There are issues with the search engine relating to both authors and addressees of correspondence. The drop-down lists are generated from distinct values of tags in the header. These tags, inherited from the Waterloo Script, contain plain text, and so the same individual is identified in a variety of different ways. It would be helpful if we could tag these names with ids from the personography, and then build our search engine drop-downs in a more intuitive fashion.
It seems best to start with the addressees, since they constitute a much smaller number (only 89 distinct values, listed below). The simplest approach would be this:
Following one of KSW's notes in this post, removed date tags from specific location in 17 files. This is presumably for consistency -- only 17 files had them -- and because I suspect some useful parsing can be done/is being done based on the first date in the text being the date the document was penned.
Items in the indexes have a link under their info popup which enables you to retrieve references to them in the correspondence, but sometimes there are no references (as in the case of peripheral bios, which are referred to in other bios but not in the actual correspondence). Previously, clicking on the "Mentions..." link simply did nothing in these cases, but I've now added a trap for this condition and an appropriate error message.
The Colonial Despatches is an XML database project which is creating a digital archive containing the original correspondence between the British Colonial Office and the colonies of Vancouver Island and British Columbia. The project lives at http://bcgenesis.uvic.ca, and the web application runs on the Pear dev Tomcat. The XML data is managed in SVN at http://revision.tapor.uvic.ca/svn/coldesp/.
|<< <||Current||> >>|