;; Phase 1 — tokenize + inverted index.
;; All cases run in ONE program evaluation (search-batch) to stay fast under load.
;; Scalar results (docFreq) are wrapped as singleton lists so the batch is a list
;; of lists.

(define
  index-cases
  (list
    (list
      "tokens basic lowercases"
      "tokens \"The Cat sat\""
      (list "the" "cat" "sat"))
    (list
      "tokens strips punctuation"
      "tokens \"Hello, World!\""
      (list "hello" "world"))
    (list "tokens collapses whitespace" "tokens \"  a   b  \"" (list "a" "b"))
    (list "tokens empty is empty" "tokens \"\"" (list))
    (list "tokens keeps digits" "tokens \"abc123 x9\"" (list "abc123" "x9"))
    (list
      "positioned attaches ordinals"
      "positioned \"a b a\""
      (list
        (list "a" 0)
        (list "b" 1)
        (list "a" 2)))
    (list
      "index + lookup single doc"
      "lookupTerm \"cat\" (indexDoc 1 \"the cat sat\" emptyIndex)"
      (list (list 1 (list 1))))
    (list
      "lookup missing term is empty"
      "lookupTerm \"dog\" (indexDoc 1 \"the cat sat\" emptyIndex)"
      (list))
    (list
      "lookup records all positions"
      "lookupTerm \"the\" (indexDoc 1 \"the cat the dog the\" emptyIndex)"
      (list (list 1 (list 0 2 4))))
    (list
      "multi-doc posting list sorted by docid"
      "lookupTerm \"x\" (indexDoc 1 \"x y\" (indexDoc 2 \"x z\" emptyIndex))"
      (list
        (list 1 (list 0))
        (list 2 (list 0))))
    (list
      "index/query case symmetry"
      "lookupTerm \"cat\" (indexDoc 1 \"CAT Cat cat\" emptyIndex)"
      (list (list 1 (list 0 1 2))))
    (list
      "re-index replaces a doc"
      "lookupTerm \"a\" (indexDoc 1 \"a a a\" (indexDoc 1 \"a\" emptyIndex))"
      (list (list 1 (list 0 1 2))))
    (list
      "delete removes a doc"
      "lookupTerm \"cat\" (deleteDoc 1 (indexDoc 1 \"the cat\" emptyIndex))"
      (list))
    (list
      "delete leaves other docs"
      "lookupTerm \"cat\" (deleteDoc 2 (indexDoc 2 \"big cat\" (indexDoc 1 \"the cat\" emptyIndex)))"
      (list (list 1 (list 1))))
    (list
      "docFreq counts docs"
      "[docFreq \"cat\" (indexDoc 2 \"a cat\" (indexDoc 1 \"the cat\" emptyIndex))]"
      (list 2))
    (list
      "docFreq zero for missing"
      "[docFreq \"zzz\" (indexDoc 1 \"a b\" emptyIndex)]"
      (list 0))
    (list
      "allTerms sorted and unique"
      "allTerms (indexDoc 1 \"banana apple cherry apple\" emptyIndex)"
      (list "apple" "banana" "cherry"))
    (list
      "allTerms merged across docs"
      "allTerms (indexDoc 2 \"d a\" (indexDoc 1 \"c b\" emptyIndex))"
      (list "a" "b" "c" "d"))))

(define
  index-results
  (search-batch "" (map (fn (c) (nth c 1)) index-cases)))

(map-indexed
  (fn
    (i c)
    (hk-test (nth c 0) (nth index-results i) (nth c 2)))
  index-cases)

{:fail hk-test-fail :pass hk-test-pass :fails hk-test-fails}