Started this project by doing a rough analysis of the application as it currently is, and reading MK's original documentation. I've sent some requests for info to JJ. I then started looking at eXist documentation and figuring out what the best approach will be to the architecture of the application. The two choices are basically to store everything in the db, including controllers, binary data such as PDFs and images, XQuery, XSLT etc.; or storing only XML data in the db, and having controllers, transforms and binaries on the file system. The former is cleaner, and is what the other two eXist 1.5 projects have done, so it's favourite, but it will require careful separation of concerns and user management inside the db, because both JJ and research assistants manage data in the db, and we don't want logic files overwritten by mistake when uploading data.
On DR's instructions.
On late duty.
Heard from AR that he's now integrated support for the Lucene SnowballAnalyzer into eXist trunk, so I built a WAR from SVN to test this with. I've added lucene-snowball-2.9.2.jar (to match the Lucene version in trunk) to WEB-INF/lib, and I'm going to start investigating both this and the way controller.xql works, ahead of porting Map of London to the latest eXist.
Talked to HR, who's happy for me to start building the site structure so that she can fill in the content. Made a start on that today.
This proved more time-consuming and complicated than I thought, because of the proliferation of page-masters and sequences over the years. Some simplification wouldn't go amiss on the codebase.
Entered editor's proofing corrections to intro.
These things are fixed:
- Missing quotes in intro (due to markup using
<quote>instead of<q>by mistake). - First letter of a para in the intro chopped off by bad copy/paste.
- Running titles in intro (they were using Review titles rather than Article titles as desired).
- Position of running titles (they were not exactly lined up with the text margin).
- TOC page header (now says "Table of Contents" instead of journal title).
- Position of contributor index bookmark in PDF TOC (I had moved the actual contributor index in the document, but not adjusted the PDF bookmark).
Leaving early.
Got a new introductory article from JT, which has now been marked up and included, along with many changes to the title pages, pagination, and order of items.