358433 - no longer break between CJK characters at inline frame boundary

Status: RESOLVED FIXED

(Core :: Layout: Text and Fonts, defect)

Description

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Bug 343445 (or something else checked in that day) regressed a testcase that I'd already added to reftests testing whether we break between CJK characters whether or not there's a frame boundary.

Steps to reproduce: load http://lxr.mozilla.org/seamonkey/source/layout/reftests/bugs/351641-2b.html

Expected results: should look just like http://lxr.mozilla.org/seamonkey/source/layout/reftests/bugs/351641-2-ref.html

Actual results: Some pairs of characters do not break.

Regressed between Linux nightlies 2006-10-18-04-trunk and 2006-10-19-04-trunk.


When you fix this, please change the reftest manifest in layout/reftests/reftest.list to remove the "f" (for expected failure) before the "==" for this test (and the comment pointing to this bug).

Comment 1

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

roc's landing today of bug 370588 apparently fixed this for Windows only; still broken on Mac and Linux.

Comment 2

[Robert O'Callahan (:roc) (email my personal email if necessary)]

Actually it only "fixed" it by causing all Japanese text to disappear and have no height, because that tinderbox doesn't have the fonts to render Japanese text.

Comment 3

[Martijn Wargers (dead)]

Fwiw, it's failing for me now on windows (mingw).

Comment 4

[ChangZhuo Chen]

It is okay for me on Linux x86_64. The character `の' are all show in one column.

Other platforms are also okay in try server [1].

[1] https://tbpl.mozilla.org/?tree=Try&rev=37c80bf7ddc5

Comment 5

[Jonathan Kew [:jfkthame]]

This was fixed by bug 367177 (note that the "fails" annotation removed from the reftest manifest), back in mid-2007. Closing.