Just a follow up to my own message on a sample/test EAD record. I
discovered that tools like oXygen allow you to create instance
documents based on schema [1]. This seemed like it would be just the
thing I needed, so I tried it out. It turned out that allowing for
every variation of use created a file of over 650MB when I stopped the
process. Clearly not something I could use.

Anyone have an idea how to make oxygen or another product create
something more reasonable? Other ideas on how to get a rather
thorough, yet reasonable testing sample EAD?

Where I think this may leave me is taking an existing document we've
created from Archivists' and editing the content into a form I could
start to use for automated testing. From there I could rely on
exception driven development [2] to discover when what I've created
does not work for perfectly valid documents.

Jason

[1] http://www.oxygenxml.com/doc/ug-oxygen/topics/xml-schema-instance-generator.html
[2] http://www.codinghorror.com/blog/2009/04/exception-driven-development.html

On Mon, Nov 15, 2010 at 8:56 AM, Jason Ronallo <[log in to unmask]> wrote:
> Hi,
> I'm working on a little EAD parser library that I may use in future
> projects. I'd like to be able to handle as many different allowable
> variations on element use as I can. I'm wanting to set up automated
> testing to see to what extent I can meet that goal. The EAD Tag
> Library has been useful in developing it as far as I have.
>
> It would be helpful to have a test record which exploits all (or most
> of) the elements in all of their allowable contexts. [1]
>
> This doesn't have to be a real example, but can contain dummy data
> that can be tested against. So the text of a unittitle could simply be
> "unittitle". Also repeatable elements within a particular context
> would only have to be a list of two to be testable.
>
> Does anyone know of a resource like this?
>
> Jason
>
> [1] Elements used mostly within mixed content would not need to be
> present for my current use case.
>