Is it possible to have Paratext use ICU custom tailoring for sort order

Question

I’ve had a user ask about using an ICU custom tailoring sort order, one that is used in Fieldworks, in Paratext when it comes to creating the sort order for the Concordance.

Is this a something that could be added in the future.
At present, as far as I can tell, Paratext cannot handle the sort order necessary for this particular language.

Answer 1

Just to check that I undesrand what you mean: does your user want to have a different sort order in the Concordance than e.g. in the Word list tool?

Comments

Thank you for replying to my question. I asked the user and was told
this “The Word List tool should also use the ICU sort if possible,
although it is not essential as the Word List would never be printed and
I can use the “Find Word” field in the Word List to find what I need.”

MSEAIT/LT

---

As far as I know, the sort order is managed in language settings. Is there a reason, why reordering the order of Alphabetic Characters in the language settings of that particular language is out of question?

---

The Alphabetic Characters sort order in the Language Settings cannot
handle some of the complexity of the language in question.

MSEAIT/LT

---

Paratext 8.0 should support the ICU sort-order notation. To use it, just replace what is in the Alphabetic Characters tab of the Language Settings dialog with the ICU sort-order rules.

Note that this feature hasn’t been tested much and should be considered experimental.

---

Thank you anon291708, I just inserted the ICU sort order from FLEx into a PT8
version of the project and it worked, or at least it didn’t present the
same message Paratext 7.5 does about not accepting some of the
notations. I’ll have to ask the user to do the same, if they have PT8,
and see how the word list is sorted. As I understand it, the PT8
version of Concordance Builder is not yet out.

MSEAIT/LT

Answer 2

After using the ICU sort order in Paratext 8, I sent the Word List XML export to the user I’m helping and he said “That looks good. (There are a few weird things with half-syllables at the very bottom of the list, but that is a fault of the project having zero-width spaces in odd/wrong places, not the fault of the sort order).” So it seems like it is working. We’ll have to wait till the Concordance Tool is ready to know for sure…but this is encouraging.
Thank you,
MSEAIT/LT