RML generally follows the XML syntax for HTML. In particular this means that all tags must be closed. Further, it is assumed that documents are encoded in UTF-8. XML namespaces are not supported.

RML documents consist of a tree of nodes, see the documents page for their general structure. The following describes some features beyond the common tree parsing behavior. For the most part, the syntax is equivalent to a subset of XML and HTML, with some additions.

Character references

RML supports character references in attributes and in data (text) encountered inside tags. When encountered, they will be translated into the specified Unicode code point.

Named character references

Numerical character references

Characters can be specified numerically based on their Unicode code points, just like in HTML.

For example, the euro sign € can be written using decimal form € or equivalently in hexadecimal form €.

Comments

Comments are supported using the familiar format.

<!-- My comment -->

Comments can also include tags, which will be ignored.

CData sections

CData sections can be used to include text that should not be parsed as RML. The section is started by <![CDATA[, followed by some data, and finally ended by ]]>.

Example

<p>A strange face appeared:
<![CDATA[
	'</"O"--"O"\>'
]]></p>

CData tags

When a CData tag is encountered in a document, all subsequent text is considered raw data until its end tag is encountered.

Structural data attributes

When a node with a structural data attribute is encountered, then all descendant RML nodes and data will be treated as a single data child of the structural node.

Note that a valid subtree is still required, and that the subtree will be considered when finding the node’s end tag.

Example

In the following, data bindings are used to dynamically display a list of elements. During RML parsing, only a single node is created, for the div tag. All its inner contents, including the h1 and p tags, are submitted as a single data child.

<div data-for="invader : invaders">
	<h1>{{ invader.name }}</h1>
	<p>Invader {{it_index + 1}} of {{ invaders.size }}.</p>
</div>

This allows the data-for view to store its inner contents without generating them immediately, so that it can later dynamically create a list of the provided elements as needed.