CRC8/Maxim function?

[avk]

In my previous post, I tried to draw your attention to the fact that all the functions mentioned above expect a sequence of bytes as input, and not their hexadecimal codes.
And by the way, have you tried passing the test string '123456789' to your instance of the CRC function? What result does it return?

[MarkMLl]

There's also a possibility that, since it's apparently not FPC, the compiler is making different size assumptions etc. e.g. that cardinals are 8-bit unsigned.

MarkMLl

[JCDes]

It is calculating as if the data was a string not bytes. "0000000500000000000000000000" calculates to "03" which is perfect for string input.
How can I get it to interpret the input as bytes $00 $00 $00 $05 $00 $00 $00 $00 $00 $00 $00 $00 $00 $00 to get "AA" result??

[lucamar]

A quick way would be to add a first step to convert the input string to another one with each pair of digits from the original converted to the corresponding character, e.g. with something like this (untested) function:

Code: Pascal  [Select][+][-]

1. function Recode(const Origin: String): String;
2. var
3.   sPos: Integer = 1;
4. begin
5.   Result := '';
6.   while sPos < Length(Origin) do begin
7.     Result := Result + Chr(StrToInt('$'+ Original.SubString(sPos,2)));
8.     Inc(sPos, 2)
9.   end;
10. end;

[JCDes]

Error: Object or record required to access method "Substring". on this line: Result := Result + Chr(StrToInt('$'+ Origin.SubString(sPos,2)));

[lucamar]

Error: Object or record required to access method "Substring". on this line: Result := Result + Chr(StrToInt('$'+ Origin.SubString(sPos,2)));

Do you have SysUtils in your uses clause? SubString() is a method of TStringHelper, the string type helper defined there.

If for some reason you don't want to use it you can use the standard Copy() function; it's basically the same:

Code: Pascal  [Select][+][-]

1. Result := Result + Chr(StrToInt('$'+ Copy(Origin, sPos, 2)));

[JCDes]

Expected 1 parameter error on line: Inc(sPos, 2)

[JCDes]

This is the code that is calculating the data as a string correctly:

Code: [Select]

function Crc8Maxim(Buffer: string):Cardinal;
var
  CrcTable: array of Cardinal;
  I: LongInt;
begin
  CrcTable := [
    $00, $5e, $bc, $e2, $61, $3f, $dd, $83, $c2, $9c, $7e, $20, $a3, $fd, $1f, $41,
    $9d, $c3, $21, $7f, $fc, $a2, $40, $1e, $5f, $01, $e3, $bd, $3e, $60, $82, $dc,
    $23, $7d, $9f, $c1, $42, $1c, $fe, $a0, $e1, $bf, $5d, $03, $80, $de, $3c, $62,
    $be, $e0, $02, $5c, $df, $81, $63, $3d, $7c, $22, $c0, $9e, $1d, $43, $a1, $ff,
    $46, $18, $fa, $a4, $27, $79, $9b, $c5, $84, $da, $38, $66, $e5, $bb, $59, $07,
    $db, $85, $67, $39, $ba, $e4, $06, $58, $19, $47, $a5, $fb, $78, $26, $c4, $9a,
    $65, $3b, $d9, $87, $04, $5a, $b8, $e6, $a7, $f9, $1b, $45, $c6, $98, $7a, $24,
    $f8, $a6, $44, $1a, $99, $c7, $25, $7b, $3a, $64, $86, $d8, $5b, $05, $e7, $b9,
    $8c, $d2, $30, $6e, $ed, $b3, $51, $0f, $4e, $10, $f2, $ac, $2f, $71, $93, $cd,
    $11, $4f, $ad, $f3, $70, $2e, $cc, $92, $d3, $8d, $6f, $31, $b2, $ec, $0e, $50,
    $af, $f1, $13, $4d, $ce, $90, $72, $2c, $6d, $33, $d1, $8f, $0c, $52, $b0, $ee,
    $32, $6c, $8e, $d0, $53, $0d, $ef, $b1, $f0, $ae, $4c, $12, $91, $cf, $2d, $73,
    $ca, $94, $76, $28, $ab, $f5, $17, $49, $08, $56, $b4, $ea, $69, $37, $d5, $8b,
    $57, $09, $eb, $b5, $36, $68, $8a, $d4, $95, $cb, $29, $77, $f4, $aa, $48, $16,
    $e9, $b7, $55, $0b, $88, $d6, $34, $6a, $2b, $75, $97, $c9, $4a, $14, $f6, $a8,
    $74, $2a, $c8, $96, $15, $4b, $a9, $f7, $b6, $e8, $0a, $54, $d7, $89, $6b, $35];

  Result := 0;
  for I := 1 to Length(Buffer) do
    Result := CrcTable[Result xor Ord(Buffer[I])];
end;

This is how I'm reading the data from the hex editor:

Code: [Select]

CS := IntToHex(GetByteHexEdit( $00),2) + IntToHex(GetByteHexEdit( $01),2) + IntToHex(GetByteHexEdit( $02),2)
+ IntToHex(GetByteHexEdit( $03),2) + IntToHex(GetByteHexEdit( $04),2) + IntToHex(GetByteHexEdit( $05),2)
+ IntToHex(GetByteHexEdit( $06),2) + IntToHex(GetByteHexEdit( $07),2) + IntToHex(GetByteHexEdit( $08),2)
+ IntToHex(GetByteHexEdit( $09),2) + IntToHex(GetByteHexEdit( $0A),2) + IntToHex(GetByteHexEdit( $0B),2)
+ IntToHex(GetByteHexEdit( $0C),2) + IntToHex(GetByteHexEdit( $0D),2);

Ir reads the data I need and gets processed but it gets processed as a string not as bytes. If I go to http://www.sunshine2k.de/coding/javascript/crc/crc_js.html and set to crc-maxim in the predifined section and select "string" I get the same result but the result I need is the one for the "bytes" setting.

"0000000500000000000000000000"
I get "03" but I should be getting "AA"

[lucamar]

Expected 1 parameter error on line: Inc(sPos, 2)

A wild guess: Some other unit in your uses is hijacking the one declared in System. Replace with:

Code: Pascal  [Select][+][-]

1. System.Inc(sPos, 2)

[JCDes]

Error, it took "system" as an undeclared variable. This tms scripter is killing me. I really appreciate you guys  helping me even if I'm not using fpc. I'm using tms scripter and it seems like the pascal syntax is limited. This is all the documentation that comes with it. https://doc.tmssoftware.com/biz/scripter/scripter-user-guide.pdf

[lucamar]

This tms scripter is killing me.

Oops! Is it an script? No wonder it barked at the type helper, then! Sorry, didn't read enough of the thread ...

If all else is working then try replacing the single Inc(sPos, 2) with either:

Code: Pascal  [Select][+][-]

1. Inc(sPos);
2. Inc(sPos);

or with:

Code: Pascal  [Select][+][-]

1. sPos := sPos + 2;

Either of them should work, according to the documentation and the error message.

ETA: I don't want to sound snarky but ... rather than copy-pasting you should try to understand what our code is trying to do so you can replace it with the appropiate one for TMS Script. In this particular case it should have been clear that it's increasing sPos by 2, so the solution(s) should have been obvious to you ... Just saying

[avk]

Hmm, what if something like that?

Code: Pascal  [Select][+][-]

1.  CS := Char(GetByteHexEdit( 0)) + Char(GetByteHexEdit( 1)) + Char(GetByteHexEdit( 2)) +
2.        Char(GetByteHexEdit( 3)) + Char(GetByteHexEdit( 4)) + Char(GetByteHexEdit( 5)) +
3.        Char(GetByteHexEdit( 6)) + Char(GetByteHexEdit( 7)) + Char(GetByteHexEdit( 8)) +
4.        Char(GetByteHexEdit( 9)) + Char(GetByteHexEdit(10)) + Char(GetByteHexEdit(11)) +
5.        Char(GetByteHexEdit(12)) + Char(GetByteHexEdit(13));
6.  

That is, you can try to get a byte sequence at once and do without further conversions.

[Thaddy]

Ahum, https://www.freepascal.org/docs-html/rtl/sysutils/tbytes.html
Also note the "see also"

[JCDes]

I really appreciate your help. The problem is that this TMS Scripter that I'm using errors out on:

 Const CrcTable:Array[0..255] of Cardinal = <------- Syntax error
inc(I,2) <----- This one says expected 1 parameter only

it allows

var  CRCTable:array;
begin
crctable := [the table here];

[MarkMLl]

I really appreciate your help. The problem is that this TMS Scripter that I'm using errors out on:

 Const CrcTable:Array[0..255] of Cardinal = <------- Syntax error
inc(I,2) <----- This one says expected 1 parameter only

it allows

var  CRCTable:array;
begin
crctable := [the table here];

Isn't that table generated by a function you gave in your first posting? Use that to generate the CRC table as a variable (which will be standard 1970s-style Pascal) rather than using language extensions to predefine it as a constant.

MarkMLl