Add manual catalog management

tools/device_mapper.py

@@ -85,6 +85,8 @@ FILE_DEFAULT_DEVICE_TYPE: Dict[str, str] = {
     "zhixuan": "phone",
 }
 
+MANUAL_SOURCE_FILE = "local/manual_catalog.json"
+
 
 BRAND_ALIASES: Dict[str, List[str]] = {
     "360": ["360", "360手机", "奇酷", "qiku"],
@@ -239,6 +241,10 @@ class DeviceRecord:
     section: str
 
 
+MANUAL_BRAND_ALIAS_OVERRIDES: Dict[str, List[str]] = {}
+MANUAL_PARENT_BRAND_OVERRIDES: Dict[str, str] = {}
+
+
 def normalize_text(text: str) -> str:
     return re.sub(r"[^0-9a-z\u4e00-\u9fff]+", "", text.lower())
 
@@ -249,6 +255,7 @@ def canonical_brand(file_stem: str) -> str:
 
 def brand_aliases(brand: str) -> List[str]:
     aliases = set(BRAND_ALIASES.get(brand, []))
+    aliases.update(MANUAL_BRAND_ALIAS_OVERRIDES.get(brand, []))
     aliases.add(brand)
     return sorted(aliases)
 
@@ -265,9 +272,108 @@ def has_keyword(text: str, keywords: Iterable[str]) -> bool:
 
 
 def resolve_parent_brand(manufacturer_brand: str) -> str:
+    if manufacturer_brand in MANUAL_PARENT_BRAND_OVERRIDES:
+        return MANUAL_PARENT_BRAND_OVERRIDES[manufacturer_brand]
     return MANUFACTURER_PARENT_BRAND.get(manufacturer_brand, manufacturer_brand)
 
 
+def reset_manual_overrides() -> None:
+    MANUAL_BRAND_ALIAS_OVERRIDES.clear()
+    MANUAL_PARENT_BRAND_OVERRIDES.clear()
+
+
+def normalize_alias_list(*groups: object) -> List[str]:
+    aliases: List[str] = []
+    seen: Set[str] = set()
+    for group in groups:
+        if group is None:
+            continue
+        items = group if isinstance(group, (list, tuple, set)) else [group]
+        for item in items:
+            text = str(item or "").strip()
+            key = normalize_text(text)
+            if not text or not key or key in seen:
+                continue
+            seen.add(key)
+            aliases.append(text)
+    return aliases
+
+
+def load_manual_catalog(repo_root: Path) -> dict[str, object]:
+    path = repo_root / MANUAL_SOURCE_FILE
+    if not path.exists():
+        return {"brands": [], "devices": []}
+    payload = json.loads(path.read_text(encoding="utf-8"))
+    if not isinstance(payload, dict):
+        raise RuntimeError(f"{MANUAL_SOURCE_FILE} 必须是 JSON 对象。")
+    brands = payload.get("brands")
+    devices = payload.get("devices")
+    return {
+        "brands": brands if isinstance(brands, list) else [],
+        "devices": devices if isinstance(devices, list) else [],
+    }
+
+
+def apply_manual_brand_overrides(manual_catalog: dict[str, object]) -> dict[str, str]:
+    reset_manual_overrides()
+    brand_parent_map: dict[str, str] = {}
+    for raw_brand in manual_catalog.get("brands", []):
+        if not isinstance(raw_brand, dict):
+            continue
+        brand_name = str(raw_brand.get("name") or "").strip()
+        if not brand_name:
+            continue
+        aliases = normalize_alias_list(brand_name, raw_brand.get("aliases"))
+        parent_brand = str(raw_brand.get("parent_brand") or brand_name).strip() or brand_name
+        MANUAL_BRAND_ALIAS_OVERRIDES[brand_name] = aliases
+        MANUAL_PARENT_BRAND_OVERRIDES[brand_name] = parent_brand
+        brand_parent_map[brand_name] = parent_brand
+    return brand_parent_map
+
+
+def parse_manual_catalog(repo_root: Path, manual_catalog: dict[str, object]) -> List[DeviceRecord]:
+    brand_parent_map = apply_manual_brand_overrides(manual_catalog)
+    records: List[DeviceRecord] = []
+
+    for raw_device in manual_catalog.get("devices", []):
+        if not isinstance(raw_device, dict):
+            continue
+        brand = str(raw_device.get("brand") or "").strip()
+        device_name = str(raw_device.get("device_name") or "").strip()
+        if not brand or not device_name:
+            continue
+
+        aliases = normalize_alias_list(
+            raw_device.get("models"),
+            device_name,
+            raw_device.get("aliases"),
+        )
+        if not aliases:
+            continue
+
+        device_type = str(raw_device.get("device_type") or "").strip() or "other"
+        section = str(raw_device.get("section") or "手动补录").strip() or "手动补录"
+        record_id = str(raw_device.get("id") or "").strip() or f"manual:{normalize_text(brand)}:{normalize_text(device_name)}"
+        parent_brand = brand_parent_map.get(brand, str(raw_device.get("parent_brand") or brand).strip() or brand)
+
+        records.append(
+            DeviceRecord(
+                id=record_id,
+                device_name=device_name,
+                brand=brand,
+                manufacturer_brand=brand,
+                parent_brand=parent_brand,
+                market_brand=brand,
+                device_type=device_type,
+                aliases=aliases,
+                source_file=MANUAL_SOURCE_FILE,
+                section=section,
+            )
+        )
+
+    return records
+
+
 def infer_market_brand(
     manufacturer_brand: str,
     device_name: str,
@@ -647,10 +753,13 @@ class DeviceMapper:
 def build_records(repo_root: Path) -> List[DeviceRecord]:
     brands_dir = repo_root / "brands"
     records: List[DeviceRecord] = []
+    manual_catalog = load_manual_catalog(repo_root)
+    apply_manual_brand_overrides(manual_catalog)
 
     for md_path in sorted(brands_dir.glob("*.md")):
         records.extend(parse_brand_file(md_path))
 
+    records.extend(parse_manual_catalog(repo_root, manual_catalog))
     return records