RefDB gets all reference data from text files (or from equivalent data read from stdin). These text files can be from various sources: you can enter them manually, you download files from search engines in the web, or you have data exported from a different database application. In almost any case, some kind of filtering has to be applied to those files in order to be valid input for RefDB.
The input filters currently shipped with RefDB are listed below. Please follow the links to the individual filters for further information.
A simple (maybe too simple) shell script to convert text files like RIS documents from DOS-style line endings to Unix-style line endings. Most RefDB tools need their input files with Unix-style line endings. This is a valuable tool to import reference databases from Windows reference managers.
A tool to convert Pubmed data in both the tagged and the XML format to RIS
A tool to convert EndNote "RIS" exported data to RIS
A tool to convert BibTeX data to RIS
A tool to convert reference data in DocBook SGML/XML documents to RIS
A tool to convert references in MARC format to RIS.
If you need other formats than those listed above, you'll either have to provide your own input filter or search the web for existing filters that convert your data to one of the supported formats. A decent set of filters is supplied by Chris Putnam's bibutils package, which interconverts bibtex, COPAC, EndNote (Refer), EndNote (XML), ISI, PubMed, MODS, and RIS data. For example, use the following command to import MODS datasets into RefDB:
#xml2ris modsdata.xml|refdbc -C addref