[Answered by wp] Saved csf file codepage problem
lazarus / 49-FPSpreadsheet

----------------------------------------------------------------------
totya · 2018-03-18 20:24

Hi!

When I save WorksheetGrid to csv file, the produced csv file codepage
is wrong, because the exel (2007) opened incorrectly.
It seems to mee, the produced csv file codepage is "UTF8 without boom"
but excel need as I see: "UTF8 with boom". Thank you!

Code: Pascal  [Select][+][-]window.onload = function(){var x1 =
document.getElementById("main_content_section"); if (x1) { var x =
document.getElementsByClassName("geshi");for (var i = 0; i unit
Unit1; {$mode objfpc}{$H+} interface uses  Classes, SysUtils,
FileUtil, Forms, Controls, Graphics, Dialogs, 
fpspreadsheetgrid; type   { TForm1 }   TForm1 = class(TForm)   
sWorksheetGrid1: TsWorksheetGrid;    procedure FormCreate(Sender:
TObject);  private    { private declarations }  public    { public
declarations }  end; var  Form1: TForm1; implementation {$R *.lfm} {
TForm1 } procedure TForm1.FormCreate(Sender: TObject);begin  with
sWorksheetGrid1 do  begin   
sWorksheetGrid1.Worksheet.WriteText(0,0,'öüóőúéáűíÖÜÓŐÚÉÁŰÍ');   
sWorksheetGrid1.SaveToSpreadsheetFile('test.csv');  end;end; end. 
Edit.: But it seems to me, save csv file with ANSI codepage is better,
because if I open "; tagged" (csv file) utf8 (with boom) file in
Excel, then I save again, the result: ";" changed to tabulator. LOL.

----------------------------------------------------------------------
wp · 2018-03-18 21:36

Since csv has so many "undefined" options fpspreadsheet declares a
global record CSVParams in unit fpscsv in which you can define
specific parameters:
Code: Pascal  [Select][+][-]window.onload = function(){var x1 =
document.getElementById("main_content_section"); if (x1) { var x =
document.getElementsByClassName("geshi");for (var i = 0; i type 
TsCSVParams = record   // W = writing, R = reading, RW =
reading/writing    SheetIndex: Integer;             // W: Index of the
sheet to be written    LineEnding: TsCSVLineEnding;     // W:
Specification for line ending to be written    Delimiter: Char;      
         // RW: Column delimiter    QuoteChar: Char;                
// RW: Character for quoting texts    Encoding: String;              
 // RW: Encoding of file (code page, such as "utf8", "cp1252" etc)   
DetectContentType: Boolean;      // R: try to convert strings to
content types    NumberFormat: String;            // W: if empty write
numbers like in sheet, otherwise use this format   
AutoDetectNumberFormat: Boolean; // R: automatically detects
decimal/thousand separator used in numbers    TrueText: String;      
        // RW: String for boolean TRUE    FalseText: String;         
    // RW: String for boolean FALSE    FormatSettings:
TFormatSettings; // RW: add'l parameters for conversion  end;   var 
CSVParams: TsCSVParams = (    SheetIndex: 0;    LineEnding: leSystem; 
  Delimiter: ';';    QuoteChar: '"';    Encoding: '';    // '' = auto-
detect when reading, UTF8 when writing    DetectContentType: true;   
NumberFormat: '';    AutoDetectNumberFormat: true;    TrueText:
'TRUE';    FalseText: 'FALSE';  ); 
Therefore, you can fine-tune the encoding of the file you just have to
adjust CSVParams.Encoding accordingly before writing. So, if Excel
wants "UTF8 with BOM", set CSVParams.Encoding := 'utf8bom', or set
CSVParams.Encoding := 'cp1252' to write an ANSI file in the central-
European codepage.

See also: http://wiki.lazarus.freepascal.org/FPSpreadsheet#Reading_and
_writing_of_CSV_files

----------------------------------------------------------------------
totya · 2018-03-18 22:04

Hi!

Very thank you for answer!

I can try only tomorrow, but I suggest, change the default csv
codepage from utf8-without-boom to utf8-with-boom. Why? Because I try
open this csv file open from LibreOffice (v5.2.2.2) too, and codepage-
detect is bad again. With utf8 with boom, the libre office autodetect
is OK, like as Excel 2007.

Thank you!

----------------------------------------------------------------------
wp · 2018-03-18 23:11

At first I thought: Good idea, and uploaded a new version with utf8bom
as default encoding for CSV files. But then I changed my mind and
reverted this: fpspreadsheet is a library spreadsheet files, so if you
want to create data for Excel or Calc you would of course write in
their native format, not in the crooked CSV. On the other hand, if you
write CSV files it is most probably for other programs, and I estimate
that most of them will not expect to have a BOM at the beginning of a
text file. Delphi 7, for example, chokes if you load a UTF8-BOM source
file.

Luckily the situation is very flexible. If your program is supposed to
write a CSV for Excel/Calc just change CSVParams.Encoding to
'utf8bom'.

----------------------------------------------------------------------
totya · 2018-03-18 23:13

Hi!

I see, you changed the default codepage, and later you changed back...
but no problem, these lines:

Code: Pascal  [Select][+][-]window.onload = function(){var x1 =
document.getElementById("main_content_section"); if (x1) { var x =
document.getElementsByClassName("geshi");for (var i = 0; i uses
...fpscsv... 
and

Code: Pascal  [Select][+][-]window.onload = function(){var x1 =
document.getElementById("main_content_section"); if (x1) { var x =
document.getElementsByClassName("geshi");for (var i = 0; i
CSVParams.Encoding:= 'cp1250'; 
... solve my problem.

And yes, my need cp1250 codepage, and not cp1252. With cp1252 few
chars missing, see demo app in the first message.

Thank you for all, master!

ps.: Why I need csv file? Because I can compare excel files easily...