DTBook To ZedAI Conversion Rules

This page gives the conversion mappings used in the DTBook-To-ZedAI module.

DTBook To ZedAI Conversion Rules

DTBook element: `a`

ZedAI element: ref

Dropped attributes: @type, @hreflang, @accesskey, @tabindex

Adjustments:

re-parent imggroup children
use inline samp
br to ln

Notes:

If dtbook:a@external = true, then we use @xlink:href instead of @ref.

DTBook element: `abbr`

ZedAI element: abbr @type='truncation'

Adjustments:

The following are rather meaningless in this context, so translate them as span. Would be nice to carry the original intent along, such as span role=’cite’`, but there aren’t roles for them all.

dfn becomes span
code becomes span
kbd becomes span
cite becomes span
sent becomes span role=”sentence”`
prodnote becomes extract

These are a little trickier as they (may) reference something else, but as there is no allowed child element in ZedAI which may contain a link or reference, that information is not carried along. This is an unlikely scenario.

q becomes span
a becomes span
annoref, noteref become span
re-parent imggroup children
use inline samp
remove br

DTBook element: `acronym`

ZedAI element: abbr

Notes:

@pronounce="no" becomes @type="initialism"
if missing @pronounce, becomes @type="initialism"
@pronounce="yes" becomes @type="acronym"

Adjustments:

The following are rather meaningless in this context, so translate them as span. Would be nice to carry the original intent along, such as span role=’cite’, but there aren’t roles for them all.

dfn becomes span
code becomes span
kbd becomes span
cite becomes span
sent becomes span
prodnote becomes extract

q becomes span
a becomes span
annoref, noteref become span
re-parent imggroup children
use inline samp
remove br

DTBook element: `address`

ZedAI element: address

Adjustments:

Normalize block/inline content model: if mixed, wrap inlines in p element.
br to ln

DTBook element: `annoref`

ZedAI element: annoref

Dropped attributes:

@type (“Type provides advisory content MIME type of the targeted id, see RFC1556.” –structguide)

DTBook element: `annotation`

ZedAI element: annotation

Adjustments:

Normalize block/inline content model: if mixed, wrap inlines in p element.
br to ln

DTBook element: `author`

ZedAI element: citation or p @property='author' @about='IDREF'

author element should be made into citation except when found inside poem/blockquote/cite, then use p @property/@about

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `bdo`

ZedAI element: span @its:dir="rtl|ltr"

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `blockquote`

ZedAI element: quote

DTBook element: `bodymatter`

ZedAI element: bodymatter

DTBook element: `book`

ZedAI element: body

DTBook element: `br`

ZedAI element: use ln to wrap lines

DTBook element: `bridgehead`

ZedAI element: hd

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `byline`

ZedAI element: citation or p @role="periodicals:byline"

Notes: for most book (non-article) use cases, byline can be citation. The exception would be anthologies, for which we can call upon the periodicals vocab and actually use "role = byline". For now, we will use just 1 vocabulary in this converter.

Adjustments:

Re-parent imggroup children
use inline samp
br to ln

DTBook element: `caption`

ZedAI element: caption

Adjustments:

Normalize block/inline content model: if mixed, wrap inlines in p element.
br to ln

DTBook element: `cite`

ZedAI element: citation

Adjustments:

title and author, are span instead of p
re-parent imggroup children
use inline samp
br to ln

DTBook element: `code`

ZedAI element: code

Adjustments:

Inline content model if only children are: em, strong, abbr, acronym, ln, sub, sup, span, bdo, pagenum, text
If any of the following occur, make all children block-level: dfn, code, samp, kbd, cite, a, img, imggroup, q, sent, w, prodnote, annoref, noteref
deal with br

DTBook element: `col`

ZedAI element: col

CSS: @align, @valign, @width (all to go in CSS)

DTBook element: `colgroup`

ZedAI element: colgroup

CSS: @align, @valign, @width

DTBook element: `covertitle`

ZedAI element: block @role='covertitle'

Adjustments:

Normalize block/inline content model: if mixed, wrap inlines in p element.
br to ln

DTBook element: `dateline`

ZedAI element: p @role="time"

Notes: Used p since ln is not suitable for all content models (e.g. level>dateline cannot be translated as section>ln)

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `dd`

ZedAI element: definition

Adjustments:

re-parent imggroup children
use inline samp * br to ln * Re-parent list, dl, div, poem, linegroup, table, sidebar * Re-parent or re-name p, dateline, epigraph, address, author, note

DTBook element: `dfn`

ZedAI element: term

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `div`

ZedAI element: block

Adjustments: Normalize block/inline content model: if mixed, wrap inlines in p element.

DTBook element: `dl`

ZedAI element: list

Adjustments: either wrap each term, definition in its own item element, or combine them into one item element

DTBook element: `docauthor`

ZedAI element: p @property="author"

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `doctitle`

ZedAI element: p @property="title" or h @property="title" when child of the frontmatter

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `dt`

ZedAI element: term

Adjustments:

The following are rather meaningless in this context, so translate them as span. Would be nice to carry the original intent along, such as span @role='cite', but there aren’t roles for them all.

dfn becomes span
code becomes span
kbd becomes span
cite becomes span
sent becomes span
prodnote becomes extract

q becomes span
a becomes span
annoref, noteref become span
inline samp
re-parent imggroup children
br to ln

DTBook element: `dtbook`

ZedAI element: document

Notes: Reference default ZedAI book profile and default vocabulary.

DTBook element: `epigraph`

ZedAI element: block @role="epigraph"

Adjustments:

Normalize block/inline content model: if mixed, wrap inlines in p element.
br to ln

DTBook element: `em`

ZedAI element: emph

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `frontmatter`

ZedAI element: frontmatter

Notes: A section is created automatically containing doctitle, docauthor, covertitle. The reason is that DTBook can have that info free-floating, whereas ZedAI must have only toc/section/biblio/glossary children and so we have to contain that info in a section element.

DTBook element: `h1`

ZedAI element: h

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `h2`

ZedAI element: h

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `h3`

ZedAI element: h

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `h4`

ZedAI element: h

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `h5`

ZedAI element: h

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `h6`

ZedAI element: h

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `hd`

ZedAI element: hd

Adjustments:

re-parent imggroup children
use inline samp
br to ln

DTBook element: `head`

ZedAI element: head

Notes:

ZedAI book profile is hardcoded
DTBook @profile is not preserved.

DTBook element: `img`

ZedAI element: object

Notes:

dtbook:@longdesc is a URI which resolves to a prodnote elsewhere the book

DTBook element: `imggroup`

ZedAI element: block (with @role='figure' if a caption is present)

Notes:

Look at DTBook multiple prodnote descriptions
There can also be multiple captions, each targeting multiple images

DTBook element: `kbd`

ZedAI element: code

Adjustments:

Same as dtb:code