Hacker News new | ask | show | jobs
by foobar__ 4773 days ago
The fact that character ranges like [a-z] can depend on the value of LC_COLLATE is also something not many people are aware of.

  $ echo "ä" | LC_COLLATE=C grep '[a-z]'
  $ echo "ä" | LC_COLLATE=en_US.UTF-8 grep '[a-z]'
  ä
For common values of LC_COLLATE, the range [a-z] does not exclude accented characters and umlauts.