PDF text extracted with PDFMiner not saved into ca_sql_search_words
even having PDFMiner installed the PDF content is not indexed in the general search.
If I debug PDFMiner runs correctly, it extracts the words which are correctly saved inside the ca_media_content_locations table but not inside the ca_sql_search_words, while on another CA installation they are correctly saved inside both tables (and the search works correctly).
What could it be? The version of CA is 1.7.6