"""Step 1 (v2): 全量采集ERP表结构到 erp_schema 通过 erp-api-gateway 的 /api/v1/query?table=xxx&limit=1 获取字段信息 运行: python3 scripts/collect_erp_schema_v2.py """ import sys, os, json, urllib.request, urllib.error, logging, time from datetime import datetime sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) from app.database import get_session_local from sqlalchemy import text logging.basicConfig(level=logging.INFO, format="%(asctime)s [%(levelname)s] %(message)s") logger = logging.getLogger("erp_schema_v2") API_BASE = "http://127.0.0.1:8300/api/v1" API_KEY = os.getenv("ERP_API_KEY", "erp-gateway-key-bhwl-2026") HEADERS = {"X-API-Key": API_KEY} def api_get(path: str) -> dict: url = f"{API_BASE}{path}" req = urllib.request.Request(url, headers=HEADERS) with urllib.request.urlopen(req, timeout=30) as resp: return json.loads(resp.read().decode()) def main(): db = get_session_local()() # 1. 获取全量表名 logger.info("获取ERP全量表名...") tables_data = api_get("/tables") all_tables = tables_data.get("tables", []) logger.info(f"共 {len(all_tables)} 张表") # 2. 已采集的 existing = set() for row in db.execute(text("SELECT table_name FROM erp_schema")).fetchall(): existing.add(row[0]) logger.info(f"已有 {len(existing)} 张,还需采集 {len(all_tables) - len(existing)} 张") # 3. 逐表采集 collected, errors, skipped = 0, 0, 0 for i, tn in enumerate(all_tables): if tn in existing: skipped += 1 continue try: data = api_get(f"/query?table={tn}&limit=1") cols = data.get("columns", []) total = data.get("total", 0) fields_json = json.dumps([{"name": c} for c in cols], ensure_ascii=False) db.execute(text(""" INSERT INTO erp_schema (table_name, total_rows, field_count, fields_json, created_at, updated_at) VALUES (:tn, :tr, :fc, :fj, NOW(), NOW()) ON DUPLICATE KEY UPDATE total_rows=:tr2, field_count=:fc2, fields_json=:fj2, updated_at=NOW() """), {"tn": tn, "tr": total, "fc": len(cols), "fj": fields_json, "tr2": total, "fc2": len(cols), "fj2": fields_json}) db.commit() collected += 1 except Exception as e: errors += 1 db.execute(text(""" INSERT INTO erp_schema (table_name, total_rows, field_count, fields_json, created_at, updated_at) VALUES (:tn, -1, 0, '[]', NOW(), NOW()) ON DUPLICATE KEY UPDATE updated_at=NOW() """), {"tn": tn}) db.commit() if (i + 1) % 100 == 0: logger.info(f"进度: {i+1}/{len(all_tables)} 已采{collected} 错误{errors} 跳过{skipped}") time.sleep(0.1) db.close() logger.info(f"\n完成! 总计:{len(all_tables)} 采集:{collected} 已有:{skipped} 错误:{errors}") if __name__ == "__main__": main()