Personally, I would take a more bare bones approach:

  1. acquire XML
  2. learn about its structure
  3. use XLST/XPath to distill XML to relational database
  4. index content of database with something like Solr
  5. use the results of #3 and #4 to create a searchable/browsable Web interface
  6. for extra credit, add functionality to the Web interface which goes beyond search
  7. go to Step #1; library work is never done

Put another way, I would stay away from some sort of "solution", because y'all will invariably: 1) spend a whole lot of time learning how to install, customize, and use the solution, and 2) spend a whole lot of time maintaining the solution.

--
Eric Morgan
University of Notre Dame