W3C home > Mailing lists > Public > public-css-testsuite@w3.org > December 2015

Re: text-baseline tests with 'sideways-lr' and 'sideways-rl'

From: Gérard Talbot <css21testsuite@gtalbot.org>
Date: Thu, 03 Dec 2015 16:51:54 -0500
To: 塩澤 元 (Shiozawa, Hajime) <hajime.shiozawa@gmail.com>
Cc: Public CSS test suite mailing list <public-css-testsuite@w3.org>, Koji Ishii <kojiishi@gluesoft.co.jp>
Message-ID: <49a7a47a179260112b7d2865f1620ac7@gtalbot.org>
Le 2015-11-25 09:13, 塩澤 元 a écrit :
> Gérard,
> I have submitted text-baseline tests with 'sideways-lr' and 
> 'sideways-rl'.
> 1. added new tests
> https://hg.csswg.org/test/rev/a027a7c1ca48


> Could you review it?


Sorry for the long delay.


line 8:    <meta name="assert" content="This test checks the generation 
of text baseline. When 'writing-mode' is 'sideways-lr' and when 
'text-orientation' is 'mixed', then the alphabetical baseline is used as 
the dominant baseline." />

line 16:       text-orientation: mixed;

As I understand 'writing-mode: sideways-lr', the writing mode is set and 
text-orientation is also set. So, as I understand this, setting 
text-orientation to whatever value should have not effect at all when 
'writing-mode' is set to 'sideways-[ lr | rl ]'. In which case, I think 
you should remove line 16 and I would reword the text assert to say:

line 8:     <meta name="assert" content="When 'writing-mode' is 
'sideways-lr', then the alphabetic baseline is used as the dominant 
baseline." />

Nit: it's "alphabetic" and not "alphabetical".

- - - - - - - - - - - -


Same comment as for text-baseline-slr-009

- - - - - - - - - - - -


'text-orientation: upright' has no effect on the orientation of latin 
glyphs because 'writing-mode' is set to 'sideways-lr'

- - - - - - - - - - - -

Once 'writing-mode: sideways- [ lr | rl ]' has been set, the 
'text-orientation' can not be reset individually. sideways- [ lr | rl ] 
is not like other shorthand properties.

When the draft spec says

The text-orientation property then determines how text is laid out 
within the line box.

this is not perfectly true or accurate. The text-orientation property 
can determine how text is laid out within the line box *_only if_* 
'writing-mode' is set to 'vertical-*'. sideways-[ lr | rl ] writing 
modes are still considered as vertical writing modes.

- - - - - - - - - - - -


line 8:    <meta name="assert" content="This test checks the generation 
of text baseline. When 'writing-mode' is 'sideways-lr' and when 
'text-orientation' is 'sideways', then the alphabetical baseline is used 
as the dominant baseline." />

Once 'writing-mode' has been set to 'sideways-[lr|rl]', then the 
dominant baseline is the alphabetic baseline. After that, setting 
'text-orientation' to any value has no effect.

[Addendum: I just had a phone discussion with Elika and she agrees that 
we should split your tests.

should be just testing the writing-mode versus dominant baseline used 
and then we should have another test making sure, verifying that 
text-orientation declarations have no effect whatsoever when the 
writing-mode is 'sideways-*'. As coded, your current 
text-baseline-s*-00[8-13] tests are testing 2 features at the same time 
and we think we should check each feature separately, in separate tests.

Okay... I think we should adjust text-baseline-srl-008 and 
text-baseline-slr-009 accordingly.

Then create 4 tests using <div>月火水Abc<br />def木金土</div>: 1 writing mode 
test with 'writing-mode: sideways-rl' with 'text-orientation: upright' 
and 3 tests with 'writing-mode: sideways-lr' with 'text-orientation: [ 
mixed | upright | sideways ]' where those 4 tests check, make sure that 
'text-orientation' declarations have no effect in those 4 'sideways-*' 
tests. Those 4 tests should probably go in section 5.1. Orienting Text: 
the 'text-orientation' property.

Hajime, do you want to do this? I can do it if you prefer..


Test Format Guidelines

Test Style Guidelines

Test Templates

CSS Naming Guidelines

Test Review Checklist

CSS Metadata
Received on Thursday, 3 December 2015 21:52:27 UTC

This archive was generated by hypermail 2.4.0 : Friday, 20 January 2023 19:58:21 UTC