Pascal decoder?

[Piru]

Excuse me, but this is a MS-DOS program.

As this is amiga.org I thought the program would have been compiled to AmigaOS binary... I'm afraid I can't do much about the DOS app. Sorry.

[Karlos]

Just how do you say "D'Oh!" in Finnish?

[ricco32097]

Anyone else want to give it a try?

[Karlos]

ricco32097 wrote:
Anyone else want to give it a try?

Perhaps someone over on the AROS sites would be better to ask? There are probably (just guessing) more x86 coders there?

[Minuous]

Then just use an MS-DOS decompiler.

Of course you will end up with C source rather than Pascal source, but that is actually an improvement :-)

[Karlos]

Arent there fundamental differences in the handling of variables etc in Pascal? I thought strings were represented differently than they are in Cfor example?

[Piru]

@Minuous

Of course you will end up with C source rather than Pascal source

Uh?

[Karlos]

Piru wrote:
@Minuous

Of course you will end up with C source rather than Pascal source

Uh?

I was thinking it would be an x86 dissassembly, myself ;-)

[Piru]

Well, I know there are apps that try to convert binaries to C code, but I doubt they produce any more readable code than x86 disassembly, anyway. :-)

[Karlos]

I suppose it depends how stripped the binaries are, I guess. I also expect OS calling mechanisms can be spotted and properly labelled.

As for the rest...

Try decompiling a duffs device loop to see what you get :-D

[Minuous]

If you use a disassembler, you'll get x86 source.

But if you use a decompiler you'll get source in a high-level language.

See http://en.wikipedia.org/wiki/Decompiler

Decompiler and disassembler aren't the same. Disassembler stops when it has created assembly code. Decompiler keeps processing and ends up with high-level language source. Of course there are many disassemblers on the Amiga but no decompilers AFAIK.

Most of the decompilers I've heard about produce C output, but in theory you could make one that produced Pascal output.

[Karlos]

I doubt a C decompiler would work well with code not written in C to start with.

There are certain programming constructs you can engineer in assembler that have absolutely no equivalent in C.

I wrote a small emulation core which was optimised in assembler, where each operation calculated the jump to the next code block directly and went there. A bit like having a big switch/case where the switch itself is never executed, but the code in each case clause is able to work out which one to jump to next relative to a base address (each block is padded to a certain 2^n size). There's absolutely no way to represent that in C.

So, code produced by pascal, could in theory contain machine level constructs that have no equivalent in C (not likely given their similarities).

[Piru]

Oh well, I still believe decompiler output is far from usable (at least the outputs I've seen, though I must admit it was years ago. Technology has gone forward since then I suppose).

[Minuous]

Yeah, they are far from perfect, but they automate some of the process: less hand-editing required to convert from assembler to C than if you did the whole thing by hand.

I wrote a program (Aminet:dev/cross/Annotate.lha) that comments Signetics 2650 assembly source with pseudo-C equivalents...a similar principle but not exactly a decompiler in the strictest sense.

[ricco32097]

File send...good luck, and thanks.

[TheMagicM]

decompile it with a AmigaBASIC like decompiler.. LOL.. get a couple of BASIC lines of code then a whole bunch of DATA statements... that should keep you busy. LOL