: If I understand you correctly, you really want to be constructing : SolrInputDocuments (not Lucene's Documents) and indexing those with : SolrJ. I don't think there is anything in the API that can read in an
I read your question differently then Otis did. My understanding is that you already have code that builds up files in the "<add><doc>..." update message syntax solr expects, but you want to modify those documents (wi/o changing your existing code) one possibility to think about is that instead of modifying the documents before sending them to Solr, you could write an UpdateProcessor tha runs direclty in Solr and gets access to those Documents after Solr has already parsed that XML (or even if the documents come from someplace else, like DIH, or a CSV file) and then make your changes. If Otis and i have *both* missunderstood your question, please clarify. -Hoss
