Conversion to UnicodeString

[jcmontherock]

I have the following instructions:

Code: Pascal  [Select][+][-]

1. aStr  := '  -- --------------------------------------------------------      ' + sLineBreak;
2. aStr += '  -- Hôte:                         127.0.0.1                       ' + sLineBreak;

aStr is defined as an AnsiString.

For only the 2nd line I got the following warning messages:
Warning: Implicit string type conversion from "AnsiString" to "UnicodeString"
Warning: Implicit string type conversion with potential data loss from "UnicodeString" to "AnsiString"

I do not understand why the compiler want to convert the second line to UnicodeString and then to AnsiString. Caused by the concatenation ?

[Seenkao]

Code: Pascal  [Select][+][-]

1. aStr = aStr + '  -- Hôte:                         127.0.0.1                       ' + sLineBreak;

same errors?

[tetrastes]

I suppose you use {$codepage utf8} directive or -FcUTF8 option. In this case compiler converts string constants with non-ASCII characters to UnicodeString (UTF-16). See produced asm:

Code: ASM  [Select][+][-]

1. # Begin asmlist al_typedconsts
2.  
3. .section .rodata.n_.Ld1,"d"
4.         .balign 8
5. .Ld1$strlab:
6.         .short  0,1
7.         .long   0
8.         .quad   -1,69
9. .Ld1:
10.         .ascii  "  -- ----------------------------------------------"
11.         .ascii  "----------      \015\012\000"
12.  
13. .section .rodata.n_.Ld2,"d"
14.         .balign 8
15. .Ld2$strlab:
16.         .short  1200,2
17.         .long   0
18.         .quad   -1,69
19. .Ld2:
20.         .short  32,32,45,45,32,72,244,116,101,58,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32
21.         .short  32,49,50,55,46,48,46,48,46,49,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,13,10,0
22. # End asmlist al_typedconsts
23.  

Tell compiler that this string constant is AnsiString explicitly:

Code: Pascal  [Select][+][-]

1.   aStr += AnsiString('  -- Hôte:                         127.0.0.1                       ') + sLineBreak;
2.  

Asm:

Code: ASM  [Select][+][-]

1. # Begin asmlist al_typedconsts
2.  
3. .section .rodata.n_.Ld1,"d"
4.         .balign 8
5. .Ld1$strlab:
6.         .short  0,1
7.         .long   0
8.         .quad   -1,69
9. .Ld1:
10.         .ascii  "  -- ----------------------------------------------"
11.         .ascii  "----------      \015\012\000"
12.  
13. .section .rodata.n_.Ld2,"d"
14.         .balign 8
15. .Ld2$strlab:
16.         .short  65001,1
17.         .long   0
18.         .quad   -1,70
19. .Ld2:
20.         .ascii  "  -- H\303\264te:                         127.0.0.1"
21.         .ascii  "                       \015\012\000"
22. # End asmlist al_typedconsts
23.  

[jcmontherock]

@seenkao: yes same messages.

@tetrastes: Yes, I use {$codepage utf8}. If I comment this directive all the warning messages disappear and the cmd program works, as well, fine. I will find the utility of that command in programs, where text is in utf-8 encoding in both cases.

Thanks a lot both of you.