From 5aa2412deef4f334085d2f4b7af44ab4fac67fb8 Mon Sep 17 00:00:00 2001 From: OoO Date: Sun, 24 May 2026 18:20:36 +0800 Subject: [PATCH] Route rescore matches to manual review --- TODO_NEXT_STEPS.txt | 1 + config.py | 2 +- docs/AI_INTELLIGENCE_MODULE_SOT.md | 2 +- docs/memory/history_logs.md | 1 + routes/dashboard_routes.py | 12 ++ .../audit_competitor_match_attempt_rescore.py | 48 +++++- services/competitor_intel_repository.py | 39 +++-- .../competitor_match_attempt_rescore_audit.py | 162 ++++++++++++++++++ tests/test_competitor_identity_revalidator.py | 14 ++ ..._competitor_match_attempt_rescore_audit.py | 68 ++++++++ 10 files changed, 322 insertions(+), 27 deletions(-) diff --git a/TODO_NEXT_STEPS.txt b/TODO_NEXT_STEPS.txt index a9697a9..b2bb136 100644 --- a/TODO_NEXT_STEPS.txt +++ b/TODO_NEXT_STEPS.txt @@ -4,6 +4,7 @@ ================================================================================ 【已完成】 + - V10.443 補 PChome rescore 人工覆核入隊：`audit_competitor_match_attempt_rescore.py --apply-accepted` 只追加 `rescore_accepted_current` attempt 進人工覆核隊列，不直接寫 `competitor_prices` / `competitor_price_history`；商品看板新增「重算可採用」分流與狀態文案，讓可救回候選先由人審確認再正式更新價差。 - V10.442 降噪 `/cicd` 舊 GitLab 探測：沒有明確啟用 `GITLAB_ENABLED=true` 與 token 時，不再打退役的 `192.168.0.110:8929` 或 SSH fallback，正式 responsive smoke 造訪 `/cicd` 只呈現空 pipeline 狀態，不污染 app logs。 - V10.441 補 PChome matcher re-score audit 與商品看板原因標籤：新增 read-only `competitor_match_attempt_rescore_audit` / `scripts/audit_competitor_match_attempt_rescore.py`，可用最新版 matcher 重新分類既有 `competitor_match_attempts`，預設不寫 DB、不更新正式價格；商品看板同步補蘭蔻/達特醫/hoi/Saugella/Lactacyd 等 focused matcher reason 中文標籤，讓「待對比」能拆成商品線不符、款式版本不符、可回刷或仍低信心。 - V10.439 收斂外部 BI / 資料協作入口：`/metabase`、`/grist` 正式頁維持 momo-pro 內部診斷 bridge，`.env.example` 與 bi profile Grist 預設改回 `https://mo.wooo.work/grist` / `GRIST_APP_HOME_URL`，並補測試禁止 `grist.wooo.work` / `awoooi` 回流到導覽設定；外部工具頁標題字級改用新版 token 與手機 media query。 diff --git a/config.py b/config.py index d3c6066..aec4ca8 100644 --- a/config.py +++ b/config.py @@ -325,7 +325,7 @@ YOUTUBE_API_KEY = os.getenv('YOUTUBE_API_KEY', '') # ========================================== # 系統版本與路徑 # ========================================== -SYSTEM_VERSION = "V10.442" +SYSTEM_VERSION = "V10.443" LOG_FILE_PATH = os.path.join(BASE_DIR, 'logs/system.log') public_url = PUBLIC_URL # 用於模板顯示 diff --git a/docs/AI_INTELLIGENCE_MODULE_SOT.md b/docs/AI_INTELLIGENCE_MODULE_SOT.md index bca7806..de4d274 100644 --- a/docs/AI_INTELLIGENCE_MODULE_SOT.md +++ b/docs/AI_INTELLIGENCE_MODULE_SOT.md @@ -2,7 +2,7 @@ > **最後更新**: 2026-05-24 (台北時間) > **狀態**: 🟢 四 AI Agent 自動化閉環已落地；LLM 路由紅線升級為 Ollama-first 三主機級聯，Gemini 備援預設關閉 -> **適用版本**: V10.442 +> **適用版本**: V10.443 --- diff --git a/docs/memory/history_logs.md b/docs/memory/history_logs.md index f4787ff..0fa1d80 100644 --- a/docs/memory/history_logs.md +++ b/docs/memory/history_logs.md @@ -13,6 +13,7 @@ ## 📅 詳細更新日誌 (考古存檔) ### 2026-05-24：PChome 近門檻身份回收第二輪 +- **V10.443 PChome rescore 人工覆核入隊**: `scripts/audit_competitor_match_attempt_rescore.py --apply-accepted` 只會把最新版 matcher 已通過門檻的舊低信心候選追加為 `rescore_accepted_current` attempt，進入商品看板人工覆核隊列；不寫 `competitor_prices`、不寫 `competitor_price_history`，必須由操作員按「採用同款」後才正式更新 PChome 價差。Dashboard 補上「重算可採用待審」分流與狀態文案，避免安全回刷候選混在一般低信心項目裡。 - **V10.442 CI/CD legacy GitLab 探測降噪**: `/cicd` 舊 GitLab pipeline API 預設改為 disabled，除非明確設定 `GITLAB_ENABLED=true` 並提供 `GITLAB_TOKEN`，否則不再打 `192.168.0.110:8929` 或 SSH fallback；正式 responsive smoke 造訪 `/cicd` 時只呈現可診斷空狀態，不再把已退役 GitLab endpoint 的 connection refused / permission denied 寫成錯誤噪音。 - **V10.441 PChome matcher re-score audit**: 新增 read-only `competitor_match_attempt_rescore_audit` 服務與 `scripts/audit_competitor_match_attempt_rescore.py`，可針對既有 `competitor_match_attempts` 用最新版 matcher 重新分類成 `accepted_current` / `unit_comparable_current` / `identity_veto_current` / `low_score_current`，預設不寫 DB、不更新正式價格；商品看板同步補蘭蔻/達特醫/hoi/Saugella/Lactacyd 等 focused matcher reason 中文標籤。正式抽樣中 31 筆舊 `strong_exact_spec_match` 低信心資料，最新版 matcher 可讀出 10 筆 gate pass、1 筆單位價、11 筆 hard veto、9 筆仍低信心，作為後續人工覆核與批次回刷前的安全量化。 - **V10.440 Mustela 爽身潤膚乳同款 anchor**: marketplace matcher 新增 `慕之幼爽身潤膚乳` identity anchor，並讓標題中插入「加量版」時仍可抽出同一身份詞；正式樣本 `【Mustela 慕之恬廊】慕之幼 加量版爽身潤膚乳 500mlX2入` vs `【慕之恬廊】慕之幼爽身潤膚乳(500毫升X2入)` 由 0.741 提升到 0.801，維持 `hard_veto=false`、人工 review 型態，不放寬全域門檻、不寫正式 `competitor_prices`。 diff --git a/routes/dashboard_routes.py b/routes/dashboard_routes.py index ac18926..b547519 100644 --- a/routes/dashboard_routes.py +++ b/routes/dashboard_routes.py @@ -61,8 +61,10 @@ REVIEW_STATUS_OPTIONS = [ 'manual_rejected', 'manual_unit_price_required', 'manual_needs_research', + 'rescore_accepted_current', ), }, + {'key': 'rescore_accepted', 'label': '重算可採用', 'statuses': ('rescore_accepted_current',)}, { 'key': 'unit_comparable', 'label': '需單位價', @@ -147,6 +149,16 @@ def _diagnostic_match_rejection_label(diagnostic_text, score_text, *, blocked=Tr def _build_pchome_match_status(attempt=None, ineligible=None): if attempt: status = attempt.get('attempt_status') or 'unknown' + if status == 'rescore_accepted_current': + score = _to_float(attempt.get('best_match_score')) + score_text = f"目前信心 {int(score * 100)}%" if score is not None else '目前信心待補' + return { + 'label': '重算可採用待審', + 'tone': 'watch', + 'score': score, + 'summary': f'{score_text}，最新版 matcher 已通過身份門檻；仍需人工採用後才寫入正式 PChome 價差', + 'detail': attempt.get('error_message') + } if status == 'manual_accepted': score = _to_float(attempt.get('best_match_score')) score_text = f"人工採用 {round(score * 100)}%" if score is not None else "人工已採用同款" diff --git a/scripts/audit_competitor_match_attempt_rescore.py b/scripts/audit_competitor_match_attempt_rescore.py index bd0f1f0..5c3beeb 100755 --- a/scripts/audit_competitor_match_attempt_rescore.py +++ b/scripts/audit_competitor_match_attempt_rescore.py @@ -19,6 +19,8 @@ if str(ROOT) not in sys.path: from services.competitor_match_attempt_rescore_audit import ( # noqa: E402 DEFAULT_RESCAN_STATUSES, build_match_attempt_rescore_audit, + fetch_match_attempt_rescore_rows, + materialize_rescore_accept_reviews, summarize_match_attempt_rescore, ) from services.competitor_price_feeder import MIN_MATCH_SCORE # noqa: E402 @@ -59,10 +61,17 @@ def main(argv: list[str] | None = None) -> int: parser.add_argument("--limit", type=int, default=100) parser.add_argument("--sample-limit", type=int, default=20) parser.add_argument("--min-score", type=float, default=MIN_MATCH_SCORE) + parser.add_argument( + "--apply-accepted", + action="store_true", + help="Append accepted-current rows to competitor_match_attempts for manual review; never writes competitor_prices.", + ) args = parser.parse_args(argv) statuses = tuple(args.statuses or DEFAULT_RESCAN_STATUSES) if args.input: + if args.apply_accepted: + parser.error("--apply-accepted requires DB mode; do not combine it with --input.") rows = [row for row in _read_jsonl(args.input) if not row.get("_invalid_json")] summary = summarize_match_attempt_rescore( rows, @@ -73,15 +82,36 @@ def main(argv: list[str] | None = None) -> int: from config import DATABASE_PATH engine = create_engine(DATABASE_PATH) - summary = build_match_attempt_rescore_audit( - engine, - source=args.source, - statuses=statuses, - reason_filter=args.reason_filter or None, - limit=args.limit, - min_score=args.min_score, - sample_limit=args.sample_limit, - ) + if args.apply_accepted: + with engine.begin() as conn: + rows = fetch_match_attempt_rescore_rows( + conn, + source=args.source, + statuses=statuses, + reason_filter=args.reason_filter or None, + limit=args.limit, + ) + summary = summarize_match_attempt_rescore( + rows, + min_score=args.min_score, + sample_limit=args.sample_limit, + ) + summary["materialize"] = materialize_rescore_accept_reviews( + conn, + rows, + source=args.source, + min_score=args.min_score, + ) + else: + summary = build_match_attempt_rescore_audit( + engine, + source=args.source, + statuses=statuses, + reason_filter=args.reason_filter or None, + limit=args.limit, + min_score=args.min_score, + sample_limit=args.sample_limit, + ) print(json.dumps(summary, ensure_ascii=False, indent=2, default=str)) return 0 diff --git a/services/competitor_intel_repository.py b/services/competitor_intel_repository.py index b443e62..8f87901 100644 --- a/services/competitor_intel_repository.py +++ b/services/competitor_intel_repository.py @@ -25,6 +25,7 @@ from sqlalchemy import inspect, text PCHOME_MATCH_SCORE_FLOOR = 0.76 UNIT_COMPARABLE_STATUSES = {"unit_comparable", "refresh_unit_comparable"} ACTIONABLE_ATTEMPT_STATUSES = { + "rescore_accepted_current", "unit_comparable", "refresh_unit_comparable", "identity_veto", @@ -41,6 +42,7 @@ ACTIONABLE_ATTEMPT_STATUSES = { "manual_needs_research", } REVIEW_STATUS_FILTER_GROUPS = { + "rescore_accepted": ("rescore_accepted_current",), "unit_comparable": ("unit_comparable", "refresh_unit_comparable", "manual_unit_price_required"), "identity_veto": ("identity_veto",), "low_score": ("low_score", "refresh_low_score", "recoverable_low_score", "true_low_confidence"), @@ -50,6 +52,7 @@ REVIEW_STATUS_FILTER_GROUPS = { "manual_closed": ("manual_rejected", "manual_unit_price_required", "manual_needs_research"), } ATTEMPT_STATUS_LABELS = { + "rescore_accepted_current": "重算可採用待審", "unit_comparable": "需單位價比較", "refresh_unit_comparable": "需單位價比較", "identity_veto": "身份否決", @@ -68,6 +71,7 @@ ATTEMPT_STATUS_LABELS = { "manual_needs_research": "人工要求補搜尋", } ATTEMPT_ACTION_LABELS = { + "rescore_accepted_current": "覆核後可人工採用同款", "unit_comparable": "人工確認檔期、贈品與單位價", "refresh_unit_comparable": "人工確認檔期、贈品與單位價", "identity_veto": "確認是否為不同商品線或規格", @@ -780,7 +784,7 @@ def fetch_competitor_review_queue(engine, limit: int = 12) -> list[dict]: """可行動的 PChome 比對覆核隊列，供 Dashboard / AI / PPT 共用。""" limit = max(1, min(int(limit or 12), 50)) return _cached_payload( - f"review_queue:v1:limit={limit}:floor={PCHOME_MATCH_SCORE_FLOOR}", + f"review_queue:v2:limit={limit}:floor={PCHOME_MATCH_SCORE_FLOOR}", lambda: _fetch_competitor_review_queue_uncached(engine, limit=limit), ) @@ -802,7 +806,7 @@ def fetch_competitor_review_queue_page( if status_filter not in REVIEW_STATUS_FILTER_GROUPS: status_filter = "" cache_key = ( - "review_queue_page:v1:" + "review_queue_page:v2:" f"page={page}:per={per_page}:q={search_query.lower()}:cat={category}:" f"status={status_filter}:" f"floor={PCHOME_MATCH_SCORE_FLOOR}" @@ -904,13 +908,14 @@ def _review_queue_cte_and_filter( la.error_message, la.attempted_at, CASE - WHEN la.attempt_status IN ('unit_comparable', 'refresh_unit_comparable') THEN 0 - WHEN la.attempt_status = 'identity_veto' THEN 1 - WHEN la.attempt_status IN ('recoverable_low_score', 'low_score', 'refresh_low_score') THEN 2 - WHEN la.attempt_status = 'protected_existing_match' THEN 3 - WHEN la.attempt_status = 'true_low_confidence' THEN 4 - WHEN la.attempt_status = 'expired_match' THEN 5 - ELSE 6 + WHEN la.attempt_status = 'rescore_accepted_current' THEN 0 + WHEN la.attempt_status IN ('unit_comparable', 'refresh_unit_comparable') THEN 1 + WHEN la.attempt_status = 'identity_veto' THEN 2 + WHEN la.attempt_status IN ('recoverable_low_score', 'low_score', 'refresh_low_score') THEN 3 + WHEN la.attempt_status = 'protected_existing_match' THEN 4 + WHEN la.attempt_status = 'true_low_confidence' THEN 5 + WHEN la.attempt_status = 'expired_match' THEN 6 + ELSE 7 END AS priority_rank FROM latest_momo lm JOIN latest_attempt la ON la.sku = lm.sku @@ -1057,6 +1062,7 @@ def _fetch_competitor_review_queue_uncached(engine, limit: int = 12) -> list[dic LEFT JOIN valid_competitor vc ON vc.sku = lm.sku WHERE vc.sku IS NULL AND la.attempt_status IN ( + 'rescore_accepted_current', 'unit_comparable', 'refresh_unit_comparable', 'identity_veto', @@ -1071,13 +1077,14 @@ def _fetch_competitor_review_queue_uncached(engine, limit: int = 12) -> list[dic ) ORDER BY CASE - WHEN la.attempt_status IN ('unit_comparable', 'refresh_unit_comparable') THEN 0 - WHEN la.attempt_status = 'identity_veto' THEN 1 - WHEN la.attempt_status IN ('recoverable_low_score', 'low_score', 'refresh_low_score') THEN 2 - WHEN la.attempt_status = 'protected_existing_match' THEN 3 - WHEN la.attempt_status = 'true_low_confidence' THEN 4 - WHEN la.attempt_status = 'expired_match' THEN 5 - ELSE 6 + WHEN la.attempt_status = 'rescore_accepted_current' THEN 0 + WHEN la.attempt_status IN ('unit_comparable', 'refresh_unit_comparable') THEN 1 + WHEN la.attempt_status = 'identity_veto' THEN 2 + WHEN la.attempt_status IN ('recoverable_low_score', 'low_score', 'refresh_low_score') THEN 3 + WHEN la.attempt_status = 'protected_existing_match' THEN 4 + WHEN la.attempt_status = 'true_low_confidence' THEN 5 + WHEN la.attempt_status = 'expired_match' THEN 6 + ELSE 7 END, lm.momo_price DESC NULLS LAST, la.best_match_score DESC NULLS LAST, diff --git a/services/competitor_match_attempt_rescore_audit.py b/services/competitor_match_attempt_rescore_audit.py index ad544a5..814ca18 100644 --- a/services/competitor_match_attempt_rescore_audit.py +++ b/services/competitor_match_attempt_rescore_audit.py @@ -6,6 +6,7 @@ from __future__ import annotations from collections import Counter from dataclasses import asdict, dataclass +import json from typing import Any, Iterable, Sequence from sqlalchemy import bindparam, text @@ -20,6 +21,7 @@ DEFAULT_RESCAN_STATUSES = ( "low_score", "refresh_low_score", ) +RESCORE_ACCEPTED_CURRENT_STATUS = "rescore_accepted_current" @dataclass(frozen=True) @@ -174,6 +176,156 @@ def summarize_match_attempt_rescore( } +def _json_expr(conn, bind_name: str) -> str: + return f"CAST(:{bind_name} AS jsonb)" if conn.dialect.name == "postgresql" else f":{bind_name}" + + +def _diagnostic_payload(decision: MatchAttemptRescoreDecision) -> dict[str, Any]: + return { + "rescore_version": "current_matcher_v1", + "stored_status": decision.stored_status, + "stored_score": decision.stored_score, + "score": decision.current_score, + "hard_veto": decision.hard_veto, + "comparison_mode": decision.comparison_mode, + "match_type": decision.match_type, + "price_basis": decision.price_basis, + "alert_tier": decision.alert_tier, + "reasons": decision.reasons, + } + + +def _diagnostic_text(decision: MatchAttemptRescoreDecision) -> str: + reasons = ",".join(decision.reasons or []) + return ( + "matcher_rescore=accepted_current; " + f"stored_status={decision.stored_status}; " + f"stored_score={decision.stored_score}; " + f"current_score={decision.current_score}; " + f"mode={decision.comparison_mode}; " + f"price_basis={decision.price_basis}; " + f"reasons={reasons}" + ) + + +def _ensure_attempt_table(conn) -> None: + from services.competitor_price_feeder import CompetitorPriceFeeder + + CompetitorPriceFeeder(engine=None)._ensure_competitor_match_attempts_table(conn) + + +def _already_materialized(conn, *, source: str, sku: str, candidate_id: str) -> bool: + row = conn.execute(text(""" + SELECT 1 + FROM competitor_match_attempts + WHERE sku = :sku + AND source = :source + AND attempt_status = :attempt_status + AND COALESCE(best_competitor_product_id, '') = :candidate_id + LIMIT 1 + """), { + "sku": sku, + "source": source, + "attempt_status": RESCORE_ACCEPTED_CURRENT_STATUS, + "candidate_id": candidate_id, + }).first() + return row is not None + + +def materialize_rescore_accept_reviews( + conn, + rows: Iterable[dict[str, Any]], + *, + source: str = "pchome", + min_score: float = MIN_MATCH_SCORE, +) -> dict[str, Any]: + """Append accepted-current rescore rows to the manual review queue. + + This never writes ``competitor_prices``. It only creates a latest + ``competitor_match_attempts`` row that Dashboard can route through the + existing human review buttons. + """ + _ensure_attempt_table(conn) + stats = { + "scanned": 0, + "materialized": 0, + "skipped_not_gate_pass": 0, + "skipped_missing_candidate": 0, + "skipped_duplicate": 0, + "samples": [], + } + search_terms_expr = _json_expr(conn, "search_terms") + diagnostic_json_expr = _json_expr(conn, "match_diagnostic_json") + diagnostic_codes_expr = _json_expr(conn, "diagnostic_codes") + + for row in rows: + stats["scanned"] += 1 + decision = classify_match_attempt_row(row, min_score=min_score) + if not decision.gate_pass: + stats["skipped_not_gate_pass"] += 1 + continue + + candidate_id = str(row.get("best_competitor_product_id") or "").strip() + candidate_name = str(row.get("best_competitor_product_name") or "").strip() + candidate_price = _to_float(row.get("best_competitor_price")) + if not candidate_id or not candidate_name or not candidate_price or candidate_price <= 0: + stats["skipped_missing_candidate"] += 1 + continue + + if _already_materialized(conn, source=source, sku=decision.sku, candidate_id=candidate_id): + stats["skipped_duplicate"] += 1 + continue + + diagnostic_payload = _diagnostic_payload(decision) + conn.execute(text(f""" + INSERT INTO competitor_match_attempts + (sku, source, momo_product_id, momo_product_name, momo_price, + search_terms, candidate_count, attempt_status, + best_competitor_product_id, best_competitor_product_name, + best_competitor_price, best_match_score, error_message, + attempted_at, competitor_product_url, competitor_image_url, + competitor_stock, match_diagnostic_json, comparison_mode, + hard_veto, diagnostic_codes) + VALUES + (:sku, :source, :momo_product_id, :momo_product_name, :momo_price, + {search_terms_expr}, :candidate_count, :attempt_status, + :best_id, :best_name, + :best_price, :best_score, :error_message, + CURRENT_TIMESTAMP, :competitor_product_url, :competitor_image_url, + :competitor_stock, {diagnostic_json_expr}, :comparison_mode, + :hard_veto, {diagnostic_codes_expr}) + """), { + "sku": decision.sku, + "source": source, + "momo_product_id": row.get("momo_product_id"), + "momo_product_name": decision.momo_product_name, + "momo_price": _to_float(row.get("momo_price")), + "search_terms": json.dumps([ + "matcher_rescore:accepted_current", + f"stored_status:{decision.stored_status}", + ], ensure_ascii=False), + "candidate_count": int(row.get("candidate_count") or 1), + "attempt_status": RESCORE_ACCEPTED_CURRENT_STATUS, + "best_id": candidate_id, + "best_name": candidate_name[:300], + "best_price": candidate_price, + "best_score": decision.current_score, + "error_message": _diagnostic_text(decision)[:1000], + "competitor_product_url": row.get("competitor_product_url"), + "competitor_image_url": row.get("competitor_image_url"), + "competitor_stock": row.get("competitor_stock"), + "match_diagnostic_json": json.dumps(diagnostic_payload, ensure_ascii=False), + "comparison_mode": decision.comparison_mode, + "hard_veto": False, + "diagnostic_codes": json.dumps(decision.reasons, ensure_ascii=False), + }) + stats["materialized"] += 1 + if len(stats["samples"]) < 10: + stats["samples"].append(decision.to_dict()) + + return stats + + def fetch_match_attempt_rescore_rows( conn, *, @@ -191,12 +343,17 @@ def fetch_match_attempt_rescore_rows( SELECT DISTINCT ON (sku, best_competitor_product_id) sku, attempt_status, + momo_product_id, momo_product_name, momo_price, + candidate_count, best_competitor_product_id, best_competitor_product_name, best_competitor_price, best_match_score, + competitor_product_url, + competitor_image_url, + competitor_stock, diagnostic_codes, attempted_at FROM competitor_match_attempts @@ -213,12 +370,17 @@ def fetch_match_attempt_rescore_rows( SELECT sku, attempt_status, + momo_product_id, momo_product_name, momo_price, + candidate_count, best_competitor_product_id, best_competitor_product_name, best_competitor_price, best_match_score, + competitor_product_url, + competitor_image_url, + competitor_stock, diagnostic_codes, attempted_at, ROW_NUMBER() OVER ( diff --git a/tests/test_competitor_identity_revalidator.py b/tests/test_competitor_identity_revalidator.py index a9becc6..ce3f091 100644 --- a/tests/test_competitor_identity_revalidator.py +++ b/tests/test_competitor_identity_revalidator.py @@ -145,6 +145,20 @@ def test_dashboard_match_status_shows_manual_review_closure_states(): assert decision["gap_amount"] is None +def test_dashboard_match_status_shows_rescore_accepted_review_state(): + from routes.dashboard_routes import _build_pchome_match_status + + status = _build_pchome_match_status({ + "attempt_status": "rescore_accepted_current", + "best_match_score": 0.801, + "error_message": "matcher_rescore=accepted_current; reasons=strong_exact_spec_match", + }) + + assert status["label"] == "重算可採用待審" + assert status["tone"] == "watch" + assert "人工採用後才寫入正式 PChome 價差" in status["summary"] + + def test_dashboard_match_status_uses_specific_matcher_reason_labels(): from routes.dashboard_routes import _build_pchome_match_status diff --git a/tests/test_competitor_match_attempt_rescore_audit.py b/tests/test_competitor_match_attempt_rescore_audit.py index 104cd8f..6a226f0 100644 --- a/tests/test_competitor_match_attempt_rescore_audit.py +++ b/tests/test_competitor_match_attempt_rescore_audit.py @@ -1,3 +1,8 @@ +import json + +from sqlalchemy import create_engine, text + + def test_match_attempt_rescore_audit_classifies_current_gate_pass_and_veto(): from services.competitor_match_attempt_rescore_audit import summarize_match_attempt_rescore @@ -61,3 +66,66 @@ def test_match_attempt_rescore_audit_skips_missing_identity_text(): assert decision.suggested_status == "skipped_missing_identity_text" assert decision.current_score is None assert decision.gate_pass is False + + +def test_match_attempt_rescore_materializes_accepted_current_for_manual_review(): + from services.competitor_match_attempt_rescore_audit import materialize_rescore_accept_reviews + + engine = create_engine("sqlite:///:memory:") + rows = [{ + "sku": "9031334", + "attempt_status": "true_low_confidence", + "momo_product_id": 9031334, + "momo_product_name": "【Mustela 慕之恬廊】慕之幼 加量版爽身潤膚乳 500mlX2入(寶寶 嬰兒乳液 公司貨 台灣獨家總代理)", + "momo_price": 1390, + "candidate_count": 1, + "best_competitor_product_id": "DDAXS-A900IHGY", + "best_competitor_product_name": "【慕之恬廊】慕之幼爽身潤膚乳(500毫升X2入)", + "best_competitor_price": 1210, + "best_match_score": 0.709, + }] + + with engine.begin() as conn: + stats = materialize_rescore_accept_reviews(conn, rows) + duplicate_stats = materialize_rescore_accept_reviews(conn, rows) + stored = conn.execute(text(""" + SELECT attempt_status, best_match_score, search_terms, + match_diagnostic_json, diagnostic_codes, error_message + FROM competitor_match_attempts + """)).mappings().all() + + assert stats["materialized"] == 1 + assert duplicate_stats["skipped_duplicate"] == 1 + assert len(stored) == 1 + assert stored[0]["attempt_status"] == "rescore_accepted_current" + assert float(stored[0]["best_match_score"]) >= 0.76 + assert "matcher_rescore:accepted_current" in json.loads(stored[0]["search_terms"]) + assert "shared_identity_anchor_reordered_line" in json.loads(stored[0]["diagnostic_codes"]) + payload = json.loads(stored[0]["match_diagnostic_json"]) + assert payload["stored_status"] == "true_low_confidence" + assert payload["price_basis"] == "manual_review" + assert "matcher_rescore=accepted_current" in stored[0]["error_message"] + + +def test_match_attempt_rescore_materialize_does_not_enqueue_veto(): + from services.competitor_match_attempt_rescore_audit import materialize_rescore_accept_reviews + + engine = create_engine("sqlite:///:memory:") + rows = [{ + "sku": "12534698", + "attempt_status": "true_low_confidence", + "momo_product_name": "【LANCOME 蘭蔻】官方直營 超極光活粹晶露150ml(LANCOME/四重酸極光水/化妝水/精華水)", + "best_competitor_product_id": "LANCOME-BAD", + "best_competitor_product_name": "LANCOME 蘭蔻 超極限肌因精華露150ml 專櫃公司貨", + "momo_price": 3555, + "best_competitor_price": 1880, + "best_match_score": 0.748, + }] + + with engine.begin() as conn: + stats = materialize_rescore_accept_reviews(conn, rows) + row_count = conn.execute(text("SELECT COUNT(*) FROM competitor_match_attempts")).scalar() + + assert stats["materialized"] == 0 + assert stats["skipped_not_gate_pass"] == 1 + assert row_count == 0