IULRDC-201 RDC: adding stemming to solr

[rdlebeau]

Adding SnowballPorterFilterFactory to solr search indexing.
On localhost, cancelling out of lando rebuild then running landocommand served to re-index the database.
(lando destroy deleted the database.)

[aploshay]

Could you add comments in the pull request and/or pertinent commit, re: the effects of each filter removed or added?

[rdlebeau]

Removing Solr token filter factory

• EnglishMinimalStemFilterFactory reduces English plural words to singular words for search purposes. We want a more aggresive approach, so we're no longer using it.

Adding Solr token filter factories

• EnglishPossessiveFilterFactory removes English possessives from search words. The word "cat's" becomes "cat" etc. in a search.

• LowerCaseFilterFactory enables case-insensitive searching. It converts uppercase letters to lowercase, while leaving numbers and other characters unchanged.

• SnowballPorterFilterFactory reduces search words to their root form using Snowball stemming algorithms. This is the big change that we want, the one that allows us to search for "Nielsen" and see "NielsenIQ" in the search results. It also reduces English plural words to singular words, so we don't lose any functionality by removing EnglishMinimalStemFilterFactory.

• NGramFilterFactory splits up search words to support partial word matching. Adding it lets us search "IQ" and get "NielsenIQ" even though it's the second half of the word. We can also search "ielsenI" for example and see the "NielsenIQ" data set in the search results.

• TrimFilterFactory removes leading and trailing whitespace from search words.