Tagging Multiple Versions

Processing Alternatives

The <alternatives> element holds multiple, logically equivalent (substitutable) versions of the same information object. The typical case is a graphic (<graphic>) which is provided in multiple formats (such as a TIF, a JPEG, and an SVG image). These various formats offer processing alternatives for use by production personnel or by software in presenting the work to the user. For example, a high-resolution TIF image may be included and designated for use in printing while the same image, provided as a lower-resolution JPEG file, may be designated for use in web display.
Other potential use cases include:
  • An equation (<disp-formula> or <inline-formula>) that is available as an image as well as both TeX- and MathML-tagged versions;
  • A video for online display and a thumbnail that represents one image from the video for use in print versions of the work (<media>); or
  • A table (<table-wrap>) for which both XHTML-inspired table tagging and an image have been provided.
The <alternatives> element can be used everywhere that <graphic> and <media> are allowed. For example, <alternatives> can be used inside a <fig> as part of the large group of options that can be inside a Figure. (See Tagging Figures.) This example shows a single figure (“Figure 3”) with four processing alternatives—3 graphics in different image formats and one media file:
...
<fig id="Fig12a">
  <label>Figure 3</label>
  <caption>
    <title>Large Poodles</title>
  </caption>
  <alternatives>
    <graphic xlink:href="poodle12.tif"/>
    <graphic xlink:href="poodle12.jpeg"/>
    <graphic xlink:href="poodle12.gif"/>
    <media mimetype="xyz" xlink:href="poodle-jump12"/>
  </alternatives>
</fig>
...
Similarly, a table (<table-wrap>) may be provided in multiple formats: for example, the table below has three alternative formats: an XML-tagged table, the equivalent image file, and a textual version with the tabular look created by lines and spacing.
...
<table-wrap>
  <object-id>...a DOI...</object-id>
  <label>Table 6.</label>
  <alternatives>
    <table frame="box" rules="all" cellpadding="5">...an XHTML-inspired formatted table...</table>
    <preformat>...a tabular form with spaces and tabs...</preformat>
    <graphic xlink:href="tab437.jpg">...a JPEG of the same table...</graphic>
  </alternatives>
  <table-wrap-foot>...</table-wrap-foot>
  <attrib>...</attrib>
  <permissions>
    <copyright-statement>...</copyright-statement>
  </permissions>
</table-wrap>
...

Alternatives in Mixed-content

Within elements that need to contain alternatives but have mixed content rather than element content models (<disp-formula>, <chem-struct>), there may be an alternative that contains just text characters. For example, the alternatives may be an equation in MathML, a JPEG image of the equation, and a plain text equation for searching. The <textual-form> element can be used inside <alternatives> to hold such plain text version alternatives. For example, a <textual-form> element might contain an equation such as a + b = c as an alternative to the same expression tagged as MathML.

Audience or Use Alternatives

The <alternatives> element names processing alternatives for a single display object such as a graphic or a table. It does not, however, accommodate versions of semantically equivalent material intended for different audiences (for patients versus for doctors). Typical cases of this type of multiple versions include:
  • Two versions of a section, a short one for print to save paper and a significantly longer one for online display;
  • A paragraph that is only in the online version and not in the print version of the same article;
  • Two or more versions of a section of semantically equivalent material prepared at different levels for different audiences (top secret versus secret versus publicly disseminated or doctor/nurses/patients); or
  • Alternative versions of a <boxed-text> for different print products (one for the magazine and one for the journal).
These textual/semantic alternatives are very different from the case of multiple formats of the same graphic, in part because there is a structural difference; a given section does or does not contain this paragraph. There is also a slight difference in complexity between this case and that of the graphic in that, for the multiple graphic formats, software can pick one alternative, whereas, for the structure, the software must build a different structural tree. While this Tag Suite does not support alternate trees, it does include an attribute (@specific-use) to record these distinctions.
The @specific-use attribute can be used to record that certain paragraphs are for print-only:
<p specific-use="print-only">See the latest updates on our website:
http://www.mulberrytech.com</p>
The @specific-use attribute can be placed on all the block-level structures, that is, on paragraphs, figures, sections, and all the block display objects. There is no recommended set of values, but publishers have used “print-only”, “web-only”, “online-only”, “voice-only”, “specific-product-name”, etc.