Re: [dita] Proposed review of DITA 2.0 elements to LwDITA components

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE reference PUBLIC "-//OASIS//DTD DITA Reference//EN" "reference.dtd">
<reference id="conref-processing-expectations">
    <title>Processing expectations for element-reference topics</title>
    <shortdesc>This topic contains "Processing expectations" sections for certain element-reference
        topics. These element-reference topics are for elements that exist in both DITA and
        LwDITA.</shortdesc>
    <refbody>
        <section id="data">
            <title>Processing expectations</title>
            <p>By default, processors <term outputclass="RFC-2119">SHOULD</term> treat
                    <xmlelement>data</xmlelement> elements as unknown metadata; the contents of
                    <xmlelement>data</xmlelement> elements <term outputclass="RFC-2119">SHOULD
                    NOT</term> be rendered.</p>
            <p>Processors that recognize a particular <xmlelement>data</xmlelement> element <term
                    outputclass="RFC-2119">MAY</term> make use of that element to trigger
                specialized rendering.</p>
        </section>
        <section id="desc">
            <title>Processing expectations</title>
            <p>When used in conjunction with <xmlelement>fig</xmlelement>
                <ph props="full-dita"> or <xmlelement>table</xmlelement>
                </ph>elements, processors <term outputclass="RFC-2119">SHOULD</term> consider the
                content of <xmlelement>desc</xmlelement> elements to be part of the content
                flow.</p>
            <p>When used in conjunction with <xmlelement>xref</xmlelement><ph props="full-dita"> or
                        <xmlelement>link</xmlelement></ph> elements, processors <term
                    outputclass="RFC-2119">MAY</term> choose to render the content of
                    <xmlelement>desc</xmlelement> elements as hover help.</p>
        </section>
        <section id="fn">
            <title>Processing expectations</title>
            <p>The two footnote types typically produce different types of output:</p>
            <dl>
                <dlentry>
                    <dt>Single-use footnote</dt>
                    <dd>When rendered, a superscript symbol (numeral or character) is produced at
                        the location of the <xmlelement>fn</xmlelement> element. The superscript
                        symbol is hyperlinked to the content of the footnote, which is placed at the
                        bottom of a PDF page or the end of an online article. The superscript symbol
                        can be specified by the value of the <xmlatt>callout</xmlatt> attribute.
                        When no <xmlatt>callout</xmlatt> value is specified, footnotes are typically
                        numbered.</dd>
                </dlentry>
                <dlentry>
                    <dt>Use-by-reference footnote</dt>
                    <dd>Nothing is rendered at the location of the <xmlelement>fn</xmlelement>
                        element. The content of a use-by-reference footnote is only rendered when it
                        is referenced by an <xmlelement>xref</xmlelement> with the
                            <xmlatt>type</xmlatt> attribute set to "fn". If an
                            <xmlelement>xref</xmlelement> with the <xmlatt>type</xmlatt> attribute
                        set to "fn" is present, a superscript symbol is rendered at the location of
                        the <xmlelement>xref</xmlelement> element. Unless conref is used, the
                            <xmlelement>fn</xmlelement> and <xmlelement>xref</xmlelement> must be
                        located in the same topic.</dd>
                </dlentry>
            </dl>
            <p>However, the details of footnote processing and formatting are implementation
                dependent. For example, a tool that renders DITA as PDF might lack support for the
                    <xmlatt>callout</xmlatt> attribute, or footnotes might be collected as end notes
                for certain types of publications.</p>
        </section>
        <section id="image">
            <title>Processing expectations</title>
            <p>The image addressed by the <xmlatt>keyref</xmlatt> or <xmlatt>href</xmlatt> attribute
                on <xmlelement>image</xmlelement> typically is rendered in the main flow of the
                content.</p>
            <p>Processors <term outputclass="RFC-2119">SHOULD</term> scale images when values are
                provided for the <xmlatt>height</xmlatt> and <xmlatt>width</xmlatt> attributes. The
                following expectations apply:</p>
            <ul>
                <li>If a height value is specified and no width value is specified, processors <term
                        outputclass="RFC-2119">SHOULD</term> scale the width by the same factor as
                    the height.</li>
                <li>If a width value is specified and no height value is specified, processors <term
                        outputclass="RFC-2119">SHOULD</term> scale the height by the same factor as
                    the width.</li>
                <li>If both a height value and width value are specified, implementations <term
                        outputclass="RFC-2119">MAY</term> ignore one of the two values when they are
                    unable to scale to each direction using different factors.</li>
            </ul>
        </section>
        <section id="map">
            <title>Processing expectations</title>
            <p>When rendering a map, processors might make use of the relationships defined in the
                map to create a Table of Contents (TOC), aggregate topics into a PDF document, or
                create links between topics in the output.</p>
            <p>The <xmlelement>title</xmlelement> element can be used to provide a title for the
                map. In some scenarios the title is purely informational; it is present only as an
                aid to the author. In other scenarios, the title might be useful or even required.
                In a map referenced by another map, the title might be discarded as topics from the
                submap are aggregated into a larger publication.</p>
        </section>
        <section id="pre">
            <title>Processing expectations</title>
            <p>Processors <term outputclass="RFC-2119">SHOULD</term> preserve line the breaks and
                spaces that are present in a <xmlelement>pre</xmlelement> element.</p>
        </section>
        <section id="section">
            <title>Processing expectations</title>
            <p>Processors <term outputclass="RFC-2119">SHOULD</term> treat the presence of more than
                one <xmlelement>title</xmlelement> element in a <xmlelement>section</xmlelement>
                element as an error.</p>
        </section>
        <section id="shortdesc">
            <title>Processing expectations</title>
            <p>Processors <term outputclass="RFC-2119">SHOULD</term> render the content of the
                    <xmlelement>shortdesc</xmlelement> element as the initial paragraph of the
                topic.</p>
            <p props="full-dita">When processors generate link previews that are based on the map
                context, they <term outputclass="RFC-2119">SHOULD</term> use the content of the
                    <xmlelement>shortdesc</xmlelement> that is located in the map rather than the
                    <xmlelement>shortdesc</xmlelement> that is located in the DITA topic. However,
                processors <term outputclass="RFC-2119">SHOULD</term> use the content of the
                    <xmlelement>shortdesc</xmlelement> element in the DITA topic when they render
                the topic itself, unless the <xmlatt>copy-to</xmlatt> attribute is specified on the
                topic reference to the element.</p>
        </section>
    </refbody>
</reference>