Docutils | Overview | About | Users | Reference | Developers

Creating reStructuredText Interpreted Text Roles

Authors:

David Goodger

Contact:
docutils-develop@lists.sourceforge.net
Date:
$Date$
Revision:
$Revision$

Interpreted text roles are an extension mechanism for inline markup in reStructuredText. This document aims to make the creation of new roles as easy and understandable as possible.

Standard roles are described in reStructuredText Interpreted Text Roles. See the Interpreted Text section in the reStructuredText Markup Specification for syntax details.

Define the Role Function

The role function creates and returns inline elements (nodes) and does any additional processing required. Its signature is as follows:

def role_fn(name, rawtext, text, lineno, inliner,
            options=None, content=None):
    code...

# Optional function attributes for customization:
role_fn.options = ...
role_fn.content = ...

Function attributes are described below (see Specify Role Function Options and Content). The role function parameters are as follows:

Role functions return a tuple of two values:

Specify Role Function Options and Content

Function attributes are for customization, and are interpreted by the "role" directive. If unspecified, role function attributes are assumed to have the value None. Two function attributes are recognized:

Note that unlike directives, the "arguments" function attribute is not supported for role customization. Directive arguments are handled by the "role" directive itself.

Register the Role

If the role is a general-use addition to the Docutils core, it must be registered with the parser and language mappings added:

  1. Register the new role using the canonical name:

    from docutils.parsers.rst import roles
    roles.register_canonical_role(name, role_function)

    This code is normally placed immediately after the definition of the role function.

  2. Add an entry to the roles dictionary in docutils/parsers/rst/languages/en.py for the role, mapping the English name to the canonical name (both lowercase). Usually the English name and the canonical name are the same. Abbreviations and other aliases may also be added here.

  3. Update all the other language modules as well. For languages in which you are proficient, please add translations. For other languages, add the English role name plus "(translation required)".

If the role is application-specific, use the register_local_role function:

from docutils.parsers.rst import roles
roles.register_local_role(name, role_function)

Examples

For the most direct and accurate information, "Use the Source, Luke!". All standard roles are documented in reStructuredText Interpreted Text Roles, and the source code implementing them is located in the docutils/parsers/rst/roles.py module. Several representative roles are described below.

Generic Roles

Many roles simply wrap a given element around the text. There's a special helper function, register_generic_role, which generates a role function from the canonical role name and node class:

register_generic_role('emphasis', nodes.emphasis)

For the implementation of register_generic_role, see the docutils.parsers.rst.roles module.

RFC Reference Role

This role allows easy references to RFCs (Request For Comments documents) by automatically providing the base URL, http://www.faqs.org/rfcs/, and appending the RFC document itself (rfcXXXX.html, where XXXX is the RFC number). For example:

See :RFC:`2822` for information about email headers.

This is equivalent to:

See `RFC 2822`__ for information about email headers.

__ http://www.faqs.org/rfcs/rfc2822.html

Here is the implementation of the role:

def rfc_reference_role(role, rawtext, text, lineno, inliner,
                       options=None, content=None):
    if "#" in text:
        rfcnum, section = utils.unescape(text).split("#", 1)
    else:
        rfcnum, section  = utils.unescape(text), None
    try:
        rfcnum = int(rfcnum)
        if rfcnum < 1:
            raise ValueError
    except ValueError:
        msg = inliner.reporter.error(
            'RFC number must be a number greater than or equal to 1; '
            '"%s" is invalid.' % text, line=lineno)
        prb = inliner.problematic(rawtext, rawtext, msg)
        return [prb], [msg]
    # Base URL mainly used by inliner.rfc_reference, so this is correct:
    ref = inliner.document.settings.rfc_base_url + inliner.rfc_url % rfcnum
    if section is not None:
        ref += "#"+section
    options = normalize_role_options(options)
    node = nodes.reference(rawtext, 'RFC ' + str(rfcnum), refuri=ref,
                           **options)
    return [node], []

register_canonical_role('rfc-reference', rfc_reference_role)

Noteworthy in the code above are:

  1. The interpreted text itself should contain the RFC number. The try clause verifies by converting it to an integer. If the conversion fails, the except clause is executed: a system message is generated, the entire interpreted text construct (in rawtext) is wrapped in a problematic node (linked to the system message), and the two are returned.

  2. The RFC reference itself is constructed from a stock URI, set as the "refuri" attribute of a "reference" element.

  3. The options function parameter, a dictionary, may contain a "class" customization attribute; it is interpreted and replaced with a "classes" attribute by the set_classes() function. The resulting "classes" attribute is passed through to the "reference" element node constructor.