Consider adding optional PDF data extraction

[jqnatividad]

Either with pdftables or Tabula

[torfsen]

For extracting the fulltext and metadata from PDFs and other files we have developed ckanext-extractor. However, it only supports text, tabular data is not treated in a special way.