From 317ffbf109bedac429684c5c606036f27e103896 Mon Sep 17 00:00:00 2001
From: Cipher208 <269750686+Cipher208@users.noreply.github.com>
Date: Wed, 7 Oct 2026 15:33:54 +0200
Subject: [PATCH] fix(graph): give the zero-results miner the same
 honest-failure contract
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

`miner_zero_results` had the same shape as the bug just fixed in
`miner_embedding`: one `except Exception` that returned `{"edges": 0}` — the
value meaning "there was nothing to surface" — and reported the reason at
`debug`, which is invisible at the level this service runs at. A failure in
`find_or_add_entity` would therefore look exactly like a quiet night, and the
record of what happened would not reach `memory.log`.

After the previous commit, `0` means something specific: this miner looked and
found no work. Leaving a second path that returns `0` on failure would make
that contract a lie again, and the two cases would be indistinguishable by the
new per-miner log line as well.

The exception now escapes to the handler in `graph_enrich` that records
`{"edges": -1}` plus a warning. The reachable `return {"edges": 0}` for "the
journal had no repeated misses" stays — that one is genuine.

This is the only other swallow in the file. The other `except Exception` left
in `graph_miners` is `_get_ner`, which probes whether spaCy is installed and
is a legitimate probe rather than a miner.

Test: a backend failure escapes `miner_zero_results`, alongside the two
existing zero-result tests that pin the real zero.

Full suite: 2086 passed.
---
 lifecycle/graph_miners.py     | 21 +++++++++++----------
 tests/test_s17_stage1_tail.py | 29 +++++++++++++++++++++++++++++
 2 files changed, 40 insertions(+), 10 deletions(-)

diff --git a/lifecycle/graph_miners.py b/lifecycle/graph_miners.py
index 3ff81d6..c3ad819 100644
--- a/lifecycle/graph_miners.py
+++ b/lifecycle/graph_miners.py
@@ -1283,16 +1283,17 @@ async def miner_zero_results(cm: AsyncConnectionManager, layer: str) -> dict[str
     if not rows:
         return {"edges": 0}
     surfaced = 0
-    try:
-        from graph.epistemic import EpistemicGraph
-
-        g = EpistemicGraph(cm=cm, layer=layer)
-        for r in rows:
-            await g.find_or_add_entity(str(r["user_id"]), str(r["query"])[:300], "question")
-            surfaced += 1
-    except Exception:
-        logger.debug("zero-result surfacing failed", exc_info=True)
-        return {"edges": 0}
+    # No try/except here, for the same reason `miner_embedding` has none: a
+    # failure that returns `{"edges": 0}` is indistinguishable from "there was
+    # nothing to surface", and this one logged at `debug`, which is invisible
+    # at the level this service actually runs at. Letting it escape reaches the
+    # handler in `graph_enrich` that records `edges: -1` with a warning.
+    from graph.epistemic import EpistemicGraph
+
+    g = EpistemicGraph(cm=cm, layer=layer)
+    for r in rows:
+        await g.find_or_add_entity(str(r["user_id"]), str(r["query"])[:300], "question")
+        surfaced += 1
     return {"edges": surfaced}
 
 
diff --git a/tests/test_s17_stage1_tail.py b/tests/test_s17_stage1_tail.py
index 75bc71a..4b74f6d 100644
--- a/tests/test_s17_stage1_tail.py
+++ b/tests/test_s17_stage1_tail.py
@@ -243,6 +243,35 @@ async def test_zero_result_single_hit_not_surfaced(migrated_cm) -> None:
     assert res["edges"] == 0, "один провал — не сигнал"
 
 
+@pytest.mark.asyncio
+async def test_zero_result_miner_lets_a_failure_escape(migrated_cm, monkeypatch) -> None:
+    """Сбой обязан выйти наружу, а не стать нулём с записью в `debug`.
+
+    07.10.2026: здесь был второй такой же глушитель, как в `miner_embedding` —
+    `except Exception` писал в `debug` (невидимо на рабочем уровне) и возвращал
+    `{"edges": 0}`, то есть «нечего выносить». Договор один на всех шахтёров:
+    ноль — это работа, которая не нашлась, а сбой пусть ловит `graph_enrich`.
+    """
+    from graph.epistemic import EpistemicGraph
+    from lifecycle.graph_miners import log_zero_result, miner_zero_results
+    from shared.constants import DB_NAME
+
+    cm = migrated_cm
+    conn = await cm.get(DB_NAME)
+    await conn.execute("DELETE FROM recall_zero_results")
+    await conn.commit()
+    for _ in range(2):
+        await log_zero_result(cm, "user", "s17zr3", "провал запроса, который надо вынести наверх")
+
+    async def _boom(self, *args, **kwargs):
+        raise RuntimeError("find_or_add_entity is down")
+
+    monkeypatch.setattr(EpistemicGraph, "find_or_add_entity", _boom)
+
+    with pytest.raises(RuntimeError, match="find_or_add_entity is down"):
+        await miner_zero_results(cm, "user")
+
+
 # ── 7. counter-signal алиасы ─────────────────────────────────────────────────