XSLT Reference

normalize-unicode()

XSLT 2.0 string function

Applies Unicode normalization (NFC, NFD, NFKC, NFKD, or FULLY-NORMALIZED) to a string, ensuring a canonical character representation.

Syntax

normalize-unicode(string, normalization-form?)

Description

normalize-unicode() converts a string to a specified Unicode normalization form. Different normalization forms control whether composed or decomposed character representations are used, and whether compatibility equivalents are collapsed.

The most common use case is ensuring consistent string comparison when data may come from different systems that represent the same character differently — for example, the letter é can be stored as a single precomposed codepoint (U+00E9) or as e followed by a combining accent (U+0065 U+0301).

Normalization forms:

Form	Name	Description
`NFC`	Canonical Decomposition + Canonical Composition	Precomposed form (default, most common)
`NFD`	Canonical Decomposition	Fully decomposed; base characters followed by combining marks
`NFKC`	Compatibility Decomposition + Canonical Composition	Collapses compatibility variants (e.g., ligatures, width variants)
`NFKD`	Compatibility Decomposition	Decomposed compatibility form
`FULLY-NORMALIZED`	W3C XML extension	NFC with additional normalization of initial combining marks

Parameters

Parameter	Type	Required	Description
`string`	xs:string?	Yes	The string to normalize.
`normalization-form`	xs:string	No	One of `NFC`, `NFD`, `NFKC`, `NFKD`, `FULLY-NORMALIZED`. Defaults to `NFC`.

Return value

xs:string — the input string in the requested normalization form. Returns "" if string is an empty sequence.

Examples

Normalizing to NFC for consistent comparison

Stylesheet:

<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet version="2.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
  <xsl:output method="xml" indent="yes"/>

  <xsl:template match="/strings">
    <normalized>
      <xsl:for-each select="s">
        <!-- Ensure NFC before comparison or storage -->
        <s><xsl:value-of select="normalize-unicode(., 'NFC')"/></s>
      </xsl:for-each>
    </normalized>
  </xsl:template>
</xsl:stylesheet>

Collapsing compatibility variants with NFKC

Input XML:

<?xml version="1.0" encoding="UTF-8"?>
<data>
  <!-- Contains ﬁ ligature (U+FB01) and ² superscript (U+00B2) -->
  <value>ﬁle²</value>
</data>

Stylesheet:

<?xml version="1.0" encoding="UTF-8"?>
<xsl:stylesheet version="2.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
  <xsl:output method="xml" indent="yes"/>

  <xsl:template match="/data">
    <result>
      <!-- NFKC: ﬁ → fi, ² → 2 -->
      <xsl:value-of select="normalize-unicode(value, 'NFKC')"/>
    </result>
  </xsl:template>
</xsl:stylesheet>

Output:

<result>file2</result>

Notes

NFC is the recommended normalization for most XML and web applications; it is the form used in HTML5 and most web APIs.
NFKC is useful for search and indexing where compatibility equivalents should be treated identically (e.g., full-width vs. half-width letters, ligatures).
NFD is mainly useful for low-level text processing or font rendering.
The normalization form argument is case-insensitive; "nfc" and "NFC" are equivalent.
If the argument is "" (empty string), the NFC form (default) is applied.