HTML attribute

An HTML attribute supplies information about an HTML element, usually through a name and a value in its start tag. Its meaning is defined for the relevant element and attribute. The source notation is converted into attribute data on the element in the Document Object Model.^[ref]

For example, <p id="notice" class="warning">Save</p> gives the paragraph the identifier notice and the class warning. CSS selectors #notice and .warning can then match it for different reasons. The text Save is element content, rather than another attribute value.

Some attributes use presence as their boolean signal. For an applicable HTML boolean attribute, disabled and disabled="false" both express presence; the latter spelling does not make it false. Removing the attribute expresses absence. Treating every value as a programming-language boolean would therefore misread the document.^[ref]

Attribute data and live interface properties are related by specified reflection rules, not by a universal identity. For some controls, an attribute describes a default while the current state can change through interaction. A JavaScript property access must consequently be interpreted using that interface's definition. This is particularly useful when debugging why the original source and the state a user sees no longer look identical.