Ordinals (a new edition) (2018, 4th edition)

[dvdrw]

Not sure if this is a bug in Affinity Publisher 1.9.3. OSX 11.0.1. …

settings: Text styles / paragraph styles / figures / ordinals checked

text in paragraph: (a new edition) (2018, 4th edition)

result: 

 

I'm a bit surprised by the (a in superscript. Is this a bug or is there a workaround? Doing all the suffixes in the book by hand is rather time consuming 🤯
But may be I'm overlooking something 

 

 

[dvdrw]

and the (a is not the only issue here:

results in:

[walt.farrell]

When you enable the Ordinals feature, it applies to a and o as they are ordinals in Spanish and some other languages, just as st, nd, th, and rd are in English.

You should not enable them for a general set of text, but only for the actual words that have the ordinal characters in them. That is, they should be used in a Character text style, for isolated words, not in a Paragraph text nor for a long run of text.

I think there is room for improvement in this. For example, the o and a are ordinals only at the end of a word as far as I know, and so it would be nice to apply them as ordinals only in that position. Also, they are not ordinals in English, so English words could have them exempted even in a final position.

But I don't think it's a bug.

[dvdrw]

thanks @walt.farrell

I'll solve it the way you suggest it.

find en replace only the 1st, 2nd, 3th, 4th, 5th, 6th, 6th, 8th, 9th and replace it via format ordinals is the easiest pragmatic solution I think

[Wosven]

It seems like a bbug, since as you said they aren't dependent of language option or at a final position or following a digit/number, or roman numbers ( primero: 1°, secundo: 2°, tercero: 3°…).

I hope one day we'll be able to use regular expressions in paragraph styles for such things, since it can be tricky: 1er, 1ers, 1re, 1res, 2nd... 3e, 3es... but also XIe (siècle), IIIe, etc. And a word like "vie" (life), or with similar ending can be finish in superscripts by error if badly implemented.

I suppose each language get similar problems with automated superscripts. But being able to code/use regular expressions to add superscripts is time saver, I would have a hard time living without.

 

[kenmcd]

16 hours ago, dvdrw said:

 

I'm a bit surprised by the (a in superscript. Is this a bug or is there a workaround?

Looks like you are using the Helvetica Now font - correct?
It has some quirks in how it handles the context for ordinals in its OpenType feature code.
I tested the text "(a new house) visitor-orientation" in a font editor and got the same results as your images above.
Took a quick look at the feature code and I am not sure why they did what they did, but it does the substitutes in that context.

If you want to leave the Ordinal box checked (which turns-on the font's Ordinals OpenType feature) to have it work properly for most occurrences, for the exceptions you can insert a zero-width space to break the context between the (a and the -o so they will not be affected.
So if there are fewer of those occurring, that may be easier than turning-off the ordinals completely.

[walt.farrell]

31 minutes ago, LibreTraining said:

So if there are fewer of those occurring, that may be easier than turning-off the ordinals completely.

Though that would have other repercussions, such as with Spell Checking, wouldn't it?

[kenmcd]

21 hours ago, walt.farrell said:

Though that would have other repercussions, such as with Spell Checking, wouldn't it?

Yes, that is a valid concern when using a zero-width space (ZWS), but in this case I do not think it will affect anything.
They use a character class for the ordinals feature code which includes the expected 0-9 so ordinals like 1ª 2º etc. work as expected, but that character class also includes other characters such as punctuation and currency signs … which is why it is affecting his text following the parenleft and hyphenminus characters.
Since the fix is to insert the ZWS between these punctuation characters and the actual words, I do not think it will affect the spellcheck.