Enomerous compile error with rime_deployer --compile

[hosiet]

• OS: Debian Sid/Debian 12
• librime version: 1.8.5+dfsg1-1 as seen at https://tracker.debian.org/pkg/librime
• rime-cantonese version: 62b4a51

Steps to reproduce

1. Initialize plum using curl -fsSL https://github.com/raw/rime/plum/master/rime-install | bash
2. Execute ./plum/rime-install cantonese
3. Execute cd ~/.config/ibus/rime/
4. Attempt to execute rime_deployer --compile ./jyut6ping3.schema.yaml , the following error message with tons of "Encode Failure" will appear:

[...]
W20230220 16:00:16.630759 45948 entry_collector.cc:147] Encode failure: '辂'.
W20230220 16:00:16.630774 45948 entry_collector.cc:147] Encode failure: '较'.
W20230220 16:00:16.630784 45948 entry_collector.cc:147] Encode failure: '辄'.
[...]

The full log exceeds the maximum line limit for GitHub issue. Please find the full output log at
error.txt

More information

This error does not appear with previous version of rime-cantonese and/or librime (1.7.x). I assume it shall either be a regression of rime-cantonese or librime.

Besides, I would like to know whether such "Encode failure" will affect the functionality of rime in any way.

[graphemecluster]

Did the error occur with the previous commit c1948ba?

[hosiet]

The same error will occur with commit c1948ba .

[tanxpyox]

These messages mean that the compiler couldn't find the corresponding code entry in the rime tables. I suspect this is because these simplified characters have not been cleaned from the corpus.

Besides, I would like to know whether such "Encode failure" will affect the functionality of rime in any way.

These are warnings, not errors - they wouldn't affect your actual usage.