CDATA

CDATA stands for Character Data. In SGML-derived languages such as XML (including XHTML), a CDATA section is a portion of an element in which the text is treated as literal character data rather than markup. This allows content that would normally be interpreted as tags or entities to be included directly without escaping.

A CDATA section begins with <![CDATA[ and ends with ]]>. The text between these markers is not parsed for markup; characters like <, >, and & are ordinary characters rather than the start of tags or entities. However, the sequence ]]> is not allowed inside a CDATA section, because it would prematurely terminate the section. If that exact sequence must appear, the content must be split into multiple CDATA blocks or escaped.

CDATA sections cannot appear inside attribute values; they are valid only in element content. They do not cause entities within the section to be expanded, since all content is treated as literal text. CDATA is a mechanism for simplifying the inclusion of code snippets, fragments of HTML, or other data that contains markup-like characters, without the need to escape them.

In practice, CDATA is primarily used in XML-based formats. In plain HTML, CDATA sections are not part of the HTML syntax, though they may appear in XHTML documents as XML constructs. Overall, CDATA provides a way to include raw text without triggering the XML parser’s markup rules, while still conforming to the document’s encoding.