I have a pile of documents I want to index.  My plan is add their text
content to a text field in a VS2005 express database, then use a full
text index to index the document contents.

In my particular situation I do not really need to store the actual
text content once the index has been updated with document data, as I
never need to see the text itself just the data in the remaining
columns, does anyone know if there is any way to remove the document
data from the database without messing up the index?

New documents will be added daily and old documents may be removed.

Thanks.