While working on a project that frequently requires that we find out
various low-level facts about our MARC records (e.g. "How many records do
we have with 510 fields where indicator1 = '1'?"), we ended up writing a
small, unsophisticated, but so-far useful tool for extracting those records
into a Solr index.
All you need to run it on a machine is a Java Development Kit (JDK), Apache
Solr, and files containing the MARC records you want to query. See the
README for more details if you're interested: