Le ven. 9 févr. 2024 à 18:30, Corentin Jabot <corentinjabot_at_[hidden]> a
écrit :

> I'm not able to quickly gauge how this change impacts the performance of
> NFC Quick Check, which would be the primary concern from a compiler
> perspective, but it seems fine.
>
The quickCheck algorithm described in Section 9 of UAX #15
<https://www.unicode.org/reports/tr15/#Detecting_Normalization_Forms>, much
like the normalization algorithm, does not change.
What the PRI background document is saying regarding the NFC_Quick_Check
property is that if, instead of using the values of the NFC_Quick_Check
property from the UCD, you are trying to compute the values yourself based
on the decompositions (don’t do that!), you might get it wrong (indeed, the
tools we use to derive those properties were getting it wrong until I fixed
them <https://github.com/unicode-org/unicodetools/pull/619>).

Note of course that you still have to go through a proper normalizer when
quickCheck says « Maybe », and if that normalizer is cleverly optimized, it
may turn out to be too clever for its own good (this is the case of the ICU
normalizer <https://unicode-org.atlassian.net/browse/ICU-22586> and the ICU4X
normalizer <https://github.com/unicode-org/icu4x/pull/4538>).