Set an UTF-8 locale as default, not POSIX

Problem

The default locale of the node image is POSIX:

$ docker run --rm node:latest locale                                       
LANG=
LANGUAGE=
LC_CTYPE="POSIX"
LC_NUMERIC="POSIX"
LC_TIME="POSIX"
LC_COLLATE="POSIX"
LC_MONETARY="POSIX"
LC_MESSAGES="POSIX"
LC_PAPER="POSIX"
LC_NAME="POSIX"
LC_ADDRESS="POSIX"
LC_TELEPHONE="POSIX"
LC_MEASUREMENT="POSIX"
LC_IDENTIFICATION="POSIX"
LC_ALL=

node:19 has the following locales built-in:

$ docker run --rm node:19 locale -a
C
C.UTF-8
POSIX

The POSIX locale doesn't recognize UTF-8 (wc -m counts characters, not bytes).

$ docker run --rm node:latest bash -c 'echo -n 💩 | wc -m'
4
$ docker run --rm node:latest bash -c 'echo -n 💩 | LC_CTYPE=C.UTF-8 wc -m'
1

It is possible to manually enable C.UTF-8:

$ docker run --rm --env LC_CTYPE=C.UTF-8 node:latest bash -c 'echo -n 💩 | wc -m'
1

However in 2023, who doesn't need a locale supporting Unicode? ASCII is a bad default as it doesn't match a 2023 development environment.

Solution

Set LC_CTYPE=C.UTF-8 in the node image. Probably in file /etc/environment.

检查 Node image 如何配置 /etc/environment，然后运行所报告的命令 docker run --rm node:latest locale 和 locale -a，以确认可用的 locale。将默认的 LC_CTYPE 设置为 C.UTF-8，并验证 wc -m Unicode 示例在不传递环境覆盖的情况下报告一个字符。