Unicode NBSP(u00A0) No-Break Space character for 64-bit

[sp0987 0]

Hi,

 

I am trying to create a .JSON file in Delphi 11 64-bit. While writing sample data to .json file , the &nbsp character was weird in 64-bit.

In 32-bit version , it was like "u00A0" and when coming to 64-bit version it was like "u=<68" and sometimes "u3<78". 

When I parse this resulted json file in .JS files , it throws an error like "invalid JSON".  Attached are samples from 32-bit and 64-bit.

 

Please suggest.

32-bit.json

64-bit.json

[Lajos Juhász 270]

A sample application that we can compile as a 64 and 32 bit would help us to help you. We can only suggest that you make improvements in your code.
 

[Virgo 18]

Are added samples supposed to be valid json? Because they both are missing end. Which raises a question, how the json was generated.

[sp0987 0]

It was just a part of json file, not the whole thing.

[Die Holländer 33]

U+1618 was added in Unicode version 3.0 in 1999.

It belongs to the block  Unified Canadian Aboriginal Syllabics in the  Basic Multilingual Plane.

The word that U+1618 forms with similar adjacent characters prevents a line break inside it.

 

>"filled with\u1618MVR\u1618information"

[sp0987 0]

Am having issue with "u3<78" . Instead of "u1618", am getting "u3<78" in some areas. When we see the 32-bit , it was 'u00A0' all over. 

[Die Holländer 33]

1 hour ago, Lajos Juhász said:

A sample application that we can compile as a 64 and 32 bit would help us to help you. We can only suggest that you make improvements in your code.
 

 

[sp0987 0]

Attaching a sample project. The project will create a json file, which when we try to use in javascript reults in error "invalid Json".

 

can be run in bot 32-bit and 64-bit with few changes. The changes were mentioned in the respected .pas files as comments. 

 

TestJson.7z

Edited June 25 by sp0987

[Stefan Glienke 1974]

Your assembly code of Int2Hex for 64bit is wrong - c is passed in RCX.

[sp0987 0]

Am sorry i didn't get you

[Brian Evans 100]

9 hours ago, sp0987 said:

Am sorry i didn't get you

In 64bit RAX does not contain the first parameter of your Int2Hex function, RCX does.  Looks like there are also other problems with the 64 bit ASM code like where the return value is stored: it should end up in RAX, 

 

Ref: Using Inline Assembly Code - RAD Studio (embarcadero.com)

Quote

The first four parameters to inline assembler functions are passed via RCX, RDX, R8, and R9 respectively, except for floating-point arguments which use XMMO, XMM1, XMM2, XMM3. The math coprocessor is not normally used from x64 code. Registers used for function parameters can be modified freely.

Ref: Assembly Procedures and Functions - RAD Studio (embarcadero.com)

Quote

64-bit

• Values 8 bytes or less in size are return in RAX.
• Real values are returned in XMM0.
• Other types are returned by a reference whose pointer value resides in RAX whose memory is allocated by the calling routine.

 

Edited June 26 by Brian Evans

[Stefan Glienke 1974]

FWIW the assembly implementation of that function is pointless given you can do it in a way more readable way - also performance cannot be a reason given the following call to mem_write causes a temporary heap allocation to convert your hex_code variable into an AnsiString.

 

Quickly slapped together (anyone who wants to further optimize this - be my guest, i cba right now):

type
  hex_code = Array [1 .. 4] of AnsiChar;

function Int2Hex(c: Word): hex_code;
const
  HexChars: array[0..15] of AnsiChar = '0123456789ABCDEF';
var
  i: NativeUInt;
begin
  i := c;
  Result[4] := HexChars[i and $0F];
  i := i shr 4;
  Result[3] := HexChars[i and $0F];
  i := i shr 4;
  Result[2] := HexChars[i and $0F];
  i := i shr 4;
  Result[1] := HexChars[i and $0F];
end;

and then instead of mem_write with AnsiString as the first argument use one that writes 4 bytes at once.

[Attila Kovacs 625]

Quote

CBA is an acronym that means can't be arsed, meaning, essentially, that a person can't be bothered to find the energy or willingness to do something. It's used in England, Australia, and New Zealand more than it is in the US. Arse is a British slang version of ass.

 

[Stefan Glienke 1974]

Can't be asked - I am using the sfw (safe for work) version 😇

[Die Holländer 33]

if S[I] in [WideChar(' ') .. WideChar('~')] Then
  mem_char(AnsiChar(S[I]), Buf)
else
Begin
  mem_char('\', Buf);
  mem_char('u', Buf);
  mem_write(IntToHex(ord(S[I]),4), Buf);
end;