they are hard to keep track of, but that is how programming can be sometimes. I don't know much C, so someone else would have to translate it to C.

Paul

Global variables are SOMETIMES necessary, but often used as a lazy excuse for not passing variables from one function to another.

Why do you want this translated to C?

--
Mats

**Paul Panks** · 03-12-2009

Because I think it would run better, with fewer Windows API errors. In C or C++, there would be a standard library which could use the Windows API for handling the keyboard, display and file I/O.

As it is now, MINI-BASIC is a fragile piece of software. It crashes on occasion with Windows API errors. It is a mystery to me why this occurs.

Paul

**matsp** · 03-12-2009

Originally Posted by Paul Panks

Because I think it would run better, with fewer Windows API errors. In C or C++, there would be a standard library which could use the Windows API for handling the keyboard, display and file I/O.

As it is now, MINI-BASIC is a fragile piece of software. It crashes on occasion with Windows API errors. It is a mystery to me why this occurs.

Paul

Right, I've done some work, and I can probably say that I wouldn't be surprised if the API errors you get are caused by global variables being changed when some other bit of code expects it to remain the same.

As to wheter C or C++ or Assembler is "more safe", I'd say all three languages are capable of doing the wrong thing, and badly written code in either of them will not work reliably.

--
Mats

**Paul Panks** · 03-12-2009

Yeah. That's true. I don't know how badly written MINI-BASIC is, but it does crash occasionally with Windows API errors. Which is unfortunate. Translating it to C or C++ shouldn't be impossible. It has a very limited subset of the traditional BASIC language. It has 26 scalar variables and 1 array -- denoted by @(I).

Paul

**matsp** · 03-12-2009

Sure. But it's not the easiest way to "translate" it.

I'm still working my way through it, and I'm just about to implement the begginnings of the parser.

Most of the code so far is following very closely the original code, but I will use C style strings in the parser, to make life a bit more C-friendly.

--
Mats

**Paul Panks** · 03-12-2009

Ok, thanks for the update.

Sounds good so far.

Paul

**leeor_net** · 03-12-2009

It probably comes down to shoddy programming -- that is usually the case with so-called fragile software. Fragile = badly written. Being written in ASM only compounds these problems when somebody is using the language without a complete and thorough understanding of internal machine code.

If you were to convert or 'port' it, I would suggest that you just rewrite it from scratch. ASM - C++ is, if you ask me, a waste of time.

Note, though, that C++ is only based on C... it is not C. (I know that's going to start an argument but I'm quoting Bjorn Stroustrup himself). You would need to choose one or the other. C++ is a lot easier when it comes to dealing with errors and the STL names it almost trivial to write an interpretter.

I would say, however, that if you're not familiar with C/C++ you should probably start with a project a little smaller than that.

**Paul Panks** · 03-12-2009

So is it possible to convert to C++?

Paul

**Paul Panks** · 03-18-2009

Originally Posted by leeor_net

It probably comes down to shoddy programming -- that is usually the case with so-called fragile software. Fragile = badly written. Being written in ASM only compounds these problems when somebody is using the language without a complete and thorough understanding of internal machine code.

If you were to convert or 'port' it, I would suggest that you just rewrite it from scratch. ASM - C++ is, if you ask me, a waste of time.

Note, though, that C++ is only based on C... it is not C. (I know that's going to start an argument but I'm quoting Bjorn Stroustrup himself). You would need to choose one or the other. C++ is a lot easier when it comes to dealing with errors and the STL names it almost trivial to write an interpretter.

I would say, however, that if you're not familiar with C/C++ you should probably start with a project a little smaller than that.

Assembly language is a different animal altogether, true...but I think porting it to C/C++ would benefit the C/C++ community. It is a useful language modeled after Palo Alto Tiny BASIC from 1976.

Paul

**anon** · 03-12-2009

It is possible to implement it from scratch (write a program that does the same thing functionally). I don't think there is any direct way to convert asm to C++ or C, since it is lacking all the higher level constructs. Assembly just contains less human readable information than the equivalent C or C++ code.

**Paul Panks** · 03-12-2009

Writing it from scratch in C would require examining how the program works first.. I believe anyway.

Paul

**leeor_net** · 03-12-2009

Originally Posted by Paul Panks

Writing it from scratch in C would require examining how the program works first.. I believe anyway.

Paul

True... or you could take a BASIC standard approach and write a new interpretter using BASIC syntax.

I think you're best bet is to just write a new basic interpretter or use an existing open-source interpretter.

It looks like you're getting a bit of help with it though so good luck!

**anon** · 03-12-2009

And converting Assembly to C would require you to study how the program works (what is input, what is output) as well as the implementation details of the assembly version. The latter you wouldn't have to do, since you'd just design that yourself any way you like.