強化 PChome 變體錯配防線

services/marketplace_product_matcher.py

@@ -475,6 +475,7 @@ FOCUSED_IDENTITY_REVIEW_ONLY_REASONS = {
     "johnsons_baby_lotion_variant_catalog",
     "im_meme_fixx_cool_setting_spray",
     "so_natural_fixx_setting_spray_catalog",
+    "kate_powder_case_catalog",
     "kate_monster_lipstick_catalog",
     "opi_gel_polish_series_catalog",
     "romand_juicy_lip_tint_2_catalog",
@@ -527,6 +528,8 @@ VARIANT_SENSITIVE_KEYWORDS = {
     "私密清潔凝露",
     "私密潔淨凝露",
     "私密淨白清潔凝露",
+    "私密防護慕絲",
+    "慕絲",
     "定妝噴霧",
     "妝前防護乳",
     "妝前乳",
@@ -543,8 +546,13 @@ VARIANT_SENSITIVE_KEYWORDS = {
     "腮紅液",
     "打亮液",
     "蜜粉餅",
+    "粉餅盒",
     "粉底棒",
     "遮瑕棒",
+    "遮瑕蜜",
+    "護手霜",
+    "滋養霜",
+    "修護乳",
     "修容打亮棒",
     "防曬",
     "防曬乳",
@@ -556,6 +564,8 @@ VARIANT_SENSITIVE_KEYWORDS = {
 VARIANT_OPTION_COLOR_WORDS = {
     "茉莉花",
     "梔子花",
+    "白茶蘭花",
+    "白茶",
     "白麝香",
     "黑麝香",
     "清新花園",
@@ -563,8 +573,18 @@ VARIANT_OPTION_COLOR_WORDS = {
     "青檸羅勒",
     "炭木香",
     "無花果",
+    "鼠尾草",
+    "海鹽",
+    "檸檬草",
+    "茶樹",
+    "櫻花",
+    "繡球花",
+    "魔髮奇緣",
     "清甜柚香",
     "杏仁牛奶",
+    "杏仁",
+    "薄荷",
+    "橙花",
     "完熟白桃",
     "琥珀橙",
     "干邑棕",
@@ -603,6 +623,12 @@ VARIANT_OPTION_COLOR_WORDS = {
     "自然色",
     "明亮色",
     "透明色",
+    "清爽型",
+    "滋潤型",
+    "橡棕",
+    "暗灰",
+    "灰棕",
+    "淺玫粉",
     "極光之藍",
     "月光銀影",
 }
@@ -1976,6 +2002,15 @@ def score_marketplace_match(
     aroma_scent_variant_conflict = _has_aroma_scent_variant_conflict(left, right)
     if aroma_scent_variant_conflict:
         reasons.append("aroma_scent_variant_conflict")
+    ingredient_line_conflict = _has_core_ingredient_line_conflict(left, right)
+    if ingredient_line_conflict:
+        reasons.append("core_ingredient_line_conflict")
+    branded_powder_line_conflict = _has_branded_powder_line_conflict(left, right)
+    if branded_powder_line_conflict:
+        reasons.append("branded_powder_line_conflict")
+    cleanser_lotion_line_conflict = _has_cleanser_lotion_line_conflict(left, right)
+    if cleanser_lotion_line_conflict:
+        reasons.append("cleanser_lotion_line_conflict")
     wax_lamp_size_letter_conflict = _has_wax_lamp_size_letter_conflict(left, right)
     if wax_lamp_size_letter_conflict:
         reasons.append("size_letter_variant_conflict")
@@ -2043,6 +2078,12 @@ def score_marketplace_match(
         hard_veto = True
     if aroma_scent_variant_conflict:
         hard_veto = True
+    if ingredient_line_conflict:
+        hard_veto = True
+    if branded_powder_line_conflict:
+        hard_veto = True
+    if cleanser_lotion_line_conflict:
+        hard_veto = True
     if wax_lamp_size_letter_conflict:
         hard_veto = True
     if nitori_diffuser_model_conflict:
@@ -3051,7 +3092,26 @@ def _has_hoi_candle_line_conflict(left: ProductIdentity, right: ProductIdentity)
 
 def _has_aroma_scent_variant_conflict(left: ProductIdentity, right: ProductIdentity) -> bool:
     pair_text = f"{left.searchable_name} {right.searchable_name}"
-    if not any(term in pair_text for term in ("香氛固體凝膠", "香氛凝膠", "空氣芳香劑", "車用香氛")):
+    if any(term in pair_text for term in ("護手霜", "融蠟燈", "蠟燭暖燈")):
+        return False
+    if not any(
+        term in pair_text
+        for term in (
+            "香氛固體凝膠",
+            "香氛凝膠",
+            "空氣芳香劑",
+            "車用香氛",
+            "車用擴香",
+            "擴香蕊",
+            "擴香罐",
+            "香薰蠟燭",
+            "香氛蠟燭",
+            "蠟燭",
+            "滾珠精油",
+            "香氛精油",
+            "植物精油",
+        )
+    ):
         return False
     if _is_multi_variant_catalog_listing(left) or _is_multi_variant_catalog_listing(right):
         return False
@@ -3070,8 +3130,19 @@ def _has_aroma_scent_variant_conflict(left: ProductIdentity, right: ProductIdent
         "青檸羅勒",
         "炭木香",
         "無花果",
+        "白茶蘭花",
+        "白茶",
+        "檸檬草",
+        "茶樹",
+        "鼠尾草",
+        "海鹽",
+        "橙花",
+        "薄荷",
+        "杏仁",
         "薰衣草",
         "茉莉",
+        "櫻花",
+        "繡球花",
         "玫瑰",
         "雪松",
         "檀香",
@@ -3085,6 +3156,52 @@ def _has_aroma_scent_variant_conflict(left: ProductIdentity, right: ProductIdent
     return False
 
 
+def _has_core_ingredient_line_conflict(left: ProductIdentity, right: ProductIdentity) -> bool:
+    pair_text = f"{left.searchable_name} {right.searchable_name}"
+    if not any(term in pair_text for term in ("油膏", "護膚油", "身體油", "精油", "霜", "乳霜")):
+        return False
+    ingredient_groups = {
+        "coconut_oil": ("椰子油", "coconut"),
+        "shea_butter": ("乳木果油", "shea"),
+    }
+    left_groups = {
+        group
+        for group, terms in ingredient_groups.items()
+        if any(term in left.searchable_name for term in terms)
+    }
+    right_groups = {
+        group
+        for group, terms in ingredient_groups.items()
+        if any(term in right.searchable_name for term in terms)
+    }
+    return bool(left_groups and right_groups and not (left_groups & right_groups))
+
+
+def _has_branded_powder_line_conflict(left: ProductIdentity, right: ProductIdentity) -> bool:
+    if not ({"港香蘭"} & (left.brand_tokens & right.brand_tokens)):
+        return False
+    if "爽身粉" not in left.searchable_name or "爽身粉" not in right.searchable_name:
+        return False
+    named_lines = ("漢本", "艾魔菈")
+    left_lines = {line for line in named_lines if line in left.searchable_name}
+    right_lines = {line for line in named_lines if line in right.searchable_name}
+    return bool(left_lines and right_lines and not (left_lines & right_lines))
+
+
+def _has_cleanser_lotion_line_conflict(left: ProductIdentity, right: ProductIdentity) -> bool:
+    if not (left.brand_tokens & right.brand_tokens):
+        return False
+    if not _has_overlapping_base_spec(left, right):
+        return False
+    cleanser_terms = ("潔膚露", "潔膚", "潔淨露", "潔面", "洗面乳", "cleanser")
+    lotion_terms = ("修護乳", "乳液", "身體乳", "潤膚乳", "lotion")
+    left_cleanser = any(term in left.searchable_name for term in cleanser_terms)
+    right_cleanser = any(term in right.searchable_name for term in cleanser_terms)
+    left_lotion = any(term in left.searchable_name for term in lotion_terms)
+    right_lotion = any(term in right.searchable_name for term in lotion_terms)
+    return bool((left_cleanser and right_lotion) or (right_cleanser and left_lotion))
+
+
 def _standalone_size_letter_tokens(identity: ProductIdentity) -> set[str]:
     text = identity.searchable_name
     return {
@@ -3174,17 +3291,38 @@ def _has_catalog_specific_variant_selection_gap(left: ProductIdentity, right: Pr
             "美體主張",
             "私密潔浴露",
             "私密潔浴",
+            "私密防護慕絲",
+            "私密慕絲",
+            "慕絲",
             "嬰兒潤膚乳",
             "定妝噴霧",
             "染眉膏",
+            "眼線膠筆",
+            "粉餅盒",
+            "遮瑕蜜",
+            "護手霜",
+            "護唇膏",
+            "護唇棒",
             "唇釉",
+            "唇膏",
             "蜜粉",
             "防曬素顏霜",
+            "車用香氛",
+            "車用擴香",
+            "車用擴香蕊",
+            "香氛擴香罐",
+            "擴香罐",
+            "擴香蕊",
+            "水性指甲油",
+            "指甲油",
+            "融蠟小夜燈",
+            "融蠟燈",
+            "滋養霜",
         )
     ):
         return False
-    left_catalog = _is_multi_variant_catalog_listing(left)
-    right_catalog = _is_multi_variant_catalog_listing(right)
+    left_catalog = _is_catalog_or_delimited_variant_listing(left)
+    right_catalog = _is_catalog_or_delimited_variant_listing(right)
     return left_catalog != right_catalog
 
 
@@ -3560,6 +3698,12 @@ def _has_focused_low_score_exact_identity_line(left: ProductIdentity, right: Pro
         and "全天候超完美定妝噴霧" in right_text
     ):
         return "so_natural_fixx_setting_spray_catalog"
+    if (
+        {"kate", "凱婷"} & (left.brand_tokens & right.brand_tokens)
+        and "粉餅盒" in left_text
+        and "粉餅盒" in right_text
+    ):
+        return "kate_powder_case_catalog"
     if (
         {"kate", "凱婷"} & (left.brand_tokens & right.brand_tokens)
         and "怪獸級持色唇膏" in left_text
@@ -3749,6 +3893,42 @@ def _is_multi_variant_catalog_listing(identity: ProductIdentity) -> bool:
     return any(phrase in text for phrase in MULTI_VARIANT_LISTING_PHRASES)
 
 
+def _normalize_variant_option(value: str) -> set[str]:
+    compact = re.sub(r"[^a-z0-9]", "", (value or "").lower())
+    if not compact:
+        return set()
+    return {compact}
+
+
+def _variant_option_compare_key(option: str) -> str:
+    if option.isdigit():
+        return option.lstrip("0") or "0"
+    return option
+
+
+def _variant_options_overlap(left_options: set[str], right_options: set[str]) -> bool:
+    if left_options & right_options:
+        return True
+    left_keys = {_variant_option_compare_key(option) for option in left_options}
+    right_keys = {_variant_option_compare_key(option) for option in right_options}
+    return bool(left_keys & right_keys)
+
+
+def _is_catalog_or_delimited_variant_listing(identity: ProductIdentity) -> bool:
+    if _is_multi_variant_catalog_listing(identity):
+        return True
+    text = identity.searchable_name
+    if re.search(r"(?<![a-z0-9])([a-z]?\d{1,3}[a-z]?)\s*(?:~|～|至|-)\s*([a-z]?\d{1,3}[a-z]?)(?![a-z0-9])", text, re.I):
+        return True
+    options = _explicit_variant_option_tokens(identity)
+    if len(options) < 2:
+        return bool(
+            re.search(r"[/／、,，.．&＆]", text)
+            and any(term in text for term in ("粉餅盒", "眼線膠筆", "眉筆", "唇膏", "唇釉", "遮瑕蜜", "車用擴香", "車用香氛"))
+        )
+    return bool(re.search(r"[/／、,，.．&＆]", text))
+
+
 def _has_catalog_variant_listing_alignment(left: ProductIdentity, right: ProductIdentity) -> bool:
     if not (_is_multi_variant_catalog_listing(left) and _is_multi_variant_catalog_listing(right)):
         return False
@@ -3823,16 +4003,15 @@ def _has_variant_descriptor_conflict(left: ProductIdentity, right: ProductIdenti
 def _explicit_variant_option_tokens(identity: ProductIdentity) -> set[str]:
     text = identity.searchable_name
     options: set[str] = set()
+    for match in re.finditer(r"(?<![a-z0-9])([a-z]?\d{1,3}[a-z]?)\s*(?:~|～|至|-)\s*([a-z]?\d{1,3}[a-z]?)(?![a-z0-9])", text, re.I):
+        for group in (match.group(1), match.group(2)):
+            options.update(_normalize_variant_option(group))
     for match in re.finditer(r"(?:#|no\.?|色號|號色)\s*([a-z]?\d{1,3}[a-z]?)(?![a-z0-9])", text, re.I):
-        value = re.sub(r"[^a-z0-9]", "", match.group(1).lower())
-        if value:
-            options.add(value)
+        options.update(_normalize_variant_option(match.group(1)))
     for match in re.finditer(r"(?<![a-z0-9])((?:0?\d){1,2})(?=[\u4e00-\u9fff])", text, re.I):
         if text[match.end(1):match.end(1) + 4] in {"號護唇膏", "號護脣膏"}:
             continue
-        value = re.sub(r"[^a-z0-9]", "", match.group(1).lower())
-        if value:
-            options.add(value)
+        options.update(_normalize_variant_option(match.group(1)))
     for color_word in VARIANT_OPTION_COLOR_WORDS:
         if color_word in text:
             options.add(color_word)
@@ -3891,9 +4070,21 @@ def _has_explicit_variant_option_conflict(
         return False
     if left_options == right_options:
         return False
-    if left_options & right_options:
+    if _variant_options_overlap(left_options, right_options):
         if _has_catalog_options_against_generic_count_alignment(left, right, left_options, right_options):
             return False
+        pair_text = f"{left.searchable_name} {right.searchable_name}"
+        if any(term in pair_text for term in ("眉筆", "眼線膠筆", "唇膏", "唇釉", "粉餅盒", "遮瑕蜜")) and (
+            (
+                len(left_options) > len(right_options)
+                and _is_catalog_or_delimited_variant_listing(left)
+            )
+            or (
+                len(right_options) > len(left_options)
+                and _is_catalog_or_delimited_variant_listing(right)
+            )
+        ):
+            return False
         if (
             len(left_options) > len(right_options)
             and _has_variant_option_selection_gap(left, left_options)
@@ -3923,15 +4114,37 @@ def _has_named_variant_selection_review(
         return True
     left_options = _explicit_variant_option_tokens(left)
     right_options = _explicit_variant_option_tokens(right)
+    if left_options and right_options:
+        for catalog_identity, catalog_options, specific_options in (
+            (left, left_options, right_options),
+            (right, right_options, left_options),
+        ):
+            if (
+                _is_catalog_or_delimited_variant_listing(catalog_identity)
+                and len(catalog_options) > len(specific_options)
+                and _variant_options_overlap(catalog_options, specific_options)
+                and _is_variant_sensitive_identity(left, right, shared_anchor)
+            ):
+                return True
     if bool(left_options) != bool(right_options):
         option_identity = left if left_options else right
         catalog_identity = right if left_options else left
         if (
             _is_variant_sensitive_identity(left, right, shared_anchor)
-            and _is_multi_variant_catalog_listing(catalog_identity)
+            and _is_catalog_or_delimited_variant_listing(catalog_identity)
             and _explicit_variant_option_tokens(option_identity)
         ):
             return True
+        if (
+            _is_variant_sensitive_identity(left, right, shared_anchor)
+            and _has_overlapping_base_spec(left, right)
+            and _explicit_variant_option_tokens(option_identity)
+            and any(
+                term in f"{left.searchable_name} {right.searchable_name}"
+                for term in ("粉餅盒", "護手霜", "護唇膏", "護唇棒", "滋養霜", "眼線膠筆", "遮瑕蜜")
+            )
+        ):
+            return True
     if bool(left_options) == bool(right_options):
         return False