Reclassify 4.A.05.A0008 as special servo

scripts/extract_catalog.py

@@ -26,6 +26,11 @@ CATEGORY_BY_PAGE = {
     **{page: "翻盖电机" for page in range(28, 30)},
 }
 
+# Business categories can intentionally differ from the source page heading.
+CATEGORY_BY_CODE = {
+    "4.A.05.A0008": "异形舵机",
+}
+
 SPEC_LABELS = {
     "标准舵机": [
         "产品尺寸", "最大砝码负载", "堵转扭矩", "空载速度", "可操作角度",
@@ -141,9 +146,9 @@ def main() -> None:
 
     with pdfplumber.open(PDF_PATH) as pdf:
         for page_number, page in enumerate(pdf.pages, 1):
-            category = CATEGORY_BY_PAGE[page_number]
+            page_category = CATEGORY_BY_PAGE[page_number]
             page_text = clean(page.extract_text() or "")
-            expected_heading = f"{category}产品明细表"
+            expected_heading = f"{page_category}产品明细表"
             if expected_heading not in page_text:
                 raise RuntimeError(
                     f"Category heading mismatch on page {page_number}: expected {expected_heading}"
@@ -169,6 +174,7 @@ def main() -> None:
                 code = clean_code(row[2])
                 if not code or not re.search(r"[A-Z].*\d", code):
                     continue
+                category = CATEGORY_BY_CODE.get(code, page_category)
 
                 explicit_name = clean(row[1])
                 if explicit_name:
@@ -185,10 +191,10 @@ def main() -> None:
                     crop_product_image(rendered, page.width, page.height, cell, IMAGE_DIR / image_name)
                     inherited_image = f"/catalog-products/{image_name}"
 
-                values = normalize_spec_values(category, row, inherited_size)
+                values = normalize_spec_values(page_category, row, inherited_size)
                 specs = {
                     label: value
-                    for label, value in zip(SPEC_LABELS[category], values)
+                    for label, value in zip(SPEC_LABELS[page_category], values)
                     if value and value != "/"
                 }