about summary refs log tree commit diff
path: root/app/lib/hashtag_normalizer.rb
diff options
context:
space:
mode:
Diffstat (limited to 'app/lib/hashtag_normalizer.rb')
-rw-r--r--app/lib/hashtag_normalizer.rb25
1 files changed, 25 insertions, 0 deletions
diff --git a/app/lib/hashtag_normalizer.rb b/app/lib/hashtag_normalizer.rb
new file mode 100644
index 000000000..c1f99e163
--- /dev/null
+++ b/app/lib/hashtag_normalizer.rb
@@ -0,0 +1,25 @@
+# frozen_string_literal: true
+
+class HashtagNormalizer
+  def normalize(str)
+    remove_invalid_characters(ascii_folding(lowercase(cjk_width(str))))
+  end
+
+  private
+
+  def remove_invalid_characters(str)
+    str.gsub(/[^[:alnum:]#{Tag::HASHTAG_SEPARATORS}]/, '')
+  end
+
+  def ascii_folding(str)
+    ASCIIFolding.new.fold(str)
+  end
+
+  def lowercase(str)
+    str.mb_chars.downcase.to_s
+  end
+
+  def cjk_width(str)
+    str.unicode_normalize(:nfkc)
+  end
+end