Ead Conversion Revision History

Revision 1.7.0, Feb 17, 2004

Added an option to generate a single container rather than Parent/ID attributes
Added a workaround to a bug in Java 1.4.2 which resulted in many input lines being ignored.

Revision 1.6, Jan 6, 2004

EAD XML files may be read and their container lists recomputed to form a new XML file
There is support for arbitrary container hierarchies. The tab delimited format changed slightly to permit them
A summary of container used is now part of the Abstract
There is full support for the [Title Box Folder] style of input
Numerous bug fixes

Revision 1.5.12, Dec 26, 2003

Revised the containers to use parent="tag" references to containers containing a id="tag" attribute. Moved the list of all containers at non-leaf levels to notes instead of containers.
Removed the option to omit boxes if they are unique since box-folder types are no longer used
Expanded folder ranges like ff3-7
Encoding to escape special xml chars has been moved earlier so that the speadsheet may now be marked up using xml entities. This addresses the problem that the markup for italics is lost when doing a cut/paste. It's still lost, but may be re-inserted in the spreadsheet form.
Completed support of input with title, box, and folder fields using inital indenting for structure
EadNormalize: line 138 null pointer bug fixed

Revision 1.5.10, Dec 17, 2003

Revised computation of container lists from the lowest levels box:folder values. Higher level containers omit folder detail if all folders in the box are represented, and lists and ranges are used for both folders and boxes.
Modified the Example.txt to better demonstrate some of the new options.
Added options to control xml generation. The options apply to files subsequently opened. They are:
- Nesting of unitdate within unittitle(default off)
- Omitting Boxes from Box:Folders if they are unique (default on)
- Generating xml entities for characters not ascii graphics (default on)
- Generating a skeletal EAD structure rather than just the DSC (default on)

Revision 1.5.8, Dec 4, 2003

Spreadsheet columns revised to handle arbitrarly deep heirarchies
Entities added for a few common publication entities from iso-pub.ent
Added handling of a "Part" line that is treated as "otherlevel" and is above Series
Added experimental handling of input with title, box, and folder fields using inital indenting for structure
Removed computation of upper level container contents temporarily
Formatting of the xml file cleaned up

Revision 1.5.6, Oct 26, 2003, unreleased

Revision 1.5.5, Oct 21, 2003, unreleased

Revision 1.5.4, Oct 17, 2003

Now accepts non-numeric series titles.
Thus the entry 'series IV Correspondence' has a title of 'IV: Correspondence' rather than ': IV Correspondence'.
A new series starts with with no default subseries or box number.
It used to take the subseries and box from the previous series.
Missing box or oversized folder id's are reported as errors.
Missing box or oversized folder titles default to the previous title.
Titles have leading '-' stripped.
Thus 'Box 3 - Correspondence' will have a title of 'Correspondence' rather than '- Correspondence'.

Revision 1.5.3, Oct15, 2003

Some characters not in unicode ISO-PUB block are not handled properly on cut/paste from a Word document into a file input to EadConversion.

Wider variety of input formats accepted, with some way of selecting one.
Parameterization of keywords for series, box, oversized folder.
Internationalization of keywords for series, box, oversized folder.
Normalization of dates and ranges of dates via ISO 8601.
Allow New input panes which may have text pasted into them directly, and allow editing of text and a way to re-convert it. This requires running in the unrestricted mode of Java Web Start. Currently the application has access only to your local system only if you permit it.