From 9d84cbfd43c7d9e01265eb36e4e6ed7b879cf50e Mon Sep 17 00:00:00 2001 From: OoO Date: Fri, 19 Jun 2026 00:41:20 +0800 Subject: [PATCH] feat: deepen pchome momo backfill guardrails --- config.py | 2 +- services/external_market_offer_service.py | 30 ++++++++++ services/marketplace_product_matcher.py | 1 + services/momo_crawler.py | 14 ++++- .../pchome_growth_momo_backfill_service.py | 38 ++++++++++++- tests/test_external_market_offer_service.py | 36 ++++++++++++ tests/test_marketplace_product_matcher.py | 14 +++++ tests/test_momo_crawler_targeted_search.py | 55 +++++++++++++++++++ tests/test_pchome_revenue_growth_service.py | 24 ++++++++ 9 files changed, 208 insertions(+), 6 deletions(-) diff --git a/config.py b/config.py index 587fbc5..51f1675 100644 --- a/config.py +++ b/config.py @@ -402,7 +402,7 @@ YOUTUBE_API_KEY = os.getenv('YOUTUBE_API_KEY', '') # ========================================== # 系統版本與路徑 # ========================================== -SYSTEM_VERSION = "V10.635" +SYSTEM_VERSION = "V10.636" LOG_FILE_PATH = os.path.join(BASE_DIR, 'logs/system.log') public_url = PUBLIC_URL # 用於模板顯示 diff --git a/services/external_market_offer_service.py b/services/external_market_offer_service.py index 7b47407..41cc8f6 100644 --- a/services/external_market_offer_service.py +++ b/services/external_market_offer_service.py @@ -684,6 +684,34 @@ def _targeted_candidate_auto_type(candidate: dict[str, Any]) -> str: return "manual_review" +def _targeted_candidate_needs_review(candidate: dict[str, Any]) -> bool: + """總價自動同步前的最後防線，避免高分但款式待確認的候選進作戰清單。""" + if candidate.get("target_hard_veto") is True: + return True + price_basis = str(candidate.get("target_price_basis") or "").strip() + alert_tier = str(candidate.get("target_alert_tier") or "").strip() + if price_basis and price_basis != "total_price": + return True + if alert_tier and alert_tier != "price_alert_exact": + return True + review_reason_markers = { + "manual_review", + "identity_review", + "unit_price_review", + "variant_selection_review", + "variant_option_conflict", + "variant_descriptor_conflict", + "makeup_catalog_selection_gap", + "commercial_condition_gap", + "count_conflict", + "bundle_offer_conflict", + "multi_component_conflict", + "component_count_conflict", + } + reasons = {str(reason or "") for reason in (candidate.get("target_match_reasons") or [])} + return bool(reasons & review_reason_markers) + + def _targeted_candidate_to_external_offer( candidate: dict[str, Any], *, @@ -692,6 +720,8 @@ def _targeted_candidate_to_external_offer( auto_type = _targeted_candidate_auto_type(candidate) if auto_type not in {"total_price", "unit_price"}: return None, "不是可自動使用的候選" + if auto_type == "total_price" and _targeted_candidate_needs_review(candidate): + return None, "候選仍需人工確認" momo_sku = str(candidate.get("product_id") or candidate.get("goodsCode") or candidate.get("id") or "").strip() pchome_product_id = str(candidate.get("target_pchome_product_id") or "").strip() diff --git a/services/marketplace_product_matcher.py b/services/marketplace_product_matcher.py index 272b91e..fe1c64f 100644 --- a/services/marketplace_product_matcher.py +++ b/services/marketplace_product_matcher.py @@ -712,6 +712,7 @@ VARIANT_OPTION_COLOR_WORDS = { "棕色", "咖啡色", "灰色", + "rose", "白色", "紅色", "粉色", diff --git a/services/momo_crawler.py b/services/momo_crawler.py index d919add..e3d2bac 100644 --- a/services/momo_crawler.py +++ b/services/momo_crawler.py @@ -659,16 +659,24 @@ def search_momo_products_for_pchome_products( continue hard_veto = bool(getattr(diagnostics, "hard_veto", False)) comparison_mode = getattr(diagnostics, "comparison_mode", "exact_identity") + diagnostic_price_basis = str(getattr(diagnostics, "price_basis", "") or "") + diagnostic_alert_tier = str(getattr(diagnostics, "alert_tier", "") or "") + diagnostic_match_type = str(getattr(diagnostics, "match_type", "") or "") unit_price_comparison = {} auto_compare_type = "manual_review" price_basis = "none" review_status = "需人工確認" - if not hard_veto and comparison_mode == "exact_identity": + if ( + not hard_veto + and comparison_mode == "exact_identity" + and diagnostic_price_basis == "total_price" + and diagnostic_alert_tier == "price_alert_exact" + ): can_auto_compare = True auto_compare_type = "total_price" price_basis = "total_price" review_status = "可直接比價" - elif comparison_mode == "unit_comparable": + elif comparison_mode == "unit_comparable" or diagnostic_price_basis == "unit_price": unit_price_comparison = build_unit_price_comparison( momo_name, pchome_name, @@ -713,6 +721,8 @@ def search_momo_products_for_pchome_products( "target_search_term": term, "target_match_reasons": list(getattr(diagnostics, "reasons", ()) or ()), "target_comparison_mode": comparison_mode, + "target_match_type": diagnostic_match_type, + "target_alert_tier": diagnostic_alert_tier, "target_hard_veto": hard_veto, "can_auto_compare": can_auto_compare, "auto_compare_type": auto_compare_type, diff --git a/services/pchome_growth_momo_backfill_service.py b/services/pchome_growth_momo_backfill_service.py index 5206fe6..c1d787b 100644 --- a/services/pchome_growth_momo_backfill_service.py +++ b/services/pchome_growth_momo_backfill_service.py @@ -4,9 +4,26 @@ from __future__ import annotations +import os from typing import Any, Callable +def _int_env(name: str, default: int, *, minimum: int, maximum: int) -> int: + try: + value = int(os.getenv(name, str(default))) + except (TypeError, ValueError): + value = default + return max(minimum, min(value, maximum)) + + +def _float_env(name: str, default: float, *, minimum: float, maximum: float) -> float: + try: + value = float(os.getenv(name, str(default))) + except (TypeError, ValueError): + value = default + return max(minimum, min(value, maximum)) + + def candidate_auto_compare_type(candidate: dict[str, Any]) -> str: auto_type = str(candidate.get("auto_compare_type") or "").strip() if auto_type in {"total_price", "unit_price"}: @@ -53,9 +70,24 @@ def _default_search_candidates(targets: list[dict[str, Any]], limit: int): return search_momo_products_for_pchome_products( targets, max_products=limit, - limit_per_product=6, - max_terms_per_product=4, - min_score=0.45, + limit_per_product=_int_env( + "PCHOME_GROWTH_MOMO_BACKFILL_LIMIT_PER_TERM", + 8, + minimum=3, + maximum=12, + ), + max_terms_per_product=_int_env( + "PCHOME_GROWTH_MOMO_BACKFILL_MAX_TERMS", + 8, + minimum=3, + maximum=10, + ), + min_score=_float_env( + "PCHOME_GROWTH_MOMO_BACKFILL_MIN_SCORE", + 0.45, + minimum=0.35, + maximum=0.8, + ), ) diff --git a/tests/test_external_market_offer_service.py b/tests/test_external_market_offer_service.py index 7e4c7c0..1b32a83 100644 --- a/tests/test_external_market_offer_service.py +++ b/tests/test_external_market_offer_service.py @@ -325,6 +325,42 @@ def test_sync_targeted_momo_candidates_writes_unit_price_offer(monkeypatch): assert stale_marks == [True] +def test_sync_targeted_momo_candidates_skips_total_price_identity_review(monkeypatch): + from services import external_market_offer_service as service + + stale_marks = [] + monkeypatch.setattr(service, "mark_pchome_growth_cache_stale", lambda: stale_marks.append(True)) + + engine = create_engine("sqlite:///:memory:") + _seed_external_offer_sync_tables(engine) + + payload = service.sync_targeted_momo_candidates_to_external_offers(engine, [ + { + "product_id": "12876190", + "name": "【LAURA MERCIER 蘿拉蜜思】煥顏透明蜜粉 29g(#Rose-國際航空版)", + "price": 809, + "target_pchome_product_id": "PCH-LAURA", + "target_pchome_name": "【Laura Mercier 蘿拉蜜思】煥顏透明蜜粉 29g", + "target_pchome_price": 899, + "target_match_score": 0.98, + "auto_compare_type": "total_price", + "target_price_basis": "manual_review", + "target_alert_tier": "identity_review", + "target_match_reasons": ["variant_selection_review", "strong_exact_spec_match"], + "target_comparison_mode": "exact_identity", + }, + ]) + + assert payload["success"] is True + assert payload["candidate_count"] == 1 + assert payload["written_count"] == 0 + assert payload["skipped_reasons"] == {"候選仍需人工確認": 1} + assert stale_marks == [] + with engine.connect() as conn: + count = conn.execute(text("SELECT COUNT(*) FROM external_offers")).scalar() + assert count == 0 + + def test_external_source_readiness_uses_legacy_momo_reference_cache(): from services.external_market_offer_service import build_external_source_readiness diff --git a/tests/test_marketplace_product_matcher.py b/tests/test_marketplace_product_matcher.py index 8a641f0..e047eeb 100644 --- a/tests/test_marketplace_product_matcher.py +++ b/tests/test_marketplace_product_matcher.py @@ -1467,6 +1467,20 @@ def test_marketplace_matcher_promotes_recent_stale_recovery_single_spec_lines(): assert diagnostics.alert_tier == "price_alert_exact" +def test_marketplace_matcher_keeps_single_sided_rose_powder_in_identity_review(): + from services.marketplace_product_matcher import score_marketplace_match + + diagnostics = score_marketplace_match( + "【LAURA MERCIER 蘿拉蜜思】煥顏透明蜜粉 Rose 29g (國際航空版)", + "【Laura Mercier 蘿拉蜜思】煥顏透明蜜粉 29g", + ) + + assert diagnostics.hard_veto is False + assert diagnostics.price_basis == "manual_review" + assert diagnostics.alert_tier == "identity_review" + assert "variant_selection_review" in diagnostics.reasons + + def test_marketplace_matcher_keeps_baby_shampoo_bath_gap_in_identity_review(): from services.marketplace_product_matcher import score_marketplace_match diff --git a/tests/test_momo_crawler_targeted_search.py b/tests/test_momo_crawler_targeted_search.py index 80cdb72..5056095 100644 --- a/tests/test_momo_crawler_targeted_search.py +++ b/tests/test_momo_crawler_targeted_search.py @@ -128,3 +128,58 @@ def test_targeted_momo_search_auto_routes_unit_comparable_candidates(): assert products[0]["target_unit_price_comparison"]["comparable"] is True assert products[0]["target_unit_price_comparison"]["unit_label"] == "ml" assert products[0]["target_gap_pct"] < 0 + + +def test_targeted_momo_search_keeps_identity_review_candidates_manual(monkeypatch): + from services.momo_crawler import MomoProduct, search_momo_products_for_pchome_products + + class FakeDiagnostics: + score = 0.98 + hard_veto = False + comparison_mode = "exact_identity" + price_basis = "manual_review" + alert_tier = "identity_review" + match_type = "exact" + reasons = ("variant_selection_review", "strong_exact_spec_match") + + class FakeCrawler: + def search_products(self, keyword, limit=10, sort_by="sSaleQty/dc"): + return True, "ok", [ + MomoProduct( + product_id="12876190", + name="【LAURA MERCIER 蘿拉蜜思】煥顏透明蜜粉 29g(#Rose-國際航空版)", + price=809, + original_price=999, + discount=19, + image_url="", + product_url="https://www.momoshop.com.tw/goods/GoodsDetail.jsp?i_code=12876190", + brand="蘿拉蜜思", + crawled_at=datetime.now(), + ) + ] + + monkeypatch.setattr( + "services.marketplace_product_matcher.score_marketplace_match", + lambda *args, **kwargs: FakeDiagnostics(), + ) + + success, message, products = search_momo_products_for_pchome_products( + [ + { + "product_id": "PCH-LAURA", + "name": "【Laura Mercier 蘿拉蜜思】煥顏透明蜜粉 29g", + "price": 899, + } + ], + crawler=FakeCrawler(), + max_terms_per_product=1, + limit_per_product=3, + min_score=0.45, + ) + + assert success is True + assert "需人工確認 1 筆" in message + assert products[0]["auto_compare_type"] == "manual_review" + assert products[0]["can_auto_compare"] is False + assert products[0]["target_price_basis"] == "none" + assert products[0]["target_alert_tier"] == "identity_review" diff --git a/tests/test_pchome_revenue_growth_service.py b/tests/test_pchome_revenue_growth_service.py index 8c9d236..b88b665 100644 --- a/tests/test_pchome_revenue_growth_service.py +++ b/tests/test_pchome_revenue_growth_service.py @@ -328,6 +328,30 @@ def test_pchome_growth_momo_backfill_service_targets_unmapped_high_sales_items() assert captured["search_limit"] == 2 +def test_pchome_growth_momo_backfill_default_search_uses_deeper_terms(monkeypatch): + from services import pchome_growth_momo_backfill_service as service + + captured = {} + + def fake_search(targets, **kwargs): + captured["targets"] = targets + captured["kwargs"] = kwargs + return True, "ok", [] + + monkeypatch.delenv("PCHOME_GROWTH_MOMO_BACKFILL_LIMIT_PER_TERM", raising=False) + monkeypatch.delenv("PCHOME_GROWTH_MOMO_BACKFILL_MAX_TERMS", raising=False) + monkeypatch.delenv("PCHOME_GROWTH_MOMO_BACKFILL_MIN_SCORE", raising=False) + monkeypatch.setattr("services.momo_crawler.search_momo_products_for_pchome_products", fake_search) + + payload = service._default_search_candidates([{"product_id": "PCH-1", "name": "商品"}], limit=3) + + assert payload == (True, "ok", []) + assert captured["kwargs"]["max_products"] == 3 + assert captured["kwargs"]["limit_per_product"] == 8 + assert captured["kwargs"]["max_terms_per_product"] == 8 + assert captured["kwargs"]["min_score"] == 0.45 + + def test_pchome_growth_momo_backfill_route_calls_shared_service(monkeypatch): from flask import Flask from routes import ai_routes as routes