bugFree UCS Outline Fonts - Bugs: bug #52383, Poor positioning of recently...


bug #52383: Poor positioning of recently encoded Malayalam marks

Submitted by:  David Corbett <dscorbett>
Submitted on:  Fri 10 Nov 2017 02:37:54 AM UTC  
Category:  individual character(s) Severity:  3 - Normal
Item Group:  letter positioning poor Status:  Fix posted
Privacy:  Public Assigned to:  Steve White <Stevan_White>
Open/Closed:  Open Release:  development

Add a New Comment(Rich Markup)

You are not logged in

Please log in, so followups can be emailed to you.


( Jump to the original submission)

Sat 11 Nov 2017 03:09:26 PM UTC, comment #6:

Using HarfBuzz 1.6.3, the candrabindu and anusvara work equally well. If the anusvara doesn’t work with some other shaper, it is probably not updated for Unicode 10. Anyway, it is not a font bug.

'abvm' doesn’t ignore mark glyphs as a special case; that is just how mark-to-base lookups work. For ⟨q dot-below dot-above⟩ you don’t need to do anything special to skip the dot below: the dot above will attach to the q as if the dot below were not there. Similarly, for ⟨consonant vowel anusvara⟩ the anusvara will attach to the consonant regardless of the vowel, assuming the vowel glyph is a mark glyph.

You should generally not change glyph classes because there is generally no reason to. When you have a reason, go for it. Characters’ glyph classes do not need to be standardized between fonts.

Revision 3754 set some vowel signs’ glyph classes explicitly. However, it skipped u_mal, uu_mal, r_voc_mal, and rr_voc_mal, which are post-base vowels too. It also set the glyph class of the pre-base and two-part vowel signs, which is unnecessary, but which I am pretty sure has no ill effects.

I don’t know why FontForge decided that the Devanagari ī should be a mark glyph but the Malayalam ī shouldn’t.

David Corbett <dscorbett>
Sat 11 Nov 2017 09:46:23 AM UTC, comment #5:

Well, that fixes the candrabindu, but I guess something else is amiss with the combining anusvara. Changes in SVN.

I've always been leery of changing the OT Glyph Claas of Unicode-encoded characters, because I guess it should be standardized, and because I was unsure who or what was responsible for it, and also it was unclear to me just what effects it would have. For example, how would I know 'abvm' ignores Mark glyphs? Where do we read about these things?

Now that I look at built font files, the Devnanagari ii-sign is automatically a Mark glyph. How come the Malayalam ii-sign isn't? Is this a bug?

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Fri 10 Nov 2017 11:51:39 PM UTC, comment #4:

You can position the candrabindu or anusvara relative to the consonant by making the post-base vowel signs (U+0D3E..0D44 and U+0D57) be mark glyphs instead of base glyphs. This will allow 'abvm' to ignore the vowels. The vowel signs will still have width, despite being marks.

David Corbett <dscorbett>
Fri 10 Nov 2017 10:19:56 PM UTC, comment #3:

My only idea was to put anchors to the left of the tall vowels. See SVN. But what of the other vowels?

This affects the candrabindu but not anusvara. I guess the latter is just too new for the software.

Are we sure of the encoding here?

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Fri 10 Nov 2017 05:33:18 PM UTC, comment #2:

The centered anusvara and candrabindu now overlap the reph.

The anusvara is still small. Is that intentional?

My last complaint was about more than the centering. The presence of the vowel sign blocks mark positioning, so the anusvara or candrabindu overlaps the matra.

David Corbett <dscorbett>
Fri 10 Nov 2017 10:07:10 AM UTC, comment #1:

Yes. I have now added anchors. In SVN.

Is the last complaint you list separate from the failure to center?

Steve White <Stevan_White>
Project AdministratorIn charge of this item.
Fri 10 Nov 2017 02:37:54 AM UTC, original submission:

U+0D3B MALAYALAM SIGN VERTICAL BAR VIRAMA should be centered. An example is ⟨സ഻⟩.

U+0D00 MALAYALAM SIGN COMBINING ANUSVARA ABOVE should be centered. An example is ⟨ഩഀ⟩. Also, the examples in its Unicode proposal give it a larger glyph than FreeFont does.

U+0D00 MALAYALAM SIGN COMBINING ANUSVARA ABOVE and U+0D01 MALAYALAM SIGN CANDRABINDU overlap high vowel signs, as in ⟨കിഀ⟩ and ⟨കിഁ⟩.

David Corbett <dscorbett>


(Note: upload size limit is set to 16384 kB, after insertion of the required escape characters.)

Attach Files:

No files currently attached


Depends on the following items: None found

Items that depend on this one: None found


Carbon-Copy List
  • -email is unavailable- added by Stevan_White (Posted a comment)
  • -email is unavailable- added by dscorbett (Submitted the item)

    Do you think this task is very important?
    If so, you can add your encouragement to it.
    This task has 0 encouragements so far.

    Only logged-in users can vote.




    Follow 2 latest changes.

    Date Changed by Updated Field Previous Value => Replaced by
    2017-11-10 Stevan_White StatusNone => Fix posted
        Assigned toNone => Stevan_White

    Back to the top

    Powered by Savane 3.4