The Unicode Consortium Discussion Forum

The Unicode Consortium Discussion Forum

 Forum Home  Unicode Home Page Code Charts Technical Reports FAQ Pages 
 
It is currently Thu Oct 30, 2014 12:31 pm

All times are UTC - 6 hours [ DST ]





Post new topic Reply to topic  [ 6 posts ] 
Author Message
 Post subject: The approach of UTR#50 is different from Japanese standards
PostPosted: Tue Sep 04, 2012 4:39 pm 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
In my understanding, the basic concept of JIS X 4051:2004 "Formatting rules for Japanese documents" and Requirements for Japanese Text Layout (JLReq: http://www.w3.org/TR/jlreq/) is different from the concept of UTR#50.

The concept of JIS X 4051 is as follows:

1. It defines Japanese traditional characters, which consists of Kanji, Katakana, Hiragana, punctuations, Arabic numbers and some symbols. Japanese traditional text consists of Japanese traditional characters.

Remark: Requirements for Japanese Text Layout (http://www.w3.org/TR/jlreq/) adds many Western characters to one of a Japanese traditional character group (cl-19).

2. Then it defines Western text: Western text consists of Western characters, Arabic numbers, and some symbols. Western text is horizontal writing only.

3. Japanese documents may consist of Japanese traditional texts and Western texts. When Japanese horizontal documents includes Western inline text such as words or short text, the Western inline text will be formatted by Western formatting rules and there should be a small gap between both texts. In short, Japanese traditional characters and Western text belongs to different world.

As for vertical writing documents, things become quite complex.

1. Almost all Japanese traditional characters are upright, excluding some punctuations and symbols. JIS X4051 includes tables of special characters which take different directions and shapes in vertical writing.
1) 41 characters need rotation
2) 46 characters take different shape
3) 9 characters are horizontal only and 4 characters are vertical only.

In addition, within general characters, there are a small number of characters which should change shape in vertical writing such as square mili-meter.

2. As for Western text within vertical Japanese documents, there are three ways of text composition:
1) Treat each character as a Japanese traditional character. Each character is upright.
2) Rotate a run of characters, such as a word and a short text, sideways.
3) Two to three characters may be written horizontally, horizontal-in-vertical setting.

As for example and details of Western text within vertical Japanese documents, please refer to "3.2.3 Mixed Text Composition in Vertical Writing Mode" of JLReq.

There are wide variation of styles regarding how to mix 1), 2) and 3) among Japanese publications.

a) Japanese major newspapers adopt 1) and 3). The big change regarding Arabic numbers was implemented between 2001 to 2009.
b) The style of book publications seems to be changing now perhaps by the influence of personal computers and internet.

3. Difficult point is: many Unicode characters belong to both of Japanese traditional characters and Western characters. Japanese specialists carefully avoided to define directions of such characters until now.

<Conclusion>
It will be needless, impossible and sometimes harmful to define the default direction of all characters, character by character. I suppose it will enough to define directions of about a hundred characters in accordance with X 4051. As a result, the definition will be similar to SVO.

Regards,

Tokushige Kobayashi
Antenna House, Inc.
http://blog.cas-ub.com/


Top
 Profile  
 
 Post subject: Re: The approach of UTR#50 is different from Japanese standa
PostPosted: Sat Sep 08, 2012 8:33 pm 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
One naive question is: Why do we need to restrict our world to that of JIS X 4051? As you correctly pointed out, there are a wide variety of usages of even a single character in a vertical line. I agree with this point. You seem to have found the SVO is the most reasonable and useful for your purposes. That's great. But if you mean UTR#50 or MVO in particular has no use, I don't understand why. Each definition set has its own limitations and benefits. Tailoring and adjustments may be necessary for different purposes. But as you mentioned there are different worlds, there can be different ways of defining character orientations.

Regards,

--Taro


Top
 Profile  
 
 Post subject: Re: The approach of UTR#50 is different from Japanese standa
PostPosted: Mon Sep 10, 2012 8:46 pm 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
tyamamot wrote:
One naive question is: Why do we need to restrict our world to that of JIS X 4051?


In the language of the whole world, there are relatively few languages ​​that write vertically. At this moment, in the languages that adopt vertical writing system, the most popular language in the world is Japanese. In the past, the most frequently used character in vertical writing was Chinese characters. There were a number of characters to write vertically in the vicinity of the cultural sphere of Chinese - Manchu, Mongolian, Tangut character, Khitan characters, Hangul, and Chunomu. However, with an implementation of Simplified Chinese, Chinese in mainland China have already migrated to the horizontal. In addition, many of the vertical writing characters disappeared. The number of characters that is written vertically is decreasing.

In the vertical writing text, how to place characters that you can write only horizontally is a very difficult issue. JIS X4051 handles this issue correctly. Though "Requirements for Japanese Text Layout" (JLReq) deals with this issue, JIS X4051 is better than JLReq.

I suppose JIS X4051 would be a single good specification in the world which specifies the method to place Latin alphanumerics in vertical writing text. For instructions on how to place horizontal characters in vertical writing text, to respect the JIS X4051 is a matter of course.

I want to ask reversely: is there any other documents that specifies how to place horizontal writing characters in vertical writing text? I would like you to teach me if there is.

tyamamot wrote:
But if you mean UTR#50 or MVO in particular has no use, I don't understand why.


Unicode character is an abstract entity. The shape of the character is said glyph, printing is the visualization of the image of glyphs. Regarding the way of the conversion process from character to glyph image, there are various differences by character. Generally speaking, to map the code point directly to the corresponding glyph image is useless, as a technique for printing characters in the world.

UTR#50 and MVO is a method of defining the orientation of glyph image of the character code point directly. As mentioned above, this approach as a general rule is incorrect. Following this, I describe in detail an example.

Glyph image/orientation in vertical text is not the characteristics of the code points of the characters. Code points of the characters and glyph images/orientation are independent variable.

For example, InDesign offers a function to display all ASCII characters in the full-width form and upright. Vice versa, I suppose that it may also represent full-width character in a proportional shape. In other words, InDesign may treat glyph image and code point of the character independently. Though I am not familiar with InDesign, it should be.

Here is another example. The contents of XML documents is a sequence of character codes.

Japanese text is a quite special in the World as it can be represented both in horizontal and vertical mode.

Re-usability will be lost if the direction-attribute is attached to the content itself. In order to be able to use a XML document both in horizontal and vertical, the content should be marked up as direction neutral. The orientation of the glyph images should be specified by stylesheet when the document is rendered vertically.

UTR#50 and MVO defines the orientation of the code points of the characters. Full-width variant codes are defined upright, ASCII characters are defined sideways. The adoption of MVO and UTR#50 will limits the re-usability of the Japanese text.

Also, in the Japanese vertical books, horizontal and vertical texts are mixed in a single page. In addition, it is normal that the body pages are vertical and index pages are horizontal. If we adopt UTR#50 and MVO, it will become inconvenient to use the horizontal text and vertical text in a single page/book.

Best regards,

Tokushige Kobayashi
Antenna House, Inc.


Top
 Profile  
 
 Post subject: Re: The approach of UTR#50 is different from Japanese standa
PostPosted: Sat Sep 15, 2012 5:55 am 
Offline

Joined: Sat Oct 29, 2011 12:06 pm
Posts: 35
I can understand that we should respect what's written in JIS X 4051, as a standard. And I can also agree that JIS X 4051 is more consistent and better than JLReq. Still, JIS X 4051's scope is narrower. And many JIS-historical elements have been inherited into JIS X 4051, better or worse. As a practical standard, I recommend JIS X 4051 to anyone. But, it doesn't mean we should not look at anything else.

Quote:
is there any other documents


JLReq is one good example for its comprehensive nature. But there have been some books on Japanese typesetting, such as those from Editors' School, whose scope of content may be limited and narrower, though.

Quote:
For example, InDesign offers a function to display all ASCII characters in the full-width form and upright. Vice versa, I suppose that it may also represent full-width character in a proportional shape. In other words, InDesign may treat glyph image and code point of the character independently.


Right. Based on the character-glyph model, one coded character can have multiple representational glyphs. But it does not mean that a coded character is completely free from the representational form, because a character needs to represent its "graphic concept", in order for the character to be correctly mapped to its representational glyphs. Without it, why can you see whether a "−90 degrees ROTATED" glyph is correctly rotated or not?

Quote:
Glyph image/orientation in vertical text is not the characteristics of the code points of the characters. Code points of the characters and glyph images/orientation are independent variable.


I completely disagree with this, for the reason mentioned above.

What I can agree with is only the fact that a coded character can have multiple different vertical orientations, depending on the context and purposes, etc.

Quote:
Re-usability will be lost if the direction-attribute is attached to the content itself. In order to be able to use a XML document both in horizontal and vertical, the content should be marked up as direction neutral. The orientation of the glyph images should be specified by stylesheet when the document is rendered vertically.


I can understand what you try to mean. And, I can even agree that there are cases where a good level of compatibility between the writing directions should be retained on the source text level. However, I don't know whether it's always possible that the content should be marked up as direction neutral or not. It's possible to minimize the writing-direction-dependency, but it will be impossible to eliminate all the writing-direction dependent glyph style selection.

Quote:
Also, in the Japanese vertical books, horizontal and vertical texts are mixed in a single page. In addition, it is normal that the body pages are vertical and index pages are horizontal.


In this case, why can you always guarantee that the same text represented in the horizontal and vertical text glyph instances appearing on this page, if any, can be the same or compatible. Why can you make the two instances always compatible, in the sense of writing-direction-independence?

Regards,

--Taro Yamamoto


Top
 Profile  
 
 Post subject: Re: The approach of UTR#50 is different from Japanese standa
PostPosted: Mon Sep 17, 2012 12:40 am 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
tyamamot wrote:
Based on the character-glyph model, one coded character can have multiple representational glyphs. But it does not mean that a coded character is completely free from the representational form, because a character needs to represent its "graphic concept", in order for the character to be correctly mapped to its representational glyphs. Without it, why can you see whether a "−90 degrees ROTATED" glyph is correctly rotated or not?

Quote:
Glyph image/orientation in vertical text is not the characteristics of the code points of the characters. Code points of the characters and glyph images/orientation are independent variable.


I completely disagree with this, for the reason mentioned above.

What I can agree with is only the fact that a coded character can have multiple different vertical orientations, depending on the context and purposes, etc.


You are correct. "Code points of the characters and glyph images/orientation are independent variable." is not an appropriate expression. The mapping from character sequence to glyph image needs very difficult procedure in some scripts. Japanese vertical writing needs a little complicated operations to map from code-points to glyph image. Until now, the operation has been done by DTP software and DTP operators. But the situation will change. Creators must markup the contents and make stylesheets. Bowsers or EPUB reader will take place of DTP software.

Quote:
It's possible to minimize the writing-direction-dependency, but it will be impossible to eliminate all the writing-direction dependent glyph style selection.


I agree that. What I want to say is that text string should be writing-direction-independent as much as possible, and glyph selection should be specified by stylesheets. That is the purpose of CSS. What I oppose to MVO strongly is that it designate R to ASCII code-points and U to its full-width variant.

Quote:
In this case, why can you always guarantee that the same text represented in the horizontal and vertical text glyph instances appearing on this page, if any, can be the same or compatible. Why can you make the two instances always compatible, in the sense of writing-direction-independence?


I can not understand what actually you mean.

Best Regards,

Tokushige Kobayashi


Top
 Profile  
 
 Post subject: Re: The approach of UTR#50 is different from Japanese standa
PostPosted: Fri Sep 21, 2012 3:31 pm 
Offline

Joined: Thu Jan 12, 2012 9:58 am
Posts: 29
TKobayashi wrote:
tyamamot wrote:
Based on the character-glyph model, one coded character can have multiple representational glyphs. But it does not mean that a coded character is completely free from the representational form, because a character needs to represent its "graphic concept", in order for the character to be correctly mapped to its representational glyphs.

.. a coded character can have multiple different vertical orientations, depending on the context and purposes, etc.


Japanese vertical writing needs a little complicated operations to map from code-points to glyph image.


JIS X4051 and JLReq say western characters (ASCII alphabets and Arabic numerals, etc.) can have two glyphs; proportional and full-width.

In horizontal writing, the glyphs of western characters should be proportional, full-width glyph may not be used.

In vertical writing, as I described in the first message, there are three ways of text composition:
<repetition>
1) Treat each western character as a Japanese traditional character. Each character is upright.
2) Rotate a run of western characters, such as a word and a short text, sideways.
3) Two to three western characters may be written horizontally, horizontal-in-vertical setting.
</repetition>

Now, I can describe the three ways more clearly with character-glyph model:
1) Treat each western character as a Japanese traditional character. Each glyph of the character is full-width and the orientation is upright.
2) Rotate a run of western characters, such as a word and a short text, sideways. Each glyph of the character is proportional and the orientation is sideways.
3) Two to three western characters may be written horizontally, horizontal-in-vertical setting. Each glyph of the character is proportional and the orientation is upright.

In all three cases, the character may be same and the width and orientation of glyph is set accordingly.

Best regards,

Tokushige Kobayashi
Antenna House, Inc.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 6 posts ] 

All times are UTC - 6 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 0 guests


Quick-mod tools:
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Jump to:  
cron
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
Template made by DEVPPL.com