Solr 1.4 has Tika built in, so it can handle Word, PDF, etc files. I keep forgetting the basic submission line. Here it is -
$ cd apache-solr-1.4.0/example
$ curl "http://localhost:8983/solr/update/extract?literal.id=doc5&commit=true" \
--data-binary @/path/to/file.pdf -H 'Content-type:application/pdf'
Where doc5 is a unique id. commit=true commits immediately so you can search for it.
Without setting the content type, it will default to xml and fail with "missing content stream".