Add Unicode whitespace filters

lib/liquid/standardfilters.rb

@@ -36,6 +36,16 @@ module StandardFilters
       %r{<style.*?</style>}m,
     )
     STRIP_HTML_TAGS = /<.*?>/m
+    UNICODE_WHITESPACE_LEFT = /\A[[:space:]]+/
+    UNICODE_WHITESPACE_RIGHT = /[[:space:]]+\z/
+    UNICODE_WHITESPACE_EDGES = Regexp.union(UNICODE_WHITESPACE_LEFT, UNICODE_WHITESPACE_RIGHT)
+    UNICODE_WHITESPACE_RUNS = /[[:space:]]+/
+    private_constant(
+      :UNICODE_WHITESPACE_EDGES,
+      :UNICODE_WHITESPACE_LEFT,
+      :UNICODE_WHITESPACE_RIGHT,
+      :UNICODE_WHITESPACE_RUNS,
+    )
 
     class << self
       def try_coerce_encoding(input, encoding:)
@@ -315,6 +325,56 @@ def squish(input)
       Utils.to_s(input).strip.gsub(/\s+/, ' ')
     end
 
+    # @liquid_public_docs
+    # @liquid_type filter
+    # @liquid_category string
+    # @liquid_summary
+    #   Strips whitespace, including Unicode whitespace, from the left and right of a string.
+    # @liquid_syntax string | strip_whitespace
+    # @liquid_return [string]
+    def strip_whitespace(input)
+      input = Utils.to_s(input)
+      input.gsub(UNICODE_WHITESPACE_EDGES, ' ').strip
+    end
+
+    # @liquid_public_docs
+    # @liquid_type filter
+    # @liquid_category string
+    # @liquid_summary
+    #   Strips whitespace, including Unicode whitespace, from the left of a string.
+    # @liquid_syntax string | lstrip_whitespace
+    # @liquid_return [string]
+    def lstrip_whitespace(input)
+      input = Utils.to_s(input)
+      input.gsub(UNICODE_WHITESPACE_LEFT, ' ').lstrip
+    end
+
+    # @liquid_public_docs
+    # @liquid_type filter
+    # @liquid_category string
+    # @liquid_summary
+    #   Strips whitespace, including Unicode whitespace, from the right of a string.
+    # @liquid_syntax string | rstrip_whitespace
+    # @liquid_return [string]
+    def rstrip_whitespace(input)
+      input = Utils.to_s(input)
+      input.gsub(UNICODE_WHITESPACE_RIGHT, ' ').rstrip
+    end
+
+    # @liquid_public_docs
+    # @liquid_type filter
+    # @liquid_category string
+    # @liquid_summary
+    #   Strips whitespace, including Unicode whitespace, from a string and collapses consecutive whitespace
+    #   to a single ASCII space.
+    # @liquid_syntax string | squish_whitespace
+    # @liquid_return [string]
+    def squish_whitespace(input)
+      return if input.nil?
+
+      Utils.to_s(input).gsub(UNICODE_WHITESPACE_RUNS, ' ').strip
+    end
+
     # @liquid_public_docs
     # @liquid_type filter
     # @liquid_category string

test/integration/standard_filter_test.rb

@@ -171,6 +171,19 @@ def test_squish_filter
     assert_equal("", Liquid::Template.parse('{{ " " | squish }}').render)
   end
 
+  def test_squish_whitespace_filter
+    unicode_spaces = "\u00A0\u202F\u2009\u2007"
+
+    assert_template_result(
+      "foo bar boo",
+      "{{ source | squish_whitespace }}",
+      { 'source' => "#{unicode_spaces}foo\u202F\u2009bar\t\n\u2007boo#{unicode_spaces}" },
+    )
+    assert_template_result("", "{{ nil | squish_whitespace }}")
+    assert_template_result("", "{{ source | squish_whitespace }}", { 'source' => unicode_spaces })
+    assert_template_result("\u200Bfoo\u200B", "{{ source | squish_whitespace }}", { 'source' => "\u200Bfoo\u200B" })
+  end
+
   def test_escape
     assert_equal('&lt;strong&gt;', @filters.escape('<strong>'))
     assert_equal('1', @filters.escape(1))
@@ -705,16 +718,52 @@ def test_strip
     assert_template_result('ab c', "{{ source | strip }}", { 'source' => " \tab c  \n \t" })
   end
 
+  def test_strip_whitespace
+    unicode_spaces = "\u00A0\u202F\u2009\u2007"
+
+    assert_template_result(
+      'ab c',
+      "{{ source | strip_whitespace }}",
+      { 'source' => "#{unicode_spaces}ab c#{unicode_spaces}" },
+    )
+    assert_template_result("a\u00A0b\u202Fc", "{{ source | strip_whitespace }}", { 'source' => "a\u00A0b\u202Fc" })
+    assert_template_result("\u200Bfoo\u200B", "{{ source | strip_whitespace }}", { 'source' => "\u200Bfoo\u200B" })
+  end
+
   def test_lstrip
     assert_template_result('ab c  ', "{{ source | lstrip }}", { 'source' => " ab c  " })
     assert_template_result("ab c  \n \t", "{{ source | lstrip }}", { 'source' => " \tab c  \n \t" })
   end
 
+  def test_lstrip_whitespace
+    unicode_spaces = "\u00A0\u202F\u2009\u2007"
+
+    assert_template_result(
+      "ab c#{unicode_spaces}",
+      "{{ source | lstrip_whitespace }}",
+      { 'source' => "#{unicode_spaces}ab c#{unicode_spaces}" },
+    )
+    assert_template_result("a\u00A0b\u202Fc", "{{ source | lstrip_whitespace }}", { 'source' => "a\u00A0b\u202Fc" })
+    assert_template_result("\u200Bfoo\u200B", "{{ source | lstrip_whitespace }}", { 'source' => "\u200Bfoo\u200B" })
+  end
+
   def test_rstrip
     assert_template_result(" ab c", "{{ source | rstrip }}", { 'source' => " ab c  " })
     assert_template_result(" \tab c", "{{ source | rstrip }}", { 'source' => " \tab c  \n \t" })
   end
 
+  def test_rstrip_whitespace
+    unicode_spaces = "\u00A0\u202F\u2009\u2007"
+
+    assert_template_result(
+      "#{unicode_spaces}ab c",
+      "{{ source | rstrip_whitespace }}",
+      { 'source' => "#{unicode_spaces}ab c#{unicode_spaces}" },
+    )
+    assert_template_result("a\u00A0b\u202Fc", "{{ source | rstrip_whitespace }}", { 'source' => "a\u00A0b\u202Fc" })
+    assert_template_result("\u200Bfoo\u200B", "{{ source | rstrip_whitespace }}", { 'source' => "\u200Bfoo\u200B" })
+  end
+
   def test_strip_newlines
     assert_template_result('abc', "{{ source | strip_newlines }}", { 'source' => "a\nb\nc" })
     assert_template_result('abc', "{{ source | strip_newlines }}", { 'source' => "a\r\nb\nc" })