[utf8vcl-general] [utf8vcl - General] RE: UTF8test project not working?

[SourceForge.net <no...@so...>]

Read and respond to this message at: 
https://sourceforge.net/forum/message.php?msg_id=4242032
By: akirabbq

Not really, there are other 2 bytes character system that is not Unicode. For
example, MBCS (widely used by CJK Chinese/Japanese/Korean) is 2 bytes long,
it consists of a leading byte ($81..$FE) + a trailing byte. The char value of
MBCS and Unicode are incompatible to each other,
MultiByteToWideChar/WideCharToMultiByte are needed for the conversion.

IMHO, the main issue of UTF-8 VCL so far is from the IDE. The IDE only stores
a string as AnsiString type when the component's property has the AnsiString
type. So if a widestring is being pasted into the object inspector, the widestring
will be converted to an AnsiString, which will destroy the integrity of the
original unicode text. The string conversion must be avoided at all cost, it
is impossible to mix different codepage ansistring into one single ansistring.

Some work have to be done on the IDE to let the IDE convert string into UTF8
during design time. For example, when some text is being typed/pasted into the
object inspector, the IDE converts the string into UTF8 first. I think it also
brings another issue that the designer must be able to display UTF8 string,
otherwise the readability of the designer will be terrible. If this cannot be
done, then the application's GUI will get stuck in AnsiString level, since the
designer wiped out all the unicode at design time. Pasting UTF8 string into
the object inspector will NOT work, since UTF8's highbyte character will mix
up with MBCS's leading byte, and trash the whole string.

______________________________________________________________________
You are receiving this email because you elected to monitor this forum.
To stop monitoring this forum, login to SourceForge.net and visit: 
https://sourceforge.net/forum/unmonitor.php?forum_id=671785