Defines a set of guards that can be used with Elixir functions to test whether a codepoint is a member of a Unicode set.

Each guard operates on a single integer codepoint since the operators permitted in a guard clause are restricted to simple comparisons that do not include string comparators.

The data that underpins these guards is generated from the Unicode character database and therefore includes a broad range of scripts well beyond the basic ASCII definitions.

Usage

Import or require the module and use the guards in a function clause when expression:

defmodule MyModule do
  import Unicode.Guards

  def character_type(codepoint) when is_upper(codepoint), do: :upper
  def character_type(codepoint) when is_lower(codepoint), do: :lower
  def character_type(codepoint) when is_digit(codepoint), do: :digit
  def character_type(_codepoint), do: :other
end

Implementation

The guards are defined as macros. Each expands, at its call site, to a membership test built as a balanced tree of comparisons over the ranges of its Unicode set. Some sets are large (is_graph/1 spans ~740 ranges), so prefer using a guard in a single function clause rather than repeating it at many when sites, since every use re-expands the membership test.

Because they are macros rather than defguard/1 definitions, they can be used in any when clause but cannot themselves be used to define another defguard.