creating asian languages IF
sloke
I would like to know if there are available any system that can be
used to create IF using asian languages, eg Chinese, even if it is
just a miminal system, without all the linguistic parsing except for
simple pattern matching. I understand there are some that is unicode,
and so can display unicode text, but what about actions?
Thanks.
Ming Hua
> I would like to know if there are available any system that can be
> used to create IF using asian languages, eg Chinese, even if it is
> just a miminal system, without all the linguistic parsing except for
> simple pattern matching.
Not that I know of. I am pretty sure there is no Chinese IF authoring
tool, the situation for other languages is less clear to me.
That said, TADS3 is designed with Unicode in mind, and I've played
with the idea of writing a Chinese TADS3 standard library. It seems a
lot of work, but not impossible.
> I understand there are some that is unicode,
> and so can display unicode text, but what about actions?
For Chinese, the displaying part is actually not simple. As for
action commands, if we are not talking about complicated command
parsing, I don't see how that is particularly difficult or different
from English.
sloke
have any pointers on how to do so?
How about writing one from scratch then? Something like with gui like
ADRIFT rather than another interpreted language.
Why do you think there isn't one for Chinese? It has not caught on?
she's long gone
popular very few Chinese people had home computers. I've done research
into Korean IF and I can tell you definitively there is none
whatsoever. There are some Korean IF fan sites but they're dedicated
to English-language games.
It shouldn't really be any harder to write an Asian-language parser
than a European-language one. But I think it would be pretty hard to
gain traction with potential audiences.
Mike Rozak
able to handle Asian, barring a few bugs (which I'm willing to fix) and some
work on word-breaking for the text-to-speech synthesis.
Would word breaking be necessary for a command parser?
"sloke" <lok...@gmail.com> wrote in message
news:48f43972-f46f-4cfb...@s36g2000prg.googlegroups.com...
Lunas Specto
it has been modified to make games in various European languages?
Ming Hua
> Can Inform 6 be modified to make games in Asian languages the same way
As far as I know, the Z-machine backend can't handle Asian languages
because it has a very small fixed-size character set. The Glulx
backend probably can, but I don't know enough about Glulx.
Ming Hua
> So displaying unicode eg Chinese is still not easy in TADS3. Do you
> have any pointers on how to do so?
Not really. The TADS 3 documentation has instructions about using
other character sets, and I made it working a long time ago. My most
recent playing with TADS 3 wasn't as successful, though. I was
hanging out on the TADS 3 mailing list then, so maybe you can find
something useful in the list archive.
> How about writing one from scratch then? Something like with gui like
> ADRIFT rather than another interpreted language.
I don't see how that would be easier than modifying/extending TADS 3.
Actually I am pretty sure it would be harder. If you know
programming, though, you are free to try whatever you feel like, of
course.
David Kinder
> As far as I know, the Z-machine backend can't handle Asian languages
> because it has a very small fixed-size character set. The Glulx
> backend probably can, but I don't know enough about Glulx.
Z-code and Glulx can both handle Unicode (though with input limitations for
Z-code): there's an example in Inform 7 that prints out in Greek and
Cyrillic, for example. At least displaying Chinese characters should be
possible in either, provided the interpreter had a font with those
characters in it.
David
Ming Hua
I see that you are listed as the author/editor of "Z-Machine Standard
1.1 Document", so I believe you know much better about Z-machine than
I do. However after reading section 3, especially 3.8.5, of Z-machine
standard 1.0 (I can't find any full version 1.1 document), my
understanding is as follows:
Z-machine (v5 or later) provides a way to map ZSCII 155-251 to any
characters in Unicode BMP (U+0001 to U+FFFF), using a "unicode
translation table".
However, Chinese has a huge character set. The most basic daily usage
probably needs about 1,500 characters, and the commonly used
characters are in the 3,000-4,000 range. If my understanding about Z-
machine's Unicode support is correct, I fail to see how it can support
a full Chinese IF, which needs much more than the 97 extra characters
Z-machine provide.
Please correct me if my understanding is wrong.
Ming
2007.12.01
Ryusui
Well, I just tested with two full songs' worth of Japanese lyrics
complete with kanji, and I didn't run into any such limitations. The
problem, however, is that Inform 7 translates Unicode characters into
substitutions based on their full names, meaning that a couple of
lines of text will run smack into the compiler's limitation on string
length. It's possible to work around this by separating paragraphs
into lines printed with individual "say" instructions, but it's fairly
kludgy. (Line breaking, to my experience, must also be handled with
explicit line break substitutions in this case.)
David Kinder
> I see that you are listed as the author/editor of "Z-Machine Standard
> 1.1 Document", so I believe you know much better about Z-machine than
> I do.
I only made some minor revisions to that: all the hard work was done by
Graham Nelson (for Standard 1.0) and then Kevin Bracey and Jason Penney.
> Z-machine (v5 or later) provides a way to map ZSCII 155-251 to any
> characters in Unicode BMP (U+0001 to U+FFFF), using a "unicode
> translation table".
This is true. However, for output you can get around this with the
@print_unicode opcode, which allows the game to output any Unicode character
in the range U+0001 to U+FFFF, regardless of whether or not it's in the
Unicode translation table.
> However, Chinese has a huge character set. The most basic daily usage
> probably needs about 1,500 characters, and the commonly used
> characters are in the 3,000-4,000 range. If my understanding about Z-
> machine's Unicode support is correct, I fail to see how it can support
> a full Chinese IF, which needs much more than the 97 extra characters
> Z-machine provide.
Output is possible, but tackling input in Z-code is a rather more difficult
problem. At that point I would give in and switch to Glulx or TADS, which
both allow arbitrary Unicode input and output.
Still, given the number of possible input characters, I don't know how a
game library could cope with Chinese input even if it could read in the
Unicode input. This may go some way to explaining why there aren't any text
adventure systems supporting Chinese that anyone's aware of ...
David
Depresiv
I don't know if there are any in Chinese, but there seems to be a
series of several pseudo-interactive fiction games in Japanese, called
"Misty", by Data West.
http://fullmotionvideo.free.fr/screen/d/misty11.png
http://fullmotionvideo.free.fr/screen/d/misty21.png
http://fullmotionvideo.free.fr/screen/d/misty31.png
http://fullmotionvideo.free.fr/screen/d/misty41.png
http://fullmotionvideo.free.fr/screen/d/misty51.png
Hope it helps.
--Depresiv