Should the EdgeNGramFilter use the same term position for the ngrams
within a single token?
As is, the EdgeNGramTokenFilter increments the term position for each
character. In analysis.jsp, with the input "hello", I get:
term position 1 2 3 4 5
term text h he hel hell hello
term type word word word word word
start,end 0,1 0,2 0,3 0,4 0,5
I would expect something more like what is generated from SOLR-357:
term position 1
term text hello
term type word
This seems like it would affect slop queries, but I don't really
understand them yet.