canonicalizations

Canonicalization is the process of converting data into a standard, or canonical, form to enable reliable comparison, deduplication, indexing, or verification. A canonical form is a unique representation among a set of equivalent representations, so that two items are equal if and only if their canonical forms match.

In computing, canonicalization covers textual and binary data. Unicode normalization (NFC, NFD, NFKC, NFKD) converts characters to a standard sequence, facilitating comparisons across languages. File and path canonicalization resolves symbolic links, relative components, and case sensitivity to produce a single, absolute path. Data deduplication and indexing also rely on canonical forms.

In the web, URL canonicalization selects a preferred URL among duplicates arising from different schemes, ports, or trailing slashes. The link rel=canonical element communicates the preferred address to browsers and crawlers. XML canonicalization (C14N) defines a deterministic byte sequence for XML documents, enabling reliable digital signatures; variants include inclusive and exclusive modes that differ in how namespace nodes are treated.

In mathematics and computer science, canonical forms represent objects in a unique standard representation, such as the canonical form of a polynomial, a matrix, or a graph, to identify equivalence classes. In practice, canonicalization aims to improve interoperability and security, but can introduce risks if input normalization reveals or creates unintended behavior, making careful validation important.