[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Inclusion of fragments
I have an application (currently DB5 and XSLT2) which needs to generate some ancillary plaintext files along with the main transformation to HTML, LaTeX, and other targets. The main source files are all DB5 chapter instances. The text for the ancillary files is in separate small DB5 documents common to all of the main instances. At certain points in each of them, values or fragments of text from the main document are required to be inserted, in order to customise the output. For example, in one of the ancillary docs: <para>This is the README file for the X series, dated Y.</para> where the main document contains the desired text for X in /chapter/info/bibliomisc[@role='series'] location, and the desired text for Y in /chapter/@revision. Other inclusions are more extensive: the introduction of each ancillary file must be the whole contents of /chapter/info/abstract. The markup for the X and Y points has not yet been chosen: olink is under consideration. The targets don't need to be given as XPaths: the markup for the main document can also be changed. There are clearly various ways of doing this with entity references, XInclude, assorted href-and-pointer mechanisms, dyn:evaluate, and doubtless others; but I wondered if anyone has experience and advice on the neatest and simplest, bearing in mind that all the references in the ancillary files will be to locations in the main document already being processed, not in other external documents. ///Peter
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]