fixing

extensions/sync_ehr_to_oa/job.py

@@ -79,6 +79,23 @@ def _normalize_job_no(v: Any) -> str:
     return s.upper()
 
 
+def _to_int_candidate(v: Any) -> int | None:
+    if v is None:
+        return None
+    if isinstance(v, dict):
+        v = _cell_value(v)
+    s = str(v).strip()
+    if not s:
+        return None
+    try:
+        # 兼容 "123.0"
+        if "." in s and s.endswith(".0"):
+            return int(float(s))
+        return int(s)
+    except Exception:
+        return None
+
+
 def _extract_oa_row_id_and_fields(row: dict[str, Any]) -> tuple[int | None, dict[str, Any]]:
     """
     兼容不同 OA export 返回结构，提取：
@@ -94,14 +111,20 @@ def _extract_oa_row_id_and_fields(row: dict[str, Any]) -> tuple[int | None, dict
         for k, v in master.items():
             if isinstance(k, str) and k.startswith("field"):
                 field_map[k] = v
-        for candidate in (row.get("id"), row.get("masterDataId"), master.get("id")):
+        # 常见主键位置优先尝试
-            if candidate is None:
+        for candidate in (
-                continue
+            row.get("id"),
-            try:
+            row.get("masterDataId"),
-                row_id = int(str(candidate))
+            master.get("id"),
+            master.get("ID"),
+            master.get("recordId"),
+            master.get("dataId"),
+            master.get("_id"),
+        ):
+            rid = _to_int_candidate(candidate)
+            if rid is not None:
+                row_id = rid
                 break
-            except Exception:
-                continue
 
     # 结构 B：masterTable.record.fields = [{name,value,showValue}, ...]
     master_table = row.get("masterTable")
@@ -117,12 +140,7 @@ def _extract_oa_row_id_and_fields(row: dict[str, Any]) -> tuple[int | None, dict
                     if name:
                         field_map[name] = fld
             if row_id is None:
-                rid = record.get("id")
+                row_id = _to_int_candidate(record.get("id"))
-                if rid is not None:
-                    try:
-                        row_id = int(str(rid))
-                    except Exception:
-                        pass
 
     # 结构 C：行级 fields 列表
     row_fields = row.get("fields")
@@ -351,6 +369,8 @@ class SyncEhrToOaFormJob(BaseJob):
             oa_id_by_job_no: dict[str, int] = {}
             oa_id_by_job_no_norm: dict[str, int] = {}
             row_parse_miss = 0
+            row_id_fallback_count = 0
+            row_id_fallback_samples: list[str] = []
             for row in rows:
                 if not isinstance(row, dict):
                     continue
@@ -368,24 +388,30 @@ class SyncEhrToOaFormJob(BaseJob):
                     continue
 
                 if row_id is None:
-                    row_parse_miss += 1
+                    # 某些 OA export 不返回主键 id；此时依赖 uniqueFiled(工号) 进行更新匹配。
-                    if verbose_trace and row_parse_miss <= 20:
+                    row_id = 0
+                    row_id_fallback_count += 1
+                    if len(row_id_fallback_samples) < 20:
+                        row_id_fallback_samples.append(job_no)
+                    if verbose_trace and row_id_fallback_count <= 20:
                         logger.info(
-                            "OA 行解析未取到记录ID：job_no=%s row_keys=%s",
+                            "OA 行解析未取到记录ID，使用回退ID=0：job_no=%s row_keys=%s",
                             job_no,
                             list(row.keys())[:20],
                         )
-                    continue
                 oa_id_by_job_no[job_no] = row_id
                 job_no_norm = _normalize_job_no(job_no)
                 if job_no_norm:
                     oa_id_by_job_no_norm[job_no_norm] = row_id
             logger.info(
-                "OA 工号索引完成：indexed_job_numbers=%s indexed_job_numbers_norm=%s parse_miss=%s",
+                "OA 工号索引完成：indexed_job_numbers=%s indexed_job_numbers_norm=%s parse_miss=%s id_fallback=%s",
                 len(oa_id_by_job_no),
                 len(oa_id_by_job_no_norm),
                 row_parse_miss,
+                row_id_fallback_count,
             )
+            if row_id_fallback_samples:
+                logger.info("OA 记录ID回退样本（前20）：%s", row_id_fallback_samples)
             if verbose_trace:
                 for job_no, row_id in list(oa_id_by_job_no.items()):
                     logger.info("OA 工号索引明细：raw=%s norm=%s row_id=%s", job_no, _normalize_job_no(job_no), row_id)