feat: 会话编排建材化+阿语号码转ASCII+12小时话术跨层同步

Co-Authored-By: Claude Code <noreply@anthropic.com>

app/services/contact_pipeline.py

@@ -10,6 +10,19 @@
 """
 import re
 
+# Unicode 十进制数字 → ASCII(阿语٠-٩/波斯语۰-۹), 中东买家号码 CRM 可直接拨。
+# 挂在归一化链最前端(normalize_contacts 的值 / digits_only / scan 与
+# latest 的捕获赋值): 存库、canon 去重比对、boundary_filter 客户回显豁免
+# 两侧全部经此转成 ASCII——阿语消息与 ASCII 提取值比对口径一致,
+# 同一号码不会被误判成"改号"触发更正邮件。
+_DIGIT_TRANSLATE = str.maketrans(
+    "٠١٢٣٤٥٦٧٨٩۰۱۲۳۴۵۶۷۸۹", "01234567890123456789")
+
+
+def _to_ascii_digits(value: str) -> str:
+    """捕获源头归一: 阿语/波斯语数字转 ASCII(translate 对 ASCII 恒等)。"""
+    return value.translate(_DIGIT_TRANSLATE)
+
 KEY_ALIASES = {
     "phone": "phone", "手机": "phone", "电话": "phone", "电话号码": "phone",
     "联系电话": "phone", "手机号": "phone", "手机号码": "phone",
@@ -38,7 +51,7 @@ def normalize_contacts(raw: dict | None) -> dict:
                 if name in key:
                     alias = target
                     break
-        val = str(v or "").strip()
+        val = _to_ascii_digits(str(v or "").strip())
         if alias and val:
             out[alias] = val
     return out
@@ -50,8 +63,11 @@ def merge_contacts(old: dict | None, new: dict | None) -> dict:
 
 
 def digits_only(s: str) -> str:
-    """提取纯数字（split_validate 位数判断与后续 canon 复用）。"""
-    return re.sub(r"\D", "", s)
+    """提取纯数字（split_validate 位数判断与后续 canon 复用）。
+    阿语/波斯语数字先转 ASCII——canon 与 boundary_filter 回显豁免两侧
+    都走本函数,天然保证同形比较(Unicode 正则的数字类本就匹配阿语数字,
+    位数判定不变;此处只统一"比对与存储的形态")。"""
+    return re.sub(r"\D", "", s.translate(_DIGIT_TRANSLATE))
 
 
 def split_validate(contacts: dict) -> tuple[dict, dict]:
@@ -72,7 +88,7 @@ def split_validate(contacts: dict) -> tuple[dict, dict]:
 EMAIL_RE = re.compile(r"[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}")
 INTL_PHONE_RE = re.compile(r"\+\d{6,15}")
 # 邻接绑定：上下文词与数字必须在同一匹配内（仅空白衔接）——禁止"消息任意位置
-# 有上下文词 + 任意位置有 7 位数字"的消息级误绑（S1：车架号/订单号/先令价格）
+# 有上下文词 + 任意位置有 7 位数字"的消息级误绑（S1：单号/订单号/先令价格）
 WHATSAPP_CTX_RE = re.compile(
     r"(?:whats\s*app|wa\.?me|(?:^|\b)wa\s*[:：])\D{0,12}(\+?\d[\d\s\-().]{5,17}\d)",
     re.IGNORECASE,
@@ -129,7 +145,7 @@ CORRECTION_CTX_RE = re.compile(
 
 def scan_user_messages(messages: list[dict]) -> dict:
     """S4: 扫描全部客户消息,返回高置信捕获(email / +国际号 / whatsapp·wa.me /
-    显式电话上下文邻接数字)。裸数字(价格/年份/里程/车架号)一律不收。"""
+    显式电话上下文邻接数字)。裸数字(价格/年份/数量/单号)一律不收。"""
     found: dict[str, str] = {}
     for msg in messages:
         if msg.get("role") != "user":
@@ -142,30 +158,30 @@ def scan_user_messages(messages: list[dict]) -> dict:
         if "phone" not in found:
             sn = standalone_number(text)
             if sn:
-                found["phone"] = sn
+                found["phone"] = _to_ascii_digits(sn)
         if "phone" not in found:
             hit = INTL_PHONE_RE.search(text)
             if hit:
-                found["phone"] = hit.group(0)
+                found["phone"] = _to_ascii_digits(hit.group(0))
         if "whatsapp" not in found:
             hit = WHATSAPP_CTX_RE.search(text)
             if hit:
-                found["whatsapp"] = hit.group(1)
+                found["whatsapp"] = _to_ascii_digits(hit.group(1))
         if "phone" not in found:
             hit = PHONE_CTX_RE.search(text)
             if hit:
-                found["phone"] = hit.group(1)
+                found["phone"] = _to_ascii_digits(hit.group(1))
         if "phone" not in found:
             hit = PHONE_ADJACENCY_SCAN_RE.search(text)
             if hit:
-                found["phone"] = hit.group(1)
+                found["phone"] = _to_ascii_digits(hit.group(1))
         # 倒装属最低置信层: whatsapp/phone 任一已被更高置信源捕获即跳过,
         # 不与之竞争(跨 key 门控故含 phone)。
         if "whatsapp" not in found and "phone" not in found:
             hit = NUMBER_FIRST_RE.search(text)
             if hit:
                 ch = _number_first_channel(hit)
-                found[ch] = hit.group(1)
+                found[ch] = _to_ascii_digits(hit.group(1))
     return found
 
 
@@ -211,21 +227,21 @@ def latest_stated_contacts(messages: list[dict]) -> dict:
             latest["email"] = hit.group(0)
         hit = WHATSAPP_CTX_RE.search(text)
         if hit:
-            latest["whatsapp"] = hit.group(1)
+            latest["whatsapp"] = _to_ascii_digits(hit.group(1))
         sn = standalone_number(text)
         if sn:
-            latest["phone"] = sn
+            latest["phone"] = _to_ascii_digits(sn)
         hit = PHONE_CTX_RE.search(text)
         if hit:
-            latest["phone"] = hit.group(1)
+            latest["phone"] = _to_ascii_digits(hit.group(1))
         elif CORRECTION_CTX_RE.search(text):
             hit = INTL_PHONE_RE.search(text)
             if hit:
-                latest["phone"] = hit.group(0)
+                latest["phone"] = _to_ascii_digits(hit.group(0))
         hit = NUMBER_FIRST_RE.search(text)
         if hit:
             ch = _number_first_channel(hit)
-            latest[ch] = hit.group(1)
+            latest[ch] = _to_ascii_digits(hit.group(1))
     return latest