Paul Eggert via tz wrote in
<[email protected]>:
|On 2026-03-07 01:38, Robert Elz wrote:
|> if there are really
|> many applications that read TZif files, aside from tzcode
|
|There are many, including the app I'm writing this email with. I daresay
|more apps read TZif via other means than via tzcode and derivatives.
It seems to me it uses ICU, and ICU says
// Check for TZ_ICU_MAGIC signature at file start. If we get a
// signature mismatch, it means we're trying to read a file which
// isn't a ICU-modified-zic-created zoneinfo file. Typically this
// means the user is passing in a "normal" zoneinfo directory, or
// a zoneinfo directory that is polluted with other files, or that
// the user passed in the wrong directory.
See how
https://github.com/unicode-org/icu/blob/main/icu4c/source/tools/tzcode/tz2icu.cpp
(cannot be read from Iran or North Korea or even Cuba i think, but
that shame aside) "builds transition tables" like grazy.
See comments like "why??".
They do perform value checks (i was only shortly glancing over).
I could imagine they would run to a carefully designed new
schematized variant rather sooner than later. (Especially if
there would be an easy scheme tester, say libucl of Vsevolod
Stakhov, both "well-known to famous":
https://github.com/vstakhov/libucl
Just my one cent.
--steffen
|
|Der Kragenbaer, The moon bear,
|der holt sich munter he cheerfully and one by one
|einen nach dem anderen runter wa.ks himself off
|(By Robert Gernhardt)
