use trigram tokenizer instead of filter to maintain ordering

[nicpottier]

This switches to using the trigram tokenizer instead of filter.

With the filter, we were turning a single token into three tokens, all with position of 0.

With the tokenizer, each trigram has a position (starting at 0), which then lets us use a phrase query to match not just the presence but ordering of the trigrams, which is what we need.

Added a test to demonstrate.

[codecov-io]

Codecov Report

Merging #5 into master will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##           master       #5   +/-   ##
=======================================
  Coverage   83.33%   83.33%           
=======================================
  Files           1        1           
  Lines         144      144           
=======================================
  Hits          120      120           
  Misses         12       12           
  Partials       12       12

Impacted Files	Coverage Δ
indexer.go	83.33% <ø> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 2236db8...e1d7b53. Read the comment docs.

[nicpottier]

merging so I can start rebuilds