Wildcarding in work filter casts too wide a net.

[Philip Burns]

Martin writes:

There is a problem with the text filter. If you look for A01353, it retrieves all files that begin with 156, which is its decade.

[Philip Burns]

This is because the dynamic document indexer always uses the standard Lucene analyzer (tokenizer) when storing index entries. The standard analyzer breaks at hyphens.

We want to use the whitespace analyzer for file names. It is not clear at this point if this can be done.

[Craig Berry]

The easy way out was to make filenames that do not contain hyphens. This needed to be done for a number of reasons, but being able to search by filename without confusing the search engine is a welcome side effect.