Closed Bug 1656040 Opened 5 years ago Closed 5 years ago

letter-spacing should be ignored for cursive scripts

Categories

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

defect

Tracking

()

RESOLVED DUPLICATE of bug 1342835

People

(Reporter: manishearth, Unassigned)

Details

Attachments

(1 file)

Attached file testcase.html

From https://drafts.csswg.org/css-text/#cursive-tracking:

If it is able, the UA may apply letter spacing to cursive scripts by translating the total extra space to be distributed to a run of such letters into some form of cursive elongation (or compression, for negative tracking values) for that run that results in an equivalent total expansion (or compression) of the run. Otherwise, if the UA cannot expand text from a cursive script without breaking its cursive connections, it must not apply spacing between any pair of that script’s typographic letter units at all (effectively treating each word as a single typographic letter unit for the purpose of letter-spacing). Both cases will result in an effective spacing of zero between such letters; however the former will preserve the sense of stretching out the text.

There are some examples of how this should apply to Arabic.

In the attached testcase, both Firefox and Chrome seem to break every letter in Arabic and N'ko, which is clearly against the spec (there are even photos of this behavior in the spec as "what not to do")

Interestingly, it's also breaking Devanagari because it's segmenting at grapheme cluster boundaries, and त्यां is two grapheme clusters despite being a single typographical letter unit.

For the cursive scripts listed I think we should just ignore the property. Kashida justification is interesting but I suspect there is a lot of work involved in making it work.

For Indic scripts which have consonant clusters that are multiple grapheme clusters but a single typographical character unit, I'm not sure what the best thing to do is here. I have in the past proposed to Unicode that we treat consonant clusters for Indic scripts that default to conjucts (e.g. Devanagari but not Tamil). While this has made it into draft versions of the spec it was removed for reasons I cannot recall.

Attachment #9166881 - Attachment mime type: text/plain → text/html
Status: NEW → RESOLVED
Closed: 5 years ago
Resolution: --- → DUPLICATE
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: