unicode-org / icu4x

Solving i18n for client-side and resource-constrained environments.
https://icu4x.unicode.org
Other
1.36k stars 174 forks source link

Use Rust instead of Python to generate UAX14 line break tables #758

Closed aethanyc closed 2 years ago

aethanyc commented 3 years ago

Currently, segmenter uses generate_properties.py to generate rule_table.rs and lb_define.rs, and we have to run the script manually to generate the table.

We may consider generate them by using Rust, and integrate the step into the build process.

aethanyc commented 3 years ago

Since the binary format for UAX14 and UAX29 rules will be similar, per discussion with @makotokato, we decide to do this after we have a prototype of UAX29. At that time, we'll know more about what the binary format will look like.

aethanyc commented 2 years ago

Duplicate of #1372 since #1372 has more information to move forward.