The EadConversion program has been updated to Version 1.6. This is a proof of concept
program to read in-depth container lists in a textual form and produce an EAD XML file.
This release incorporates a number of improvements and new features, particularly the ability
to read the container list from an existing EAD XML file, analyze it, and generate an updated
XML file using parent/id references for containers as well as a summary of container use.
It is available as a Java Web Start application. You may have to bring up Java Web Start's
preferences and clear the Applications Folder to get version 1.6 if you have used earlier
versions. You can start it from either:
<http://agileimage.com/html/ead/EadConversion.jnlp>
or, if you are willing to trust the application with full access to the web and your local system,
<http://agileimage.com/html/ead/EadConversionFull.jnlp>
The ability to read an EAD XML file and generate a spreadsheet and an updated XML file now
permits editing to occur at any stage of the conversion process. XML markup can now be
added to titles in the spreadsheet form. EadConversion handles the parent/id container
relationships, but also tries to infer a 'box' or 'oversize' container for a 'folder' container from
ancestors, and failing that from the last one encountered. Input XML files should be valid EAD
files, not just well formed. An ead.dtd file from the agileimage.com web site will be, if
necessary, used during the processing.
EadConversion also adds a note in the upper levels of the component hierarchy with a concise
list of all the containers used by components below it. Since this list is computed from the
lowest level components, it may be used to help track down errors in the markup.
The intermediate spreadsheet has changed format slightly to accommodate an arbitrary
hierarchy of containers, although the old format is still accepted as input. This ability to use a
physical container structure like 'carton:folder:envelope' extends the idea of the 'box-folder'
type in a container. It also was the driving reason behind changing the generated XML to use
parent references to the id of enclosing containers since that permits representing an
arbitrary physical hierarchy.
The input format used at the University of Wyoming is now fully supported as well. The format
comes from cutting tables with columns for title, box, and folder and uses leading tabs in the
title column to denote the logical structure. Dates are at the end of the first line of the title
and are recognized as dates. Several examples of the different input formats are now included
in the Help menu.
A User Guide is available through the Help menu or from:
<http://agileimage.com/html/ead/EadConversionGuide.jnlp>
There are references to the issues of parent/id use on the home page for EadConversion at:
<http://agileimage.com/html/ead/>
I would be interested in examples of files that could not be converted properly as well as
some of those that were okay. The more examples I have, the more likely it becomes that
variants are handled properly.
I also would like to encourage discussion on this list server of the issues involved in
automatically converting textural container lists into EAD documents.
Enjoy,
Paul
|