On 2020-Jan-23, Tom Lane wrote:
> That particular case could be improved by stopping at a dash ... but
> since this code is also used to match strings like "A.M.", we can't
> just exclude punctuation in general. Breaking at whitespace seems
> like a reasonable compromise.
Yeah, and there are cases where dashes are used in names -- browsing
through glibc for example I quickly found Akan, for which the month names are:
mon "Sanda-<U0186>p<U025B>p<U0254>n";/
"Kwakwar-<U0186>gyefuo";/
"Eb<U0254>w-<U0186>benem";/
and so on. Even whitespace is problematic for some languages, such as Afan,
mon "Qunxa Garablu";/
"Naharsi Kudo";/
"Ciggilta Kudo";/
(etc) but I think whitespace-splitting is going to be more comprehensible
in the vast majority of cases, even if not perfect.
--
Álvaro Herrera https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services