Disable Tika to improve FITS performance

[escowles]

FYI, we were trying to figure out why FITS was taking so long, and noticed that it was running Tika, which in turn runs Tesseract OCR. This is incredibly slow, and really unnecessary to do characterization. Disabling Tika reduced FITS execution time for us from 90 seconds to 20.

Here's a snippet we added to our Ansible config to comment out Tika in the FITS config file:

- name: Disable Tika
  lineinfile:
    dest='{{ fits_install }}/fits-{{ fits_version }}/xml/fits.xml'
    regexp='.*(<.*edu.harvard.hul.ois.fits.tools.tika.TikaTool.*>).*'
    backrefs=yes
    line='<!-- \1 -->'
    state=present