Commits on Dec 2, 2021

1. fix: Make QGramIndex use less memory; fix bug; improve tests&docs …

   I've added some more documentation and clarifying tests to `QGramIndex`,
   so that the implementation is more clear without looking at the code.
   
   I found one small bug: q-grams of value `0` (i.e. `AAA`) wouldn't be
   filtered by `max_count`. I've added a test that would fail if the
   `.skip(1)` was still there.
   
   The size of the `pos` vector was always `|text|`, but it's sufficient to
   only allocate memory for the number of unfiltered q-grams. Please verify
   this. I've added a `text="AAAAA"`, `pattern="AAA"` test that should show
   that things are still working.

   

   RagnarGrootKoerkamp committed Dec 2, 2021
   Copy the full SHA

   adb363a View commit details

   Browse the repository at this point in the history