DC Code XML

This repository contains the Code of the District of Columbia in XML format.

The DC Code in XML makes use of a custom schema that draws on ideas from Akoma Ntoso and United States Legislative Markup.

The schema can almost entirely be seen in the following four examples.

(1) Root element. The root element is a <level> element. It contains some metadata and nested <level> elements for the first level divisions of the code.

<level type="document">
	<heading>Code of the District of Columbia</heading>

	<meta>
		<recency>
			<last-act>DC Act 20-101</last-act>
			<effective-date>2014-09-09</effective-date>
			<publication-date>2014-09-24T17:03:41-04:00</publication-date>
		</recency>
	</meta>

	<!-- <level> elements as shown below -->

</level>

(2) "Big" levels that appear in the Table of Contents (toc). These are divisions, titles, chapters, and so on above the level of a section.

<level type="toc">
	<prefix>Title</prefix>
	<num>1</num>
	<heading>Government Organization</heading>

	<!-- other <level> elements nested here -->

</level>

(3) Sections and annotations.

<level type="section">
	<num>2-531</num>
	<heading>Public policy.</heading>

	<text>The public policy of the District of Columbia is that all persons are entitled
	to full and complete information regarding the affairs of government and the official
	acts of those who represent them as public officials and employees. To that end,
	provisions of this subchapter shall be construed with the view toward expansion of public
	access and the minimization of costs and time delays to persons requesting information.</text>

	<level type="annotations">
	  <level>
	    <heading>History</heading>
	    <text>(Oct. 21, 1968, Pub. L. 90-614, title II, ...</text>
	  </level>
	</level>
</level>

(4) "Little" levels which are rendered as paragraphs (often numbered).

<level>
	<num>(c)</num>
	<heading>United States</heading>

	<text>The United States, its officers, employees, and agents, and its agencies shall not:</text>

	<level>
		<num>(1)</num>
		<text>Be responsible for the payment of any judgments, liabilities or costs resulting
		from any action or proceeding against the District of Columbia or its agencies,
		officers, employees, or agents;</text>
	</level>

	<!-- other <level> elements nested here -->
</level>

XIncludes

The Code is one logical XML document (approximately 75 megabytes), but it has been broken apart into approximately 26,000 smaller files to make it more manageable. The Code begins in index.xml and XInclude tags stitch all of the documents together. All XIncludes target <level> elements at or above the section level. We have split the code at titles, chapters, subchapters, and so on down to sections.

Each XInclude element has an href attribute with a relative path to the file that is logically included at that point.

The top-level index.xml file looks like this:

<level type="document">
  <heading>Code of the District of Columbia</heading>
  <meta>
    ...
  </meta>
  <level type="toc">
    <prefix>Division</prefix>
    <num>I</num>
    <heading>Government of District</heading>
    <ns0:include xmlns:ns0="http://www.w3.org/2001/XInclude" href="Title-1/index.xml"/>
    <ns0:include xmlns:ns0="http://www.w3.org/2001/XInclude" href="Title-2/index.xml"/>
    <ns0:include xmlns:ns0="http://www.w3.org/2001/XInclude" href="Title-3/index.xml"/>
    ...
  </level>
  <level type="toc">
    <prefix>Division</prefix>
    <num>II</num>
    <heading>Judiciary and Judicial Procedure</heading>
    <ns0:include xmlns:ns0="http://www.w3.org/2001/XInclude" href="Title-11/index.xml"/>
    <ns0:include xmlns:ns0="http://www.w3.org/2001/XInclude" href="Title-12/index.xml"/>
    ...
</level>

The namespace prefix ns0 is subject to change.

Document Element

The root element is a <level>, like most other nodes in the schema, with type="document". We are not using namespaces.

This document element contains a <meta> element with document metadata.

The <meta> element is followed by <level> elements representing the Divisions of the Code. In the actual root file index.xml in this repository some of the <level> elements shown below are replaced by XIncludes, but we show them replaced with their referenced content here to highlight the intended semantics.

<?xml version='1.0' encoding='utf-8'?>
<level type="document">

  <heading>Code of the District of Columbia</heading>

  <meta>
    ...
  </meta>

  <level type="toc">
    <prefix>Division</prefix>
    <num>I</num>
    <heading>Government of District</heading>

    <level type="toc">
      <prefix>Title</prefix>
      <num>1</num>
      <heading>Government Organization</heading>

      <level type="toc">
        <prefix>Chapter</prefix>
        <num>1</num>
        <heading>District of Columbia Government Development</heading>

        <level type="toc">
          <prefix>Subchapter</prefix>
          <num>I</num>
          <heading>District of Columbia Establishment</heading>

          <level type="section">
            <num>1-101</num>
            <heading>Territorial area.</heading>
            <text>The District of Columbia is that portion of the territory
            of the United States ceded by the State of Maryland for the
            permanent seat of government of the United States, including the
            river Potomac in its course through the District, and the islands
            therein.</text>

Metadata

The document's metadata describes the recency of the DC Code publication:

<recency>
  <last-act>DC Act 20-101</last-act>
  <effective-date>2014-09-09</effective-date>
  <publication-date>2014-09-24T17:03:41-04:00</publication-date>
</recency>

The last-act field is a textual field describing, in human-readable form, the last act that has been codified in the document. For corresponding machine-readable data, see the file update-chart.csv at the root of the repository.

The effective-date field contains an ISO 8601-formatted date corresponding to the effective date of the most recently codified act. The DC Code is expected to accurately reflect the law on this date.

The publication-date field contains an ISO 8601-formatted date indicating when the document was published by the DC Council. The publication date may be before or after the effective date.

Levels

All hierarchy of the Code is represented by <level> elements. This includes both "big" levels (titles, chapters, etc.), sections, and "little" levels (subsections, paragraphs, and so on).

Type Attribute

The type attribute indicates the type of the level, as follows:

document: The root of the entire document. There is one such element and it is at the top of the hierarchy.
toc: These are divisions, titles, chapters, and so on above the level of a section. toc is a mnemonic for Table of Contents. These levels appear in the Table of Contents of the code.
section: Sections of the code (e.g. § 1-101), the primary level at which text and annotations occur and what citations target.
placeholder: The level is section-like but it represents Code sections that no longer exist. They are documented specifically in a section below.
annotations: Annotation blocks.
appendices: Appendices to a section (see Title-16/Chapter-9/16-916.01.xml).
form: Blocks that are rendered with internal whitespace preserved. These elements contains quoted language that the Code is directing to be used on government forms.
When type is omitted, the level is a numbered paragraph (i.e. (1) The public policy....) or an unnumbered block with a heading. Inside annotations, subheadings are represented this way.

Level Numbering

A <level> element must contain at least one of <prefix>, <num>, or <heading>.

<prefix> is used only in type="toc" levels to provide the text that appears before the level's number.

The inner text of this element is one of Article, Chapter, Division, Part, Subchapter, Subdivision, Subpart, Subtitle, Title, and Unit.

<num> indicates the level's number.

For toc levels, this usually is a number, letter, or roman numeral.
For sections, the number is roughly in the form 1-101 (title, hyphen, section number). It never includes the §-symbol.
For "little" levels, this is the paragraph's numbering including any parenthesis around the numbering. An example would be (12A-i).
When <num> is omitted, the level is simply unnumbered. document and placeholder levels never have <num> elements. For placeholders, see below.

<heading> contains the level's name or heading.

For toc levels, sections, placeholders, and subheadings in annotations, this is the level's title or heading, such as Government Organization.
A heading can also appear on "little" levels and would appear as something like italicized text.

Content Elements

Following the <level>'s numbering elements described just above, a <level> must contain one or more <level> or <text> (see below) elements.

These elements must be rendered in document order.
For "little" levels, the level's <num> and/or <heading> are typically rendered inside and at the start of the first <text> paragraph.

Text

Text content is contained within <text> elements inside <level>s.

Each <text> element is a separate paragraph.
<text> may be interleaved with other <text> and <level> elements and must be rendered in document order.
<text> elements contain a subset of HTML.

This is an example of a typical <text> element:

<text>One <span style="font-style: italic; ">ex officio </span>Commissioner,
the Deputy Mayor for Planning and Economic Development;</text>

An optional class attribute can be one of:

centered: The paragraph is rendered with centered text alignment.

Inside the <text> element is HTML conforming to the following restrictions:

Only bare text, <span>, and <table> elements may appear.
<span> elements may have a style attribute containing only the following CSS:
font-weight: bold;
font-style: italic;
text-decoration: underline;
font-family: monospace;
<table> elements may contain thead, tr, th, and td elements. th and td elements may have colspan attributes (as in HTML), and their content is restricted to the same subset of HTML as for <text> elements. When a <table> element appears inside a <text> node, it is the only content in the the <text> node (because each <text> node represents a paragraph).

Placeholders

Placeholder levels are similar to levels with type="section" but represent Code sections that have been repealed, omitted, etc. Here is an example:

<level type="placeholder">
  <reason>Not funded</reason>
  <section>1-136.01</section>
  <heading>Establishment of the 51st State Commission.</heading>
  <text>[Not funded].</text>
  <level type="annotations">
    ...

These levels do not have a <num> element. The <heading> element typically contains the name of the section that was removed. These levels may also have these other child elements:

<reason>: The reason for the placeholder.
When present, one of Expired, Not funded, Omitted, Repealed, Reserved, and Transferred.
When not present, the reason for the placeholder is not indicated in the code.
<section>: The number of the omitted section (e.g. 1-136.01).

In place of <section>, a placeholder may instead contain:

<section-start>: The number of the first omitted section in a range of omitted sections.
<section-end>: The number of the last omitted section in a range of omitted sections.
<section-range-type>: This element contains the text range if all sections between start and end are omitted (usually rendered as "A to B") or list if the start and end name two sections but a range is not implied (usually rendered as "A, B").

DCCouncil/dc-code-prototype