You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
If you want a DUCET collator, you can get it from class UCA. But you can also get a cached version from WriteCollationData.
If you want a CLDR collator, you need to call getCollator(type) in class WriteCollationData. That code works with class UCA. In particular, when we were still reordering characters for CLDR (before PR #762), the Remap class for doing so was in class UCA.
This seems misplaced, like an unnecessary back-and-forth. It seems like getCollator(type) should be in class UCA, and WriteCollationData should focus on, well, writing collation data.
A possible alternative is to create a separate class for making a CLDR collator, but there does not seem to be enough special code that we would need that.
So, try to move getCollator(type) from class WriteCollationData to class UCA.
On Wed, May 1, 2024 at 12:52 PM Markus Scherer ***@***.***> wrote:
If you want a DUCET collator, you can get it from class UCA. But you can
also get a cached version from WriteCollationData.
If you want a CLDR collator, you need to call getCollator(type) in class
WriteCollationData. That code works with class UCA. In particular, when we
were still reordering characters for CLDR, the Remap class for doing so was
in class UCA.
This seems misplaced, like an unnecessary back-and-forth. It seems like
getCollator(type) should be in class UCA, and WriteCollationData should
focus on, well, *writing* collation data.
A possible alternative is to create a separate class for making a CLDR
collator, but there does not seem to be enough special code that we would
need that.
So, try to move getCollator(type) from class WriteCollationData to class
UCA.
@macchiati <https://github.com/macchiati> FYI
—
Reply to this email directly, view it on GitHub
<#793>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACJLEMBVB7LTOQMRNBXODXDZAFBYPAVCNFSM6AAAAABHCPCY7KVHI2DSMVQWIX3LMV43ASLTON2WKOZSGI3TIMBZGMZTQNI>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
If you want a DUCET collator, you can get it from class UCA. But you can also get a cached version from WriteCollationData.
If you want a CLDR collator, you need to call getCollator(type) in class WriteCollationData. That code works with class UCA. In particular, when we were still reordering characters for CLDR (before PR #762), the Remap class for doing so was in class UCA.
This seems misplaced, like an unnecessary back-and-forth. It seems like getCollator(type) should be in class UCA, and WriteCollationData should focus on, well, writing collation data.
A possible alternative is to create a separate class for making a CLDR collator, but there does not seem to be enough special code that we would need that.
So, try to move getCollator(type) from class WriteCollationData to class UCA.
@macchiati FYI
The text was updated successfully, but these errors were encountered: