dita — archive

[Eliot Kimber <ekimber@innodata-isogen.com>]

At least with DITA 1.1, where we can't add attributes, a reasonably common pattern is this (using the base types, which would normally be specialized for a specific application): as opposed to: (That is, using a nested phrase to clearly separate the PCDATA content from the metadata and better control the order of occurrence of the text and the data.) A typical specialization would be to name the nested "ph" element "text", indicating that its purpose is to contain the text content of the outer phrase: Given the general utility of this pattern, it seems reasonable to provide the "text" specialization of "ph" as a base type so specializers don't have to repeatedly re-invent it. Note that in neither DTDs nor schemas is it possible to prevent the occurrence of PCDATA content before or after a subelement of a given element, although in schemas you can control the order of subelements within an element with a "mixed" content model. This is one reason why using the "text" pattern is generally better than just allowing the text. While from an abstract processing standpoint it shouldn't matter where the <data> elements go (their semantic doesn't change based on position and their existence shouldn't affect the processing of the base content), from an authoring standpoint it can be less confusing for authors to have clear markup rules (for example, authors don't have to worry about white space around the <data> elements). In addition, naive processors might assume, for example, that the <data> elements will be the first child node of their parents, which won't necessarily be the case without the "text" element (because a text node could precede the <data> element). Cheers, Eliot -- W. Eliot Kimber Professional Services Innodata Isogen 8500 N. Mopac, Suite 402 Austin, TX 78759 (214) 954-5198 ekimber@innodata-isogen.com www.innodata-isogen.com </data></data></data></data>