Dutch j-acute

[LettError]

Thank you for this resource!

I understand the best practice for the Dutch /J+acute /j+acute is to use the combining acute U+0301. This is also the method used by the iOS keyboards,.

So j + U+0301 and then swap the j for dotlessj to get rid of the dot.

This assumes that in Dutch any j following iacute is expected to be j-acute. This is not the always the case.

For instance the Dutch sentence Gaan we naar Fuji? Néé, we gaan naar Fíji! shows a context in which the j should not become j-acute.

[koeberlin]

Many thanks for your feedback, Erik!

So would you say a combination of ccmp (for getting rid of the dot) and mark (to place the acute) might be desirable, while the Durch locl code could be dropped?

[colinmford]

My approach to supporting the jacute is as follows, your mileage my vary. This at least fulfills Apple's requirements for "support" of Dutch. When the J/jacute is typed on a Dutch iOS keyboard (by holding on j and selecting it in the pop-up), it places j + U+0301 as Erik mentioned. Fonts without U+0301 now do not technically support Dutch according to Apple, and may not show up in font menus if the user's system language is set to Dutch.

1. Include J, j, dotlessj (U+0237, or no unicode, this glyph is purely to remove the dot and is not needed for Dutch support), and acutecmb (U+0301)
2. ccmp as follows:

feature ccmp {
    ...
    sub j' acutecmb by dotlessj;
    ...
} ccmp;

3. set up anchors for J, dotlessj, and acutecmb so that a mark feature can be generated for placement.

I don't use a languagesystem latn NLD set up or anything like that.

[colinmford]

@moyogo Yes, I typically also write the ccmp feature to accommodate other dotted letters, the code was just a sample for this particular situation.

[moyogo]

@colinmford Got it, my bad.
@koeberlin Unfortunately Glyphs only does this for i, j, but not for į, ị, ḭ, ɨ, і (i-cy), for example į́ in Navajo, ị́ in Igbo when tones are marked, ḭ́ in some African languages, ɨ́ in some American or African languages, і́ in Ukrainian when stress is marked, or other phonetic or theoretical ones like ᶖ, ɉ, ј (je-cy), ⁱ, etc. that all have the Unicode Soft_Dotted property.

[LettError]

Please do these things for living languages! Joost van den Vondel is quite dead ;)

[weiweihuanghuang]

@colinmford Got it, my bad. @koeberlin Unfortunately Glyphs only does this for i, j, but not for į, ị, ḭ, ɨ, і (i-cy), for example į́ in Navajo, ị́ in Igbo when tones are marked, ḭ́ in some African languages, ɨ́ in some American or African languages, і́ in Ukrainian when stress is marked, or other phonetic or theoretical ones like ᶖ, ɉ, ј (je-cy), ⁱ, etc. that all have the Unicode Soft_Dotted property.

Just to clarify does that mean something like į iogonek needs to be swapped for /dotlessi/ogonekcomb to make ı̨́ like:
sub iogonek' acutecomb by dotlessi ogonekcomb acutecomb;

[colinmford]

@weiweihuanghuang or if the ogonek has to be attached smoothly for the design if the iogonek (as some times has to occur) you would make a dotlessiogonek (or however you want to name that) glyph, then the feature would be something like

@soft_dotted = [i        j        iogonek];
@dotless     = [dotlessi dotlessj dotlessiogonek];

sub @soft_dotted' acutecmb by @dotless;

And expand those classes to include any other soft-dotted glyphs you might need, as @moyogo alludes to.