Lucene 8 upgrade checklist

[jimczi]

This is a meta issue to track the progress of upgrading master to Lucene 8. The branch can be found here. Below is the list of things to be changed/removed for the upgrade:

• Levenstein distance renaming.
• TermContext renaming (TermStates).
• Collectors should implement scoreMode instead of needsScore. (@jpountz)
• IndexSearcher#createNormalizedWeight has been removed. (@jpountz)
• StandardFilter removal. (@jimczi)
• EarlyTerminatingSortingCollector removal. (@jpountz)
• BoostingQuery removal (replaced with FunctionScoreQuery#boostByQuery). (@jimczi)
• Similarity#SimWeight removal. (@jpountz)
• BasicModel removals (after effects, BE, D and P). (@jpountz)
• Version#LUCENE_6* removal. (@jimczi Remove unsupported Version.V_5_* #32937)
• TopDocs#totalHits rewriting.
• TopDocs#computeMaxScore replacement
• Explanation#getValue() is a Number instead of a float

Test failures

• PercolateQueryBuilder#createMultiDocumentSearcher, PercolatorQuerySearchIT: @jpountz
• FunctionScoreTests#testSimpleWeightedFunction: produces negative scores (Forbid negative values for "weight" in Function Score Query #31927)
• ScriptedSimilarityTests#testBasics, ScriptedSimilarityTests#testInit: score is called on non-scoring clauses (to get the max score, see ImpactsDISI)
• FieldSubsetReaderTests#testFieldNames, FieldSubsetReaderTests#testFieldNamesMissing, FieldSubsetReaderTests#testFieldThreeFields: FieldNamesTerms return invalid statistics (e.g. -1) Security for _field_names field should not override field statistics #33261
• TransportTwoNodesSearchIT#testDfsQueryThenFetch, TransportTwoNodesSearchIT#testDfsQueryThenFetchWithSort: new explanation format and message

\cc @elastic/es-distributed to sync up.

[jasontedor]

Pinging @dnhatn