Type mismatch - Inno Setup Unicode 5.5

[fiazhnd]

I face error "Type Mismatch' in Inno Setup Unicode 5.5.9 and the same code I use in Delphi work well.

Code: Pascal  [Select][+][-]

1.  
2. function GetHardwareId(HDD, NIC, CPU, BIOS: LongBool; lpHWID: PAnsiChar; nMaxCount: Integer): Integer; external 'GetHardwareId@files:hwid.dll stdcall setuponly';
3.  
4. function GetHWID(): AnsiChar;
5. var
6.   sHWID: array[0..254] of AnsiChar;
7.   iRtn : Integer;
8.   HDD, NIC, CPU, BIOS: LongBool;
9. begin
10.  
11.   SetAppName('');
12.  
13.   HDD := True;
14.    NIC := False;
15.     CPU := True;
16.      BIOS := True;
17.  
18.   iRtn:= GetHardwareId(HDD, NIC, CPU, BIOS, sHWID, SizeOf(sHWID)); // <= Error
19.   if iRtn > 0 then
20.   Edit0.Text := sHWID;
21. end;

[Thaddy]

That's not nearly enough information. Are you sure you tried this in {$mode delphi} or {$mode delphiunicode} and platform Windows?

I face error "Type Mismatch'

that's not all information: the error has the full type mismatch line: what type it expects and what type you try to use.
That is essential information. We need that to help you.
Your code is also very badly formatted.

My current guesswork ends up with:

Code: Pascal  [Select][+][-]

1.   sHWID: array[0..254] of AnsiChar;

should be

Code: Pascal  [Select][+][-]

1.   sHWID: array[0..255] of AnsiChar;

and

Code: Pascal  [Select][+][-]

1. SizeOf(sHWID))

should be

Code: Pascal  [Select][+][-]

1. Strlen(sHWID)

It is a complete mess....maybe this will help:

Code: Pascal  [Select][+][-]

1. {$ifdef fpc}{$mode delphi}{$H+}{$endif}
2. var s: array [0..255] of AnsiChar;
3. begin
4.  s := Pchar('testmewhat my length is?');
5.  writeln('Array length: ',SizeOf(s),' Array length: ', length(s),' Content length!: ' , strlen(s));
6. end.

See the difference?

[Zvoni]

uhm?
The 5th Argument of the Function expects a Pointer to AnsiChar, but he throws the Array at it?

i would have expected something like

iRtn:= GetHardwareId(HDD, NIC, CPU, BIOS, @sHWID[0], Length(sHWID));

[Josh]

Hi

Are you trying this in INNOSetup itself?
surely your protection should be in your app?

AFAIK If innosetup could use the wmi to get the information, the info would only be valid for the developer machine(when you compile the script) not the client machine. I could be wrong here though.

If its an innosetup specific, maybe posting on their groups would be better
http://www.jrsoftware.org/newsgroups.php

[Thaddy]

uhm?
The 5th Argument of the Function expects a Pointer to AnsiChar, but he throws the Array at it?

i would have expected something like

iRtn:= GetHardwareId(HDD, NIC, CPU, BIOS, @sHWID[0], Length(sHWID));

Well, I already showed why your code will fail too.... I gave a simple example, compiles and works and shows exactly what is wrong.
Like

Code: Pascal  [Select][+][-]

1.  Length(sHWID));// this is a mistake! just like SizeOf() is

Some people don't believe simple examples. It is very hard to write simple examples...

[Thaddy]

AFAIK If innosetup could use the wmi to get the information, the info would only be valid for the developer machine(when you compile the script) not the client machine. I could be wrong here though.

Using WMI is not an option and not needed (because Innosetup uses the underlying lowlevel calls anyway!!!) . But InnoSetup is windows platform only.
What he is doing is not giving enough information:
The compiler tells you what it expects and what it gets......

[Zvoni]

Thaddy, i said expected.

Since he's using an Array i would have expected "Length", but you're correct.
Since it's a Pointer to AnsiChar expected, strLen would be correct

EDIT: What actually baffles me: If he knows, that his external function expects a PAnsiChar, why is he using an Array??
Something like this would have been my first try (untested)

Code: Pascal  [Select][+][-]

1. Var
2.    sHWID:PAnsiChar;
3. .
4. .
5. .
6. sHWID:=GetMem(256);
7. iRtn:= GetHardwareId(HDD, NIC, CPU, BIOS, sHWID, StrLen(sHWID));
8. .
9. .
10. .
11. FreeMem(sHWID);
12.  

[jamie]

@sHWID[0]
should be enough on the original verison..

[Thaddy]

@sHWID[0]
should be enough on the original verison..

No it is not. The library is NOT written in Pascal. Pascal allows embedded #0's in a string. C and C++ (and more brackety languages)
You should always determine the length of such a type of string with compatible code such as strlen.

Comments such as yours introduce bugs, not solve them, for a Pascal programmer.
I suppose now you know the errors of your ways?

BTW even using an index my small example shows clearly that is just as wrong....

[jamie]

what ever..