[erlang-questions] unicode in string literals

Richard Carlsson <>
Mon Jul 30 15:28:12 CEST 2012


On 07/30/2012 03:13 PM, Masklinn wrote:
> I'd expect the string manipulation functions of Erlang assume that as
> well (that strings are lists of "bytes"), don't they? E.g. that `words`
> splits on 0x20 (and maybe 0xA0), not on the {{Zs}} general category?

Yes, the old "string" module in the Erlang stdlib is not much use for 
working with Unicode strings. You should use something like the "ux" 
library (https://github.com/freeakk/ux) or Erlang bindings to ICU (can't 
seem to find the link, but I think there are more than one 
implementation of such bindings)

    /Richard




More information about the erlang-questions mailing list