Processing Document-oriented XML

• Need to be able to work at 4 levels:

  • document level: to grab cross-references, number clause titles... often in a separate pass,
  • complex element level: tables, lists with internal references, chapter,
  • simple element processing: change a tag into an other tag (often adding the initial element as a class attribute),
  • within text: generate links from URLs, or from text elements, parse element text or attribute values.