MENU


Diacritical Marks

A diacritic is a mark added to a letter, such as the accent in é. Unicode can store é as the single precomposed character U+00E9, or as e followed by U+0301 COMBINING ACUTE ACCENT. Combining marks live mainly in the block U+0300 to U+036F (decimal 768 to 879); place one after any character with a numeric reference. Marks stack, which is how "Zalgo" text piles them above and below a letter.

Common combining marks (U+0300 block)
Mark Reference Result Mark Reference Result
Grave, acute à á à á Caron, cedilla č ç č ç
Circumflex, tilde â ñ â ñ Dot below, ring ạ å ạ å
Diaeresis ü ü Long stroke a̶ a̶
Stacking combining marks on base lettersHTMLLive
<p style="font: 22px Georgia, serif">
  Vie&#x323;&#x302;t Nam &middot; s&#x336;t&#x336;r&#x336;i&#x336;k&#x336;e&#x336; &middot;
  Z&#x351;&#x352;&#x353;a&#x300;&#x301;&#x302;&#x303;l&#x308;&#x30A;g&#x327;o&#x336;</p>
Browser output of Listing 2.54
Browser output of 54

The two encodings look identical but compare as different strings, so a search for "café" can miss "café". Normalize input before comparing or storing it: combined.normalize('NFC') composes, 'NFD' decomposes, and str.normalize('NFD').replace(/\p{M}/gu, '') turns "Crème brûlée" into the search key "Creme brulee". To count characters as a reader sees them, use Intl.Segmenter (Baseline since April 2024) rather than length. Strip marks only for search keys and slugs, never for display: in Vietnamese or Czech the mark changes the word.