Bug 165084 – REGRESSION(r205396): ComplexPath is unable to render correctly when some text fallback is involved

• Component: Text
• Status: RESOLVED FIXED

[Ebrahim Byagowi]

Created attachment 295459 [details]
Actual

Steps to repro:
data:text/html;charset=utf8,<u style="font-family:Helvetica, Arial, sans-serif" dir="RTL">AAA %D8%B4%D8%B3%DB%8C%D8%B4%D8%B3%DB%8C %D8%B4%D8%B3%DB%8C%D8%B4%D8%B3%DB%8C %D8%B4%D8%B3%DB%8C%D8%B4%D8%B3%DB%8C %DA%AF%D8%B2%DB%8C%D9%86%D9%87%D9%94 %D8%B4%D8%B3%DB%8C%D8%B4%D8%B3%DB%8C.

Actual:
Some part of the text fell under the underline

Expected:
Same rendering with Firefox and Chrome.

[Myles C. Maxfield]

Which OS are you using?

I've attached a screenshot showing it working correctly on 16B2555.

[Myles C. Maxfield]

Created attachment 295535 [details]
Screenshot on 16B2555

[Myles C. Maxfield]

Created attachment 295536 [details]
Screenshot on 16B2555

[Myles C. Maxfield]

Do you have any custom fonts installed?

[Ebrahim Byagowi]

This seems like a newly regression and I am seeing it just with Safari Technology Preview 18 and not with the default installed version of Safari. Are you testing with it?

[Radar WebKit Bug Importer]

<rdar://problem/29464658>

[Ebrahim Byagowi]

Another test case which doesn't involve fallback but an immediate start of another tag after a word containing diacritics,

https://fa.wikipedia.org/wiki/User:Ebrahim/t1?printable=yes&withCSS=MediaWiki:Gadget-Vazir.css

third line (after web font load) should be more like Firefox and Chrome.

[Myles C. Maxfield]

Created attachment 296227 [details]
Raw AppKit

[Myles C. Maxfield]

(In reply to comment #8)
> Created attachment 296227 [details]
> Raw AppKit

Added a screenshot of the test case using raw AppKit. The problem doesn't occur there, which means this is a WebKit bug.

I am able to reproduce this on a recent build.

[Myles C. Maxfield]

Created attachment 296228 [details]
WebKit

[Myles C. Maxfield]

The run with the following code points:
U+06af U+0632 U+06cc U+0646 U+0647 U+0654

has both an initial advance as well as origins.

[Myles C. Maxfield]

This is caused by r205396

[Myles C. Maxfield]

*** Bug 165464 has been marked as a duplicate of this bug. ***

[Myles C. Maxfield]

Created attachment 296282 [details]
Patch

[Myles C. Maxfield]

Created attachment 296283 [details]
Patch

[Build Bot]

Comment on attachment 296283 [details]
Patch

Attachment 296283 [details] did not pass ios-sim-ews (ios-simulator-wk2):
Output: http://webkit-queues.webkit.org/results/2632700

New failing tests:
fast/text/initial-advance-in-intermediate-run-complex.html
imported/blink/fast/text/international/vertical-positioning-with-combining-marks.html

[Build Bot]

Created attachment 296285 [details]
Archive of layout-test-results from ews123 for ios-simulator-wk2

The attached test failures were seen while running run-webkit-tests on the ios-sim-ews.
Bot: ews123  Port: ios-simulator-wk2  Platform: Mac OS X 10.11.6

[Myles C. Maxfield]

Created attachment 296300 [details]
Patch

[Myles C. Maxfield]

Created attachment 296301 [details]
Patch

[Simon Fraser (smfr)]

Comment on attachment 296301 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=296301&action=review

> Source/WebCore/platform/graphics/mac/ComplexTextController.cpp:674
> +        if (!complexTextRun.glyphOrigins() && r && m_adjustedBaseAdvances.size()) {

Can we rename 'r'?

[Myles C. Maxfield]

Committed r209410: <http://trac.webkit.org/changeset/209410>