Output depends on $LANG env variable

[javalsai]

Took me a LONG while to debug this, working properly on terminal but not on a hyprland binding...

Basically, unipicker will output properly the chosen character if the $LANG env variable is (apparently) set to an installed locale, if not (empty, c...), it will output an unknown sequence (appears to be only the first byte of the sequence).

^{rg is just like grep}

[javalsai]

Apparently, it's something about sed?

[javalsai]

My guess is that sed without LANG treats all text as simple byte sequences, meanwhile, when given an "advanced" LANG, it's able to parse UTF sequences and include the whole character in the match group, instead of just one byte.

EDIT: Even something like C.UTF-8 as LANG would work whilst being generic, it's the encoding what matters.