Full width punctuations U+2018, U+2019, U+201C and U+201D in Chinese

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Set font to Noto Sans CJK Traditional Chinese in browser and word processor 
2. Browse a web page or edit a document containing these characters

What is the expected output? What do you see instead?
These punctuation marks should be rendered proportionally. Instead they are 
full-width.

What version of the product are you using? On what operating system?
Noto Sans CJK Traditional Chinese, initial release.

Please provide any additional information below.
The texts in the attached screenshots are:
1. How Hooters Came to Symbolize Bond Market’s Desperation
2. 「這很好」英文可以說 “It’s good”

Original issue reported on code.google.com by ping...@google.com on 24 Jul 2014 at 8:33

Attachments:

• [apostrophe on plurk.png](https://storage.googleapis.com/google-code-attachments/noto/issue-94/comment-0/apostrophe on plurk.png)
• [Screen Shot 2014-07-24 at 11.30.48 AM.png](https://storage.googleapis.com/google-code-attachments/noto/issue-94/comment-0/Screen Shot 2014-07-24 at 11.30.48 AM.png)

[GoogleCodeExporter]

[deleted comment]

[GoogleCodeExporter]

Original comment by roozbeh@google.com on 24 Jul 2014 at 8:35

[GoogleCodeExporter]

Replacing the previous screenshot with a more blurred version on personal name.

Original comment by ping...@google.com on 24 Jul 2014 at 8:35

Attachments:

• [apostrophe on plurk.png](https://storage.googleapis.com/google-code-attachments/noto/issue-94/comment-3/apostrophe on plurk.png)

[GoogleCodeExporter]

Original comment by roozbeh@google.com on 25 Jul 2014 at 6:27

• Changed title: Full width punctuations U+2018, U+2019, U+201C and U+201D in Chinese

[GoogleCodeExporter]

Original comment by roozbeh@google.com on 25 Jul 2014 at 6:33

• Changed state: Accepted

[GoogleCodeExporter]

I want to get a claification on a couple of things:
- Are proportional shapes also desired for Traditional Chinese text? I 
understand  proportional shapes are desired for Western text, i.e., the "It's 
good" case.
- If we do context aware shapes, i.e., proportional shapes for English and full 
width shapes for T. Chinese in a Chinese-English mixed text, should we consider 
both sides of the quotation marks, or only left side of the quotation mark?


Also, for S Chinese, full-width forms are preferred when used for S chinese 
text and proportional forms are preferred for Western text.  If we have a good 
solution for the context aware shapping, we can also also apply that to S 
Chinese (and maybe Japanese and Korean too). 

Original comment by xian...@google.com on 30 Jul 2014 at 12:08

[GoogleCodeExporter]

Original comment by xian...@google.com on 30 Jul 2014 at 12:08

[GoogleCodeExporter]

The Japanese and Korean fonts and OTC font instances already use proportional 
glyphs for these four characters, and additionally include CJK-optimized forms 
(that are aligned to the em-box rather than to the Western cap height) which 
are accessible via the 'locl' GSUB feature when the text is language-tagged to 
a CJK one.

The ideal way to handle this is via the 'locl' GSUB feature, but the problem is 
that only a small number of environments support it, and it also depends on the 
user properly language-tagging the text. Another approach is the 'calt' 
(contextual alternates) GSUB feature, but there are Western and Chinese use 
cases that would break this.

In the end, I suspect that simply choosing a default and having some mechanism 
for accessing the other forms, such as the 'pwid' and 'fwid' GSUB feature, is 
the best overall approach.

Original comment by ken.lu...@gmail.com on 30 Jul 2014 at 12:56

[GoogleCodeExporter]

Issue 179 has been merged into this issue.

Original comment by roozbeh@google.com on 23 Oct 2014 at 4:50

[GoogleCodeExporter]

Original comment by xian...@google.com on 18 Feb 2015 at 7:44

[ghost]

Moved to notofonts/noto-cjk#5