Standardized Variation Sequences

Unicode deals with characters, not glyphs. A single character may have numerous different graphical representations, but since all these glyphs still represent the same underlying entity with the same semantics they are not assigned to different code points (unless it is required for compatibility with another standard, of course). Higher‐level protocols can be used to force a particular rendering on a piece of text if the author so desires.

However, language isn’t always that straightforward and the boundary between character and glyphic variant can be blurry sometimes. It can be useful or even necessary to differentiate between variant forms of a character in some contexts while in other contexts these same differences may be completely unsubstantial. In such edge cases, variation sequences are employed to serve both scenarios.

A variation sequence consists of some base character, immediately followed by an invisible, zero‐width combining mark called a variation selector. The presence of the variation selector restricts the intended appearance of the base character to one particular subset of all its possible graphical representations. This means that applications that care about these specific differences know to apply a distinct glyph to the base character while other applications can safely ignore the variation selector and display the base character however they see fit. The underlying character remains the same, after all. The meaning doesn’t change; only the presentation does.

Standardized variation sequences – the subject of this page – make use of the generic variation selectors 1 through 14 (U+FE00–U+FE0D) and the Mongolian free variation selectors (U+180B–U+180D, U+180F). They can be defined for pretty much any purpose. There also exist emoji variation sequences which use variation selectors 15 and 16 (U+FE0E, U+FE0F) to select either the text‐style or emoji‐style rendering of an appropriate base character, and ideographic variation sequences which use variation selectors 17 through 256 (U+E0100–U+E01EF) and are exclusive to CJK Unified Ideographs.

Only variation sequences explicitly defined as part of the Unicode Character Database, Unicode Emoji, or the Ideographic Variation Database are valid; a variation selector in any other context must not affect the glyphic appearance of the character it applies to.

Some standardized variants apply only to select positional forms of a base character (cf. Cursive Joining). This page uses little boxes labelled “ZWJ” to mark the cursive joining points of these variation sequences. For example, a character flanked on either side by a ZWJ is shown in its medial form.

For the sequences corresponding to CJK Compatibility Ideographs, the respective ideograph is shown in grey alongside the variation sequence. A well‐made font will display them with identical glyphs. These sequences are the preferred representation because compatibility ideographs are canonically equivalent to unified ideographs, meaning the distinction is irrevocably lost in normalised text.

See The Unicode Standard, section 23.4: Variation Selectors for more information.

Source: StandardizedVariants.txt