Bug 123846
Summary: | unicode-bidi: plaintext does not work properly when a line wraps | ||
---|---|---|---|
Product: | WebKit | Reporter: | Ryosuke Niwa <rniwa> |
Component: | Text | Assignee: | Nobody <webkit-unassigned> |
Status: | RESOLVED CONFIGURATION CHANGED | ||
Severity: | Normal | CC: | ahmad.saleem792, ap, enrica, igor.oliveira, jonlee, mitz, mmaxfield, rniwa, zalan |
Priority: | P2 | Keywords: | BlinkMergeCandidate |
Version: | 528+ (Nightly build) | ||
Hardware: | Unspecified | ||
OS: | Unspecified |
Ryosuke Niwa
Merge https://chromium.googlesource.com/chromium/blink/+/d5088d76fcf75eaf23e30be43ebf119b2343dfed
unicode-bidi: plaintext never worked properly when there is a line wrapping in a
span with unicode-bidi: plaintext.
The problem happens because constructBidiRunsForSegment does not have
enough logic to understand the difference between line wrapping vs line
breaking, the current patch just adds the necessary logic(e.g when a text has a
line wrapping it needs to determine the directionality of the paragraph from the beginning of the isolatedInline)
Attachments | ||
---|---|---|
Add attachment proposed patch, testcase, etc. |
Ahmad Saleem
I took test case from the patch:
Link (removed -webkit-) - https://jsfiddle.net/q8m3o1tu/show
Link (not removed -webkit-) - https://jsfiddle.net/t9r2s41n/1/show
All browsers (Chrome Canary 108, Firefox Nightly 107 and Safari 16) render all boxes same with the same direction of arrows and stuff (no wrapping problem).
Do we need this anymore? Thanks!
zalan
There's also a hidden IFC progression here.