From 16fe22669a24fd7bc5c2a756d808fb92c9222c01 Mon Sep 17 00:00:00 2001
From: giles <giles.bradshaw@rose-ash.com>
Date: Fri, 8 May 2026 11:19:29 +0000
Subject: [PATCH] =?UTF-8?q?mk:=20cyclic-graph=20behaviour=20test=20?=
 =?UTF-8?q?=E2=80=94=20motivates=20Phase=207=20tabling?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Demonstrates the naive-patho behaviour on a 2-cycle (a <-> b, plus
b -> c). Without Phase-7 tabling, the search produces ever-longer
paths: (a b), (a b a b), (a b a b a b), ... `run 5` truncates to a
finite prefix; `run*` diverges. Documenting this as a regression-style
test gives Phase 7 a concrete starting point.

3 new tests, 331/331 cumulative.
---
 lib/minikanren/tests/cyclic-graph.sx | 48 ++++++++++++++++++++++++++++
 plans/minikanren-on-sx.md            |  5 +++
 2 files changed, 53 insertions(+)
 create mode 100644 lib/minikanren/tests/cyclic-graph.sx

diff --git a/lib/minikanren/tests/cyclic-graph.sx b/lib/minikanren/tests/cyclic-graph.sx
new file mode 100644
index 00000000..f40675d2
--- /dev/null
+++ b/lib/minikanren/tests/cyclic-graph.sx
@@ -0,0 +1,48 @@
+;; lib/minikanren/tests/cyclic-graph.sx — demonstrates the naive-patho
+;; behaviour on a cyclic graph. Without Phase-7 tabling/SLG, the search
+;; produces ever-longer paths revisiting the cycle. `run n` truncates;
+;; `run*` would diverge.
+
+(define cyclic-edges (list (list :a :b) (list :b :a) (list :b :c)))
+
+(define cyclic-edgeo (fn (x y) (membero (list x y) cyclic-edges)))
+
+(define
+  cyclic-patho
+  (fn
+    (x y path)
+    (conde
+      ((cyclic-edgeo x y) (== path (list x y)))
+      ((fresh (z mid) (cyclic-edgeo x z) (cyclic-patho z y mid) (conso x mid path))))))
+
+;; --- direct edge ---
+
+(mk-test
+  "cyclic-direct"
+  (run 1 q (cyclic-patho :a :b q))
+  (list (list :a :b)))
+
+;; --- runs first 5 paths from a to b: bare edge, then increasing
+;;     numbers of cycle traversals (a->b->a->b, etc.) ---
+
+(mk-test
+  "cyclic-enumerates-prefix-via-run-n"
+  (let
+    ((paths (run 5 q (cyclic-patho :a :b q))))
+    (and
+      (= (len paths) 5)
+      (and
+        (every? (fn (p) (= (first p) :a)) paths)
+        (every? (fn (p) (= (last p) :b)) paths))))
+  true)
+
+(mk-test
+  "cyclic-finds-c-via-cycle-or-direct"
+  (let
+    ((paths (run 3 q (cyclic-patho :a :c q))))
+    (and
+      (>= (len paths) 1)
+      (some (fn (p) (= p (list :a :b :c))) paths)))
+  true)
+
+(mk-tests-run!)
diff --git a/plans/minikanren-on-sx.md b/plans/minikanren-on-sx.md
index d40a86c9..ca39f8c5 100644
--- a/plans/minikanren-on-sx.md
+++ b/plans/minikanren-on-sx.md
@@ -173,6 +173,11 @@ _(none yet)_
 
 _Newest first._
 
+- **2026-05-08** — **Cyclic graph behaviour test (motivates Phase 7 tabling)**:
+  Demonstrates that naive patho on a 2-cycle generates ever-longer paths.
+  `run 5` truncates to a finite prefix; `run*` would diverge. This is
+  exactly the test case Phase 7 (tabled / SLG resolution) is designed
+  to fix. 3 new tests, 331/331 cumulative.
 - **2026-05-08** — **numbero / stringo / symbolo (type predicates)**: ground-only
   type tests via project. Compose with `membero` for type-filtered enumeration:
   `(fresh (x) (membero x (1 "a" 2 "b" 3)) (numbero x) (== q x))` → `(1 2 3)`.