Bug 80072 – [chromium] Japanese vertical text displayed in uneven spacing

• Component: Layout and Rendering
• Status: RESOLVED WONTFIX

[Xiaomei Ji]

Created attachment 129773 [details]
spacing and parenthesis in cr-linux

load http://dl.dropbox.com/u/64492542/epub3/verttest.html.

right under the title, you could see that the beginning parenthesis "(" is displayed as is without rotation for vertical text, while ending parenthesis is.

Also, if you look at the text closely, you could see that text are bit crowded. Looks like after ruby signs appear, the lines without ruby sign are closely adjacent to the previous line, instead of having a space (with the same width as where ruby signs are displayed) in between.

I attached the snapshot for both mac and linux for comparison.

[Xiaomei Ji]

Created attachment 129774 [details]
spacing and parenthesis in cr-mac

[Koan-Sin Tan]

The parenthesis problem was caused by a typo. The opening one is a FULLWIDTH LEFT PARENTHESIS (U+FF08) instead of U+0028. To display it correctly, you need a font with proper vert/vrt2 tables.

[Koan-Sin Tan]

regarding the space problem, on my system, it doesn't look like the space between vertical lines is gone.

[Koan-Sin Tan]

Created attachment 130564 [details]
screenshot 

it seems to me line space is there

[Xiaomei Ji]

Created attachment 130903 [details]
safari screenshot

I attached screenshot from Safari as well.
If you look closely on the cr-mac and Safari screenshots,  you will see the line space between each line is fixed width.

If you look at your screenshot, you can see the line space between each line is not fixed.
If you look at the 1st and 2nd line in main content, there is wider space in between. It is approximately the same wide as the line space between 2nd and 3rd line (while 3rd line has ruby).
If you look at space between line 4th and 5th, you can see the space in between is narrower.

What I heard from Japanese and traditional Chinese speaker is that the space between lines should be fixed.

[Koji Ishii]

I suspect "wrong parenthesis" part is dup of bug 69282 for cr-linux. The plan is to finish Apple Win port in bug 48459, port it to cr-win in bug 51450, and [e-mail] will port it to bug 69282. It's a bit long way though.

Line pitch with ruby seems to be a separate bug to me. Probably you should change this bug to the line pitch with ruby. For regular Japanese publications, author usually set line-height to 1.8 or 2, and the bug will not appear. It only happens when line-height isn't big enough to show ruby.

[Xiaomei Ji]

(In reply to comment #6)
> I suspect "wrong parenthesis" part is dup of bug 69282 for cr-linux. The plan is to finish Apple Win port in bug 48459, port it to cr-win in bug 51450, and [e-mail] will port it to bug 69282. It's a bit long way though.

I do not quite understand why this is related to  bug 48459.
vertical display is not supported in Windows port yet.
But it is supported in cr-linux. As to the wrong parenthesis, if there is no vert or vrt2 table, we will need to create a separate mapping table, is it?

[Koji Ishii]

I see your point, I thought you're using fonts with vert, and I'm not very aware of the current state of cr-linux port.

So there are two issues:
1. Support vertical flow with fonts without vert table (note that the current CSS spec doesn't use vrt2; just vert)
2. Line height is calculated incorrectly when there's ruby and line height isn't big enough to hold ruby

I think you should file two separate bugs. The former is a kind of feature addition. How common is such font? Right now, the CSS spec doesn't support such fonts and therefore it will likely be webkit extension to the spec (remember, the spec is still WD.) You may want to discuss at [e-mail] to allow UA to support such fonts if you think you really need it, so that it becomes part of the official spec.

The later is a separate bug in ruby code, and it reproduces on other ports as well. It's hard to work on if the two are mixed in single bug.

[Xiaomei Ji]

I've changed the bug title. And the 1st issue is probably a dup of bug 69282. I will file a separate one if I found out that it is not a dup.

[Stephen Chenney]

https://code.google.com/p/chromium/issues/detail?id=231292