Open nordzilla opened 3 years ago
@nordzilla Please assign this to an appropriate 0.6 sprint milestone.
@sffc Was there any plan to still implement this?
PR https://github.com/unicode-org/icu4x/pull/2913 has been open since December 2022 implementing some of this. I hope @nordzilla would pick it back up, but I believe he wouldn't mind some help to bring it over the finish line.
I can definitely help out wherever needed on this if no one is picking it back up. Might need a guidance on the ideal direction though.
Ok let's talk about a plan at the next team meeting!
Notes from 2024-06-20
chrono
has done part of that work. So it seems like a tzdb utility is worthwhile. Start at the base.chrono
does codegen, no?chrono_tz
is a whole crate. It seems like something that gets plugged directly into chrono
.
One of the formats we should be able to generate ICU4X TimeZone data from is the human-readable file format provided by the IANA TZDB
There are two ways we can go about this:
1) Build the
zic
C binary included in the IANA TZDB, and use it to generate TZif files. Then use our TZif capabilities from TODO(NUMBER) to generate the ICU4X data. 2) Write a custom parser to go directly from IANA files to ICU4X data.This will then match the combined
--tz-src-format iana
and--tz-src-path
options in thedatagen
crate (see TODO(Update datagen crate ... from file system))This will also match the
--tz-src-version
options in thedatagen
crate.For more context see
1009
Depends on:
998
999