345 lines
12 KiB
Python
345 lines
12 KiB
Python
"""
|
||
陆运选项规则：运输类型、线路类别、运输分类、中港装载方式。
|
||
|
||
本文件职责：从销售原话收取标准项、校验联动、生成补问清单、判断「确定查价」意图。
|
||
联动错误只纠正那一项能选什么，不整单作废。
|
||
禁止：改六态；发明枚举；把选项当货物字段。
|
||
线程：纯函数，无共享可变状态，可在 Worker 槽调用。
|
||
"""
|
||
|
||
from __future__ import annotations
|
||
|
||
from typing import Any
|
||
|
||
# 与后台员工管理陆运「负责线路」同一套字，禁止改字。
|
||
TRANSPORT_TYPES = (
|
||
"国内运输拼车",
|
||
"国内运输整车",
|
||
"跨境集拼",
|
||
"跨境整车",
|
||
"中港",
|
||
)
|
||
DOMESTIC_ROUTES = ("国内长途/零担", "国内城配/拖车/打包")
|
||
CROSS_ROUTES = ("中港/中亚/中欧", "东南亚")
|
||
ROUTES_BY_TYPE = {
|
||
"国内运输拼车": DOMESTIC_ROUTES,
|
||
"国内运输整车": DOMESTIC_ROUTES,
|
||
"跨境集拼": CROSS_ROUTES,
|
||
"跨境整车": CROSS_ROUTES,
|
||
"中港": CROSS_ROUTES,
|
||
}
|
||
CROSS_CLASSES = (
|
||
"普件，中港、东南亚常用线路/常规车型",
|
||
"普件，中亚、中欧、东南亚涉及国内+国外段",
|
||
"普件，中亚、中欧、东南亚涉及境外段清派等特殊要求",
|
||
"急件",
|
||
)
|
||
DOMESTIC_CLASSES = (
|
||
"普件，常用线路/常规车型",
|
||
"普件，非常用线路/临时线路/特种车型",
|
||
"急件",
|
||
)
|
||
CLASSES_BY_ROUTE = {
|
||
"中港/中亚/中欧": CROSS_CLASSES,
|
||
"东南亚": CROSS_CLASSES,
|
||
"国内长途/零担": DOMESTIC_CLASSES,
|
||
"国内城配/拖车/打包": DOMESTIC_CLASSES,
|
||
}
|
||
HK_LOADS = ("整车", "零担/集拼")
|
||
|
||
# 口语别名 → 标准运输类型。长词优先，避免「拼车」误伤「整车」。
|
||
_TYPE_ALIASES = (
|
||
("国内运输拼车", "国内运输拼车"),
|
||
("国内运输整车", "国内运输整车"),
|
||
("国内拼车", "国内运输拼车"),
|
||
("国内整车", "国内运输整车"),
|
||
("跨境集拼", "跨境集拼"),
|
||
("跨境整车", "跨境整车"),
|
||
("中港", "中港"),
|
||
)
|
||
|
||
_ROUTE_ALIASES = (
|
||
("国内城配/拖车/打包", "国内城配/拖车/打包"),
|
||
("国内长途/零担", "国内长途/零担"),
|
||
("中港/中亚/中欧", "中港/中亚/中欧"),
|
||
("国内城配", "国内城配/拖车/打包"),
|
||
("国内长途", "国内长途/零担"),
|
||
("东南亚", "东南亚"),
|
||
)
|
||
|
||
# 分类全称从长到短；「急件」单独出现才算急件，避免普件句被误收。
|
||
_CLASS_CANON = tuple(
|
||
sorted(
|
||
{*(CROSS_CLASSES + DOMESTIC_CLASSES)},
|
||
key=len,
|
||
reverse=True,
|
||
)
|
||
)
|
||
|
||
_HK_LOAD_ALIASES = (
|
||
("零担/集拼", "零担/集拼"),
|
||
("零担", "零担/集拼"),
|
||
("集拼", "零担/集拼"),
|
||
("整车", "整车"),
|
||
)
|
||
|
||
_CONFIRM_PHRASES = ("可以查了", "确定", "确认", "查价")
|
||
_CONFIRM_EXCLUDE = ("始发", "目的", "体积", "车型", "改成", "重量")
|
||
|
||
|
||
def canonicalize_transport_type(raw: str) -> str:
|
||
"""
|
||
把销售原话收成标准运输类型。
|
||
|
||
参数：任意一句。返回 TRANSPORT_TYPES 之一或空串。
|
||
副作用：无。先匹配更长别名。
|
||
"""
|
||
text = (raw or "").strip()
|
||
if not text:
|
||
return ""
|
||
for needle, canon in _TYPE_ALIASES:
|
||
if needle in text:
|
||
return canon
|
||
return ""
|
||
|
||
|
||
def canonicalize_route(raw: str) -> str:
|
||
"""把原话收成标准线路类别；对不上返回空串。"""
|
||
text = (raw or "").strip()
|
||
if not text:
|
||
return ""
|
||
for needle, canon in _ROUTE_ALIASES:
|
||
if needle in text:
|
||
return canon
|
||
return ""
|
||
|
||
|
||
def canonicalize_land_class(raw: str) -> str:
|
||
"""
|
||
把原话收成标准运输分类。
|
||
|
||
先匹配完整分类句；整句只有「急件」才收急件，避免普件描述被截成急件。
|
||
"""
|
||
text = (raw or "").strip()
|
||
if not text:
|
||
return ""
|
||
for canon in _CLASS_CANON:
|
||
if canon == "急件":
|
||
continue
|
||
if canon in text:
|
||
return canon
|
||
stripped = text.replace("运输分类", "").replace("：", "").replace(":", "").strip()
|
||
if stripped == "急件" or text.strip() == "急件":
|
||
return "急件"
|
||
if "急件" in text and "普件" not in text:
|
||
return "急件"
|
||
return ""
|
||
|
||
|
||
def canonicalize_hk_load(raw: str) -> str:
|
||
"""中港装载方式：整车或零担/集拼。"""
|
||
text = (raw or "").strip()
|
||
if not text:
|
||
return ""
|
||
for needle, canon in _HK_LOAD_ALIASES:
|
||
if needle in text:
|
||
return canon
|
||
return ""
|
||
|
||
|
||
def harvest_land_options(text: str, current: dict[str, str] | None = None) -> dict[str, str]:
|
||
"""
|
||
从一句原话收取陆运选项，合并到已有值上。
|
||
|
||
已有非空项默认保留；本句明确写出另一标准项时才覆盖。
|
||
已有值对不上标准枚举则丢掉，避免模型把「陆运」写成运输类型后整张清单重问。
|
||
不改货物字段。返回新 dict，不修改 current。
|
||
"""
|
||
merged = {
|
||
k: str(v).strip()
|
||
for k, v in dict(current or {}).items()
|
||
if str(v).strip()
|
||
}
|
||
_keep_canonical_options(merged)
|
||
blob = text or ""
|
||
found_type = canonicalize_transport_type(blob)
|
||
found_route = canonicalize_route(blob)
|
||
found_class = canonicalize_land_class(blob)
|
||
found_load = canonicalize_hk_load(blob)
|
||
if found_type:
|
||
merged["运输类型"] = found_type
|
||
if found_route:
|
||
merged["线路类别"] = found_route
|
||
if found_class:
|
||
merged["运输分类"] = found_class
|
||
if found_load:
|
||
merged["中港装载方式"] = found_load
|
||
return merged
|
||
|
||
|
||
def _keep_canonical_options(merged: dict[str, str]) -> None:
|
||
"""只留能对上枚举的选项；模型胡写的「陆运」不能冒充运输类型。"""
|
||
t = canonicalize_transport_type(merged.get("运输类型") or "")
|
||
if t:
|
||
merged["运输类型"] = t
|
||
else:
|
||
merged.pop("运输类型", None)
|
||
r = canonicalize_route(merged.get("线路类别") or "")
|
||
if r:
|
||
merged["线路类别"] = r
|
||
else:
|
||
merged.pop("线路类别", None)
|
||
c = canonicalize_land_class(merged.get("运输分类") or "")
|
||
if c:
|
||
merged["运输分类"] = c
|
||
else:
|
||
merged.pop("运输分类", None)
|
||
load = canonicalize_hk_load(merged.get("中港装载方式") or "")
|
||
if load:
|
||
merged["中港装载方式"] = load
|
||
else:
|
||
merged.pop("中港装载方式", None)
|
||
|
||
|
||
def validate_land_options(opts: dict[str, str]) -> dict[str, Any]:
|
||
"""
|
||
校验选项是否齐全且联动正确。
|
||
|
||
返回 ok / missing / errors。errors 每项 {field, message}，只纠正错的那一项。
|
||
中港缺装载方式进 missing，不改货物字段。
|
||
副作用：无。
|
||
"""
|
||
src = {k: str(v).strip() for k, v in dict(opts or {}).items() if str(v).strip()}
|
||
t = canonicalize_transport_type(src.get("运输类型") or "") or src.get("运输类型") or ""
|
||
r = canonicalize_route(src.get("线路类别") or "") or src.get("线路类别") or ""
|
||
c = canonicalize_land_class(src.get("运输分类") or "") or src.get("运输分类") or ""
|
||
load = canonicalize_hk_load(src.get("中港装载方式") or "") or src.get("中港装载方式") or ""
|
||
|
||
missing: list[str] = []
|
||
errors: list[dict[str, str]] = []
|
||
|
||
if t not in TRANSPORT_TYPES:
|
||
missing.append("运输类型")
|
||
t = ""
|
||
|
||
if t:
|
||
allowed_routes = ROUTES_BY_TYPE.get(t) or ()
|
||
if not r:
|
||
missing.append("线路类别")
|
||
elif r not in allowed_routes:
|
||
errors.append(
|
||
{
|
||
"field": "线路类别",
|
||
"message": "线路类别只能选：" + "、".join(f"「{x}」" for x in allowed_routes),
|
||
}
|
||
)
|
||
r = ""
|
||
elif r:
|
||
# 类型还没有，线路先记下但不判联动
|
||
pass
|
||
|
||
if r:
|
||
allowed_cls = CLASSES_BY_ROUTE.get(r) or ()
|
||
if not c:
|
||
missing.append("运输分类")
|
||
elif c not in allowed_cls:
|
||
errors.append(
|
||
{
|
||
"field": "运输分类",
|
||
"message": "运输分类只能选：" + "、".join(f"「{x}」" for x in allowed_cls),
|
||
}
|
||
)
|
||
c = ""
|
||
elif "运输分类" not in missing and not c:
|
||
missing.append("运输分类")
|
||
|
||
if t == "中港":
|
||
if load not in HK_LOADS:
|
||
missing.append("中港装载方式")
|
||
ok = not missing and not errors
|
||
return {
|
||
"ok": ok,
|
||
"missing": missing,
|
||
"errors": errors,
|
||
"normalized": {
|
||
"运输类型": t,
|
||
"线路类别": r,
|
||
"运输分类": c,
|
||
"中港装载方式": load if t == "中港" else "",
|
||
},
|
||
}
|
||
|
||
|
||
def option_prompt(*, current: dict[str, str]) -> str:
|
||
"""
|
||
按已识别项生成补问清单。
|
||
|
||
全缺：一次列出类型、线路、分类、中港装载。
|
||
类型已知：不再问类型，只列合法线路/分类。
|
||
不修改 current。
|
||
"""
|
||
src = {k: str(v).strip() for k, v in dict(current or {}).items() if str(v).strip()}
|
||
t = canonicalize_transport_type(src.get("运输类型") or "") or src.get("运输类型") or ""
|
||
r = canonicalize_route(src.get("线路类别") or "") or src.get("线路类别") or ""
|
||
lines: list[str] = []
|
||
|
||
if t not in TRANSPORT_TYPES:
|
||
lines.append("请一次选择：运输类型、线路类别、运输分类。")
|
||
lines.append("")
|
||
lines.append("运输类型：")
|
||
lines.append("、".join(TRANSPORT_TYPES))
|
||
lines.append("")
|
||
lines.append("线路类别（跟运输类型联动）：")
|
||
lines.append("国内运输拼车 / 国内运输整车：" + "、".join(DOMESTIC_ROUTES))
|
||
lines.append("跨境集拼 / 跨境整车 / 中港：" + "、".join(CROSS_ROUTES))
|
||
lines.append("")
|
||
lines.append("运输分类（跟线路类别联动）：")
|
||
lines.append("中港/中亚/中欧、东南亚：" + "；".join(CROSS_CLASSES))
|
||
lines.append("国内长途/零担、国内城配/拖车/打包：" + "；".join(DOMESTIC_CLASSES))
|
||
lines.append("")
|
||
lines.append("运输类型选中港时，还要选择：" + "、".join(HK_LOADS))
|
||
return "\n".join(lines).strip()
|
||
|
||
# 类型已认出来：不再问运输类型
|
||
allowed_routes = ROUTES_BY_TYPE.get(t) or ()
|
||
if r not in allowed_routes:
|
||
lines.append(f"运输类型已选「{t}」。请选择线路类别：")
|
||
lines.append("、".join(allowed_routes))
|
||
route_for_class = ""
|
||
else:
|
||
route_for_class = r
|
||
|
||
if route_for_class:
|
||
allowed_cls = CLASSES_BY_ROUTE.get(route_for_class) or ()
|
||
if src.get("运输分类") not in allowed_cls:
|
||
lines.append(f"线路类别已选「{route_for_class}」。请选择运输分类：")
|
||
lines.append("；".join(allowed_cls))
|
||
elif t in TRANSPORT_TYPES:
|
||
# 线路未定，把该类型下各线路的分类一并列出，方便一次粘贴
|
||
seen: set[str] = set()
|
||
lines.append("运输分类（跟线路类别联动）：")
|
||
for route in allowed_routes:
|
||
if route in seen:
|
||
continue
|
||
seen.add(route)
|
||
lines.append(f"{route}：" + "；".join(CLASSES_BY_ROUTE.get(route) or ()))
|
||
|
||
if t == "中港" and (src.get("中港装载方式") not in HK_LOADS):
|
||
lines.append("中港请再选择装载方式：" + "、".join(HK_LOADS))
|
||
return "\n".join(lines).strip()
|
||
|
||
|
||
def looks_like_confirm(text: str) -> bool:
|
||
"""
|
||
判断是否在表示确定查价。
|
||
|
||
「确定」「确认」「查价」「可以查了」及相近说法都算。
|
||
同一句像在改始发/目的/体积/车型/「改成」则不当确定，避免改内容被当成查价。
|
||
"""
|
||
raw = (text or "").strip()
|
||
if not raw:
|
||
return False
|
||
if any(x in raw for x in _CONFIRM_EXCLUDE):
|
||
return False
|
||
return any(p in raw for p in _CONFIRM_PHRASES)
|