search: Phase 2 query AST + boolean/phrase eval + 28 tests

Query ADT (Term|And|Or|Not|Phrase) and evalQuery over docid-sorted posting lists: boolean ops as linear merges, Not over the allDocs universe, Phrase via positional adjacency. Batched both test suites into one program eval each (search-batch) so they finish under heavy CPU load. 46/46. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-06 18:47:42 +00:00
parent b8cf3eb1b8
commit 0f0da0319c
9 changed files with 264 additions and 125 deletions
--- a/lib/search/tests/boolean.sx
+++ b/lib/search/tests/boolean.sx
@@ -0,0 +1,123 @@
+;; Phase 2 — query AST + boolean/phrase evaluation (hand-built Query values).
+;; Corpus:
+;;   doc 1 "the quick brown dog"   -> the quick brown dog
+;;   doc 2 "a quick brown fox"     -> a quick brown fox
+;;   doc 3 "the dog barks loudly"  -> the dog barks loudly
+;; All queries run in ONE program evaluation (search-batch) to stay fast.
+
+(define
+  search-corpus
+  "idx = indexDoc 3 \"the dog barks loudly\" (indexDoc 2 \"a quick brown fox\" (indexDoc 1 \"the quick brown dog\" emptyIndex))\n")
+
+(define
+  bool-cases
+  (list
+    (list
+      "term in two docs"
+      "evalQuery idx (Term \"quick\")"
+      (list 1 2))
+    (list
+      "term in two docs (the)"
+      "evalQuery idx (Term \"the\")"
+      (list 1 3))
+    (list "term in one doc" "evalQuery idx (Term \"fox\")" (list 2))
+    (list "term absent" "evalQuery idx (Term \"zzz\")" (list))
+    (list
+      "term case-sensitive at AST level"
+      "evalQuery idx (Term \"QUICK\")"
+      (list))
+    (list "term on empty index" "evalQuery emptyIndex (Term \"cat\")" (list))
+    (list
+      "and both terms"
+      "evalQuery idx (And (Term \"quick\") (Term \"brown\"))"
+      (list 1 2))
+    (list
+      "and overlap subset"
+      "evalQuery idx (And (Term \"the\") (Term \"dog\"))"
+      (list 1 3))
+    (list
+      "and disjoint is empty"
+      "evalQuery idx (And (Term \"the\") (Term \"fox\"))"
+      (list))
+    (list
+      "and right-nested"
+      "evalQuery idx (And (Term \"the\") (And (Term \"dog\") (Term \"barks\")))"
+      (list 3))
+    (list
+      "or two singletons"
+      "evalQuery idx (Or (Term \"fox\") (Term \"barks\"))"
+      (list 2 3))
+    (list
+      "or all docs"
+      "evalQuery idx (Or (Term \"quick\") (Term \"the\"))"
+      (list 1 2 3))
+    (list
+      "or with absent term"
+      "evalQuery idx (Or (Term \"fox\") (Term \"zzz\"))"
+      (list 2))
+    (list "not term" "evalQuery idx (Not (Term \"the\"))" (list 2))
+    (list "not term 2" "evalQuery idx (Not (Term \"quick\"))" (list 3))
+    (list
+      "and with not"
+      "evalQuery idx (And (Term \"quick\") (Not (Term \"the\")))"
+      (list 2))
+    (list
+      "double negation"
+      "evalQuery idx (Not (Not (Term \"fox\")))"
+      (list 2))
+    (list
+      "or of and with term"
+      "evalQuery idx (Or (And (Term \"the\") (Term \"dog\")) (Term \"fox\"))"
+      (list 1 2 3))
+    (list
+      "phrase adjacent both docs"
+      "evalQuery idx (Phrase [\"quick\", \"brown\"])"
+      (list 1 2))
+    (list
+      "phrase adjacent one doc"
+      "evalQuery idx (Phrase [\"brown\", \"dog\"])"
+      (list 1))
+    (list
+      "phrase the quick"
+      "evalQuery idx (Phrase [\"the\", \"quick\"])"
+      (list 1))
+    (list
+      "phrase dog barks"
+      "evalQuery idx (Phrase [\"dog\", \"barks\"])"
+      (list 3))
+    (list
+      "phrase non-adjacent empty"
+      "evalQuery idx (Phrase [\"quick\", \"dog\"])"
+      (list))
+    (list
+      "phrase order matters"
+      "evalQuery idx (Phrase [\"brown\", \"quick\"])"
+      (list))
+    (list
+      "phrase single term"
+      "evalQuery idx (Phrase [\"dog\"])"
+      (list 1 3))
+    (list
+      "phrase three terms"
+      "evalQuery idx (Phrase [\"the\", \"dog\", \"barks\"])"
+      (list 3))
+    (list
+      "and of phrase and term"
+      "evalQuery idx (And (Phrase [\"quick\", \"brown\"]) (Term \"dog\"))"
+      (list 1))
+    (list
+      "not of phrase"
+      "evalQuery idx (Not (Phrase [\"quick\", \"brown\"]))"
+      (list 3))))
+
+(define
+  bool-results
+  (search-batch search-corpus (map (fn (c) (nth c 1)) bool-cases)))
+
+(map-indexed
+  (fn
+    (i c)
+    (hk-test (nth c 0) (nth bool-results i) (nth c 2)))
+  bool-cases)
+
+{:fail hk-test-fail :pass hk-test-pass :fails hk-test-fails}