Ticket #4173 (new enhancement)

SVN Diffs for #4173

 

Opened 4 years ago

Last modified 6 months ago

Pick best available canonically-equivalent glyphs, Phase III

Reported by: mader(at)jtcsv.com Assigned to: eric
Priority: minor Milestone: 4.4
Component: layout Version: 3.0
Keywords: layout Cc:
Load: Xref: 3732, 4172
Java Version: Operating System: all
Project (C/J): ICU4C Weeks: 2
Review:

Description (Last modified by yoshito)

This is part 3 from jitterbug 3732. (It was described incorrectly in that bug)

Here's a better statement of the problem: assume that the font contains an a-umlaut and does not contain an a-dot_below. Then:

a + dot_below + umlaut

should be changed to:

a-umlaut + dot_below

Attachments

Change History

12/31/69 17:40:52 changed by notes

Another thought: given a + dot_below + umlaut, the canonical iterator will generate a + umlaut + dot_below and a_umlaut + dot_below as two alternatives. We could feed all two-accent combinations into the iterator, but we should be able to handle an aribtrary number of accents.

12/31/69 17:40:53 changed by auditor

  • Thu Jan 13 18:32:00 2005 emader changed notes
  • Mon Jan 17 18:03:18 2005 emader sent reply 1
  • Mon Jan 17 18:03:31 2005 emader changed notes2: xref: "3732" to "3732, 4172",
  • Mon Jan 17 18:03:31 2005 emader changed notes
  • Thu May 19 19:04:35 2005 weiv changed notes2: target: "3.4" to "3.6",
  • Thu May 19 19:04:35 2005 weiv changed notes
  • Fri Sep 16 10:44:41 2005 weiv changed notes2: priority: "committed" to "high",
  • Fri Sep 16 10:44:41 2005 weiv changed notes
  • Tue Oct 17 23:20:11 2006 emader changed notes2: target: "3.6" to "UNSCH",
  • Tue Oct 17 23:20:11 2006 emader changed notes

01/17/05 17:03:18 changed by Eric Mader <mader(at)jtcsv.com>

We might be able to use mark classes to make this work: if dot_below and umlaut are in two different classes, we could have a ligature table that skips over all marks but umlaut and does a + umlaut -> a_umlaut.

Another thought: given a + dot_below + umlaut, the canonical iterator will generate a + umlaut + dot_below and a_umlaut + dot_below as two alternatives. We could feed all two-accent combinations into the iterator, but we should be able to handle an aribtrary number of accents.

10/03/07 10:51:01 changed by eric

  • load changed.
  • java changed.
  • revw changed.
  • summary changed from RFE: Pick best available canonically-equivalent glyphs, Phase III to Pick best available canonically-equivalent glyphs, Phase III.

07/09/08 15:15:11 changed by yoshito

  • priority changed from major to assess.
  • description changed.

07/21/08 09:01:16 changed by hchapman

  • priority changed from assess to minor.
  • milestone changed from UNSCH to 4.1.1.

07/25/08 11:23:46 changed by yoshito

  • milestone changed from 4.1.1 to 4.4.

Add/Change #4173 (Pick best available canonically-equivalent glyphs, Phase III)




Anti spam check: