[SOLVED] Operators overriding and use of PChars in code for AVRs

[ccrause]

2. What is your opinion about PChars (=arrays[0..n] of char) in code for AVRs? Is this feature fully supported by the compiler? In the wiki page https://wiki.freepascal.org/AVR_Programming there is no mention about PChars, but I think it was written at  one point that PChars work. If yes - what is better to use in practice: strings[n] or arrays[0..n] of char?

As Thaddy and other mentioned, you can use either style, but shortstrings is more convenient to work with. I would avoid ansistrings unless you also have other dynamic data allocation requirements since you have to reserve a block of SRAM for the memory manager to work with.

[ackarwow]

I give up. This is agnostic to processors..

I have problem with compile my code for operator * for TFloat32 even on Windows. Maybe I did some mistake?
TFloat32=UInt32.

[d.ioannidis]

Hi,

<snip>

Code: Pascal  [Select][+][-]

1. uses
2.   float32;
3.  
4. {$mode objfpc}
5. operator * (const f1, f2: TFloat32): TFloat32; <-- HERE
6. begin
7.   Result:=Float32Mul(f1, f2);
8. end;
9.  

and FPC throws an error:

Code: Text  [Select][+][-]

1. Target OS: Embedded
2. Compiling C:\Programs\avr\avrpascal\examples\Operators.pas
3. Operators.pas(12,46) Error: Impossible operator overload
4. Operators.pas(17,11) Fatal: Syntax error, "identifier" expected but "OPERATOR" found
5. Fatal: Compilation aborted
6.  

What type is the TFloat32 ? If it's a record then it should work .

If it's a class, object I don't think that its supported . AFAIK, Only the enumerator operator supports classes .

EDIT: I wrote my reply while you posted :

A long shot but maybe this works ?

Code: Pascal  [Select][+][-]

1. operator * (const f1, f2: UInt32): UInt32;

regards,

[ackarwow]

2. What is your opinion about PChars (=arrays[0..n] of char) in code for AVRs? Is this feature fully supported by the compiler? In the wiki page https://wiki.freepascal.org/AVR_Programming there is no mention about PChars, but I think it was written at  one point that PChars work. If yes - what is better to use in practice: strings[n] or arrays[0..n] of char?

As Thaddy and other mentioned, you can use either style, but shortstrings is more convenient to work with. I would avoid ansistrings unless you also have other dynamic data allocation requirements since you have to reserve a block of SRAM for the memory manager to work with.

@ccrause
Thank you for comment. What about passing parameters of string type (=shortstring=255 bytes?) into procedures/functions. Isn't it better to use PChar (of any size defined earlier)?

[ackarwow]

(...)
What type is the TFloat32 ? If it's a record then it should work .

If it's a class, object I don't think that its supported . AFAIK, Only the enumerator operator supports classes .

regards,

TFloat32=UInt32;

[ackarwow]

(...)

Code: Pascal  [Select][+][-]

1. operator * (const f1, f2: UInt32): UInt32;

Unfortunately, this doesn't work either. I am testing on simple unit added to project... Maybe in wrong way... 

[d.ioannidis]

Hi,

I give up. This is agnostic to processors..

I have problem with compile my code for operator * for TFloat32 even on Windows. Maybe I did some mistake?
TFloat32=UInt32.

it doesn't work for me either . On Windows for x86 and for AVR4 ...

Maybe a bug ?

regards,

[Thaddy]

The bug is laziness in the programmer.I ff'ing showed you.

[ccrause]

Lets start with a simple example:

Code: Pascal  [Select][+][-]

1. program test;
2.  
3. type
4.   TComplex = record
5.     r, i: int8;
6.   end;
7.  
8. operator + (const a, b: TComplex): TComplex;
9. begin
10.   Result.r := a.r + b.r;
11.   Result.i := a.i + b.i;
12. end;
13.  
14. var
15.   x1: TComplex = (r: 1; i: 2);
16.   x2: TComplex = (r: 1; i: -4);
17.   y: TComplex;
18.  
19. begin
20.   y := x1 + x2;
21.   writeln('y.r = ', y.r, ' (expected: 2)');
22.   writeln('y.i = ', y.i, ' (expected: -2)');
23. end.

Compile (for dramatic effect I compile fro the command line):

Code: Text  [Select][+][-]

1. $ ~/fpc/installs/lib/fpc/3.3.1/ppcrossavr -Tembedded -Wpavrsim -Mobjfpc -viwn test.pp
2. Target OS: Embedded
3. Compiling test.pp
4. Assembling test
5. Linking test
6. 24 lines compiled, 0.2 sec, 5324 bytes code, 8593 bytes data

Now test generated bin in fp-avrsim (no need to wear out flash to do this trivial test):

Code: Text  [Select][+][-]

1. $ ~/LazProjs/fp-avrsim-cc/avrsim test.bin
2. y.r = 2 (expected: 2)
3. y.i = -2 (expected: -2)

[d.ioannidis]

Hi,

The bug is laziness in the programmer.I ff'ing showed you.

Oh I agree .... Did you tried now ?

Lets start with a simple example:
<snip>

I'm sorry but I never used operator overloading . Can you show a multiplication example with simple types ?

regards,

[ackarwow]

Lets start with a simple example:

Code: Pascal  [Select][+][-]

1. program test;
2.  
3. type
4.   TComplex = record
5.     r, i: int8;
6.   end;
7.  
8. operator + (const a, b: TComplex): TComplex;
9. begin
10.   Result.r := a.r + b.r;
11.   Result.i := a.i + b.i;
12. end;
13.  
14. var
15.   x1: TComplex = (r: 1; i: 2);
16.   x2: TComplex = (r: 1; i: -4);
17.   y: TComplex;
18.  
19. begin
20.   y := x1 + x2;
21.   writeln('y.r = ', y.r, ' (expected: 2)');
22.   writeln('y.i = ', y.i, ' (expected: -2)');
23. end.

Compile (for dramatic effect I compile fro the command line):

Code: Text  [Select][+][-]

1. $ ~/fpc/installs/lib/fpc/3.3.1/ppcrossavr -Tembedded -Wpavrsim -Mobjfpc -viwn test.pp
2. Target OS: Embedded
3. Compiling test.pp
4. Assembling test
5. Linking test
6. 24 lines compiled, 0.2 sec, 5324 bytes code, 8593 bytes data

Now test generated bin in fp-avrsim (no need to wear out flash to do this trivial test):

Code: Text  [Select][+][-]

1. $ ~/LazProjs/fp-avrsim-cc/avrsim test.bin
2. y.r = 2 (expected: 2)
3. y.i = -2 (expected: -2)

@ccrause
Yes, thank you. This code is compiling without problems.
Let's try with "more complex" one

Code: Pascal  [Select][+][-]

1. program Operators;
2.  
3. {$mode objfpc}
4.  
5. type
6.   TFloat32=UInt32;
7.  
8. operator * (const f1, f2: TFloat32): TFloat32;
9. begin
10.   Result:=f1*f2; //only for test
11. end;
12.  
13. begin
14.   //nothing to do
15. end.
16.  

Code: Text  [Select][+][-]

1. Target OS: Embedded
2. Compiling C:\Programs\avr\avrpascal\examples\Operators.pas
3. Operators.pas(8,46) Error: Impossible operator overload
4. Operators.pas(17) Fatal: There were 1 errors compiling module, stopping
5. Fatal: Compilation aborted
6.  

Where is my mistake? 

[TRon]

Where is my mistake? 

https://www.freepascal.org/docs-html/ref/refse102.html#x222-24600015.1

Free Pascal supports operator overloading. This means that it is possible to define the action of some operators on self-defined types, and thus allow the use of these types in mathematical expressions.

[Thaddy]

Tnx, Ron, saves me time to respond.

[ackarwow]

Where is my mistake? 

https://www.freepascal.org/docs-html/ref/refse102.html#x222-24600015.1

Free Pascal supports operator overloading. This means that it is possible to define the action of some operators on self-defined types, and thus allow the use of these types in mathematical expressions.

@Tron thank you, that really helped. But that's not very good information.

This code works:

Code: Pascal  [Select][+][-]

1. program Operators;
2.  
3. {$mode objfpc}
4.  
5. type
6.   TFloat32=array[0..3] of UInt8;// UInt32;
7.  
8. operator * (const f1, f2: TFloat32): TFloat32;
9. begin
10.   Result:=f1*f2; //only for test
11. end;
12.  
13. begin
14.   //nothing to do
15. end.
16.  

But array in this case is not very comfortable construction...

[d.ioannidis]

Hi,

Where is my mistake? 

https://www.freepascal.org/docs-html/ref/refse102.html#x222-24600015.1

Free Pascal supports operator overloading. This means that it is possible to define the action of some operators on self-defined types, and thus allow the use of these types in mathematical expressions.

OK got it  .. so RTFM error  and apologies 

Sorry for the noise ...

Regards,