Hello,
I am developing a console application which receives an UTF-8 encoded text through stdin.
The text is, after possible modification, sent to SAPI5. The SAPI5 generated interface wrappers have parameters of type string. In FPC 2.6.4, I used UTF8toANSI on various places and that worked well. Now after reading about FPC 3.x's unicode support, I thought that adding this line say after begin in the main program could be enough: SetMultiByteConversionCodePage(CP_UTF8);
It however doesn't work.
How can I solve this without conversion to ANSI?
Thanks
Mailing lists / fpc-pascal / 2015-11
FPC 3.x for Windows and code page conversion
5 messages · 4 people ·
- Lubos Pintes 2
- Jonas Maebe 1
- Mattias Gaertner 1
- Michael Van Canneyt 1
Reply by email Subscribe to this list Your subscription
Read-only archive. Use j and k to move between messages.
Show quoted text 9 lines
Hello,
I am developing a console application which receives an UTF-8 encoded
text through stdin.
The text is, after possible modification, sent to SAPI5. The SAPI5
generated interface wrappers have parameters of type string.
In FPC 2.6.4, I used UTF8toANSI on various places and that worked well.
Now after reading about FPC 3.x's unicode support, I thought that adding
this line say after begin in the main program could be enough:
SetMultiByteConversionCodePage(CP_UTF8);
The above setting makes UTF8ToAnsi to convert the string to UTF-16 and back to UTF-8.
It however doesn't work. How can I solve this without conversion to ANSI?
Do you want to convert an UTF8 string to the system codepage? Are you using any Lazarus units?
Mattias
I am developing a console application which receives an UTF-8 encoded text through stdin.…
Lubos Pintes wrote on Mon, 23 Nov 2015:
Show quoted text 10 lines
I am developing a console application which receives an UTF-8
encoded text through stdin.
The text is, after possible modification, sent to SAPI5. The SAPI5
generated interface wrappers have parameters of type string.
In FPC 2.6.4, I used UTF8toANSI on various places and that worked
well. Now after reading about FPC 3.x's unicode support, I thought
that adding this line say after begin in the main program could be
enough:
SetMultiByteConversionCodePage(CP_UTF8);
It however doesn't work.
That routine sets the default code page for ansistrings. It does not change I/O.
How can I solve this without conversion to ANSI?
The code page can be set per I/O channel:
http://www.freepascal.org/docs-html/3.0.0/rtl/system/settextcodepage.html (and Michael: I just noticed a typo on that page, under "Description" it says GetTextCodePage instead of SetTextCodePage)
Jonas
Hello,
Thanks for replies, I will try to explain this in other way: My system code page is windows-1250. When a string is sent to a COM component, it is somehow converted to UTF-16. Thus FPC knows how to do the conversion. Now I know that I am receiving text in UTF-8. Thus I called UTF8ToANSI to uselessly convert the UTF-8 to ANSI and then FPC converted it to UTF-16 to make COM happy.
Although it worked well in many cases, a data loss may occur. So to reformulate this question, how to avoid this useless conversion step? I thought that setting the default code page to UTF-8 just do the trick.
Dňa 23. 11. 2015 o 17:29 Jonas Maebe napísal(a):
Show quoted text 31 lines
Lubos Pintes wrote on Mon, 23 Nov 2015:
> I am developing a console application which receives an UTF-8 encoded
> text through stdin.
> The text is, after possible modification, sent to SAPI5. The SAPI5
> generated interface wrappers have parameters of type string.
> In FPC 2.6.4, I used UTF8toANSI on various places and that worked
> well. Now after reading about FPC 3.x's unicode support, I thought
> that adding this line say after begin in the main program could be
> enough:
> SetMultiByteConversionCodePage(CP_UTF8);
> It however doesn't work.
That routine sets the default code page for ansistrings. It does not
change I/O.
> How can I solve this without conversion to ANSI?
The code page can be set per I/O channel:
http://www.freepascal.org/docs-html/3.0.0/rtl/system/settextcodepage.html (and
Michael: I just noticed a typo on that page, under "Description" it says
GetTextCodePage instead of SetTextCodePage)
Jonas
_______________________________________________
fpc-pascal maillist -
http://lists.freepascal.org/cgi-bin/mailman/listinfo/fpc-pascal
Lubos Pintes wrote 7 lines
Hello,
Thanks for replies, I will try to explain this in other way:
My system code page is windows-1250. When a string is sent to a COM
component, it is somehow converted to UTF-16. Thus FPC knows how to do
the conversion. Now I know that I am receiving text in UTF-8. Thus I
called UTF8ToANSI to uselessly convert the UTF-8 to ANSI and then FPC
converted it to UTF-16 to make COM happy.
Do not convert to Ansi.
You should use unicodestring or widestring variables, and send that to COM.
Although it worked well in many cases, a data loss may occur. So to reformulate this question, how to avoid this useless conversion step? I thought that setting the default code page to UTF-8 just do the trick.
COM only works with UTF16. The best way is to declare all strings that you intend to send to COM to Widestring or UnicodeString.
Michael.