chore: import existing codebase as baseline

2026-09-26 01:12:10 +08:00
commit 0ac0c4eb56
77 changed files with 14155 additions and 0 deletions
@@ -0,0 +1,109 @@
+import math
+
+from .analysis import METRICS
+
+RESERVED_DATA_HEADERS = {
+    "period",
+    "customertoken",
+    "revenue",
+    "cost",
+    "grossinput",
+    "marginrate",
+    "grosssource",
+    "listamount",
+    "selfoperated",
+    "customerpresent",
+    "rowcount",
+    "grosscalc",
+}
+
+
+def available_metrics(ds):
+    roles = {column["role"] for column in ds.mapping["columns"]}
+    dimensions = {column.get("key") for column in ds.mapping["columns"] if column["role"] == "dimension"}
+    products = bool(ds.mapping.get("self_operated_products")) and "product" in dimensions
+    gross = "gross" in roles or (bool(ds.mapping.get("margin_rates")) and "product" in dimensions)
+    customer, list_amount, cost = "customer" in roles, "list_amount" in roles, "cost" in roles
+    available = {"revenue"}
+    if cost:
+        available.add("cost")
+    if gross:
+        available.update(("gross", "margin"))
+    if customer:
+        available.update(("customers", "basket"))
+    if list_amount:
+        available.add("discount_rate")
+    if customer and list_amount:
+        available.add("list_basket")
+    if cost and (ds.mapping.get("roi_formula") == "profit_over_cost" or gross):
+        available.add("roi")
+    if products:
+        available.add("self_revenue")
+        if customer:
+            available.add("upgrade_rate")
+    return {key: value for key, value in METRICS.items() if key in available}
+
+
+OPERATORS = {"add": "+", "subtract": "-", "multiply": "*", "divide": "/"}
+
+
+def validate_column(value):
+    if not isinstance(value, dict) or set(value) != {"name", "meaning", "scope", "missing", "expression"}:
+        raise ValueError("自定义列定义字段无效")
+    for key, limit in (("name", 80), ("meaning", 500), ("scope", 200), ("missing", 300)):
+        text = value[key]
+        if not isinstance(text, str) or not text.strip() or len(text) > limit:
+            raise ValueError(f"自定义列{key}说明不完整")
+    if value["missing"] != "结果不可用（#N/A）；零分母不可用":
+        raise ValueError("暂不支持自定义缺失值策略；请选择结果不可用（#N/A）及零分母不可用")
+    _validate_expression(value["expression"])
+    return value
+
+
+def _validate_expression(node, depth=0):
+    if depth > 8 or not isinstance(node, dict):
+        raise ValueError("自定义计算公式结构无效")
+    if set(node) == {"metric"}:
+        if not isinstance(node["metric"], str) or node["metric"] not in METRICS:
+            raise ValueError("公式引用了不支持的指标")
+        return
+    if set(node) == {"number"}:
+        if (
+            type(node["number"]) not in (int, float)
+            or not math.isfinite(node["number"])
+            or abs(node["number"]) > 1e12
+        ):
+            raise ValueError("公式常数无效")
+        return
+    if (
+        set(node) == {"operator", "left", "right"}
+        and isinstance(node["operator"], str)
+        and node["operator"] in OPERATORS
+    ):
+        _validate_expression(node["left"], depth + 1)
+        _validate_expression(node["right"], depth + 1)
+        return
+    raise ValueError("公式包含不支持的字段或运算")
+
+
+def metrics_used(node):
+    if "metric" in node:
+        return {node["metric"]}
+    if "number" in node:
+        return set()
+    return metrics_used(node["left"]) | metrics_used(node["right"])
+
+
+def ensure_available_metrics(ds, column):
+    unavailable = metrics_used(column["expression"]) - available_metrics(ds).keys()
+    if unavailable:
+        names = "、".join(METRICS[key] for key in sorted(unavailable))
+        raise ValueError(f"当前数据集未映射计算“{names}”所需字段，不能生成此自定义列")
+
+
+def to_excel(node, references):
+    if "metric" in node:
+        return references[node["metric"]]
+    if "number" in node:
+        return str(node["number"])
+    return f"({to_excel(node['left'], references)}{OPERATORS[node['operator']]}{to_excel(node['right'], references)})"