I released ods2csv: https://hackage.haskell.org/package/ods2csv It extracts plain data from ODS/FODS documents without the need to install or run libreoffice, since it is based on a simple XML scanner. With "libreoffice --headless" or "unoconv" I had endless hassle to cope with running instances of LibreOffice and selecting a certain table. In contrast to that, ods2csv is lightweight and fast.