sj_nmtybt_server/scripts/dedupe-pctj-stats.js

/**
 * 清理批次统计脏数据：同一 (c_batch_id, c_pay_bank) 只保留一条。
 * 远端无法可靠物理删，将多余行 state 置为 -1（业务查询 states=0 不可见）。
 *
 * 用法：node scripts/dedupe-pctj-stats.js
 *       node scripts/dedupe-pctj-stats.js --dry-run
 */
const http = require("http");
const qs = require("querystring");

const HOST = "121.43.55.7";
const PORT = 2101;
const DRY = process.argv.includes("--dry-run");
const COL = { columnId: 1846, modelId: 1969, name: "批次统计" };

function request(method, urlPath, fields, token) {
  return new Promise((resolve, reject) => {
    const body = qs.stringify(fields || {});
    const headers = {
      "Content-Type": "application/x-www-form-urlencoded",
      "Content-Length": Buffer.byteLength(body),
    };
    if (token) {
      headers.Token = token;
      headers.token = token;
    }
    const req = http.request(
      { hostname: HOST, port: PORT, path: urlPath, method: method || "POST", headers },
      (res) => {
        let data = "";
        res.on("data", (c) => (data += c));
        res.on("end", () => {
          try {
            resolve(JSON.parse(data));
          } catch (e) {
            resolve({ raw: data, status: res.statusCode });
          }
        });
      }
    );
    req.on("error", reject);
    req.write(body);
    req.end();
  });
}

async function login() {
  const res = await request("POST", "/proxy_oauth/user/login", {
    userName: "user_liu",
    password: "WE176852439@lmx",
    clientId: "1",
  });
  if (res.code != 200) throw new Error("login fail: " + JSON.stringify(res));
  return res.message;
}

async function listAll(token) {
  const rows = [];
  for (let page = 0; page < 80; page++) {
    const res = await request(
      "POST",
      "/proxy_dms/content/selectContentList",
      {
        columnId: String(COL.columnId),
        modelId: String(COL.modelId),
        page: String(page),
        pageSize: "100",
        states: "0",
      },
      token
    );
    if (res.code == 202) break;
    if (res.code != 200) {
      throw new Error("list fail: " + JSON.stringify(res).slice(0, 400));
    }
    const data = (res.content && res.content.data) || [];
    rows.push.apply(rows, data);
    const count = Number((res.content && res.content.count) || 0);
    if (rows.length >= count || data.length === 0) break;
  }
  return rows;
}

async function softDelete(token, id, reason) {
  return request(
    "POST",
    "/proxy_dms/content/updateContent",
    {
      columnId: String(COL.columnId),
      modelId: String(COL.modelId),
      content: JSON.stringify({
        id: id,
        state: -1,
        c_remarks: "PCTJ_DEDUPE|" + reason + "|" + Date.now(),
      }),
    },
    token
  );
}

function score(row) {
  // 优先保留：有金额/人数、更新时间较新、有 grain
  const should = Number(row.c_should_pay_count || 0) + Number(row.c_actual_pay_count || 0);
  const amt = Number(row.c_should_pay_amount || 0) + Number(row.c_actual_pay_amount || 0);
  const t = Date.parse(row.update_time || row.create_time || 0) || 0;
  const grain = row.c_grain ? 1 : 0;
  return should * 1000 + amt + t / 1e12 + grain * 10;
}

(async () => {
  const token = await login();
  const rows = await listAll(token);
  console.log("可见统计行=" + rows.length + (DRY ? " [dry-run]" : ""));

  const groups = new Map();
  const orphan = [];
  for (const r of rows) {
    const bid = (r.c_batch_id || "").trim();
    const bank = (r.c_pay_bank || "").trim() || "(空)";
    if (!bid) {
      orphan.push(r);
      continue;
    }
    const key = bid + "||" + bank;
    if (!groups.has(key)) groups.set(key, []);
    groups.get(key).push(r);
  }

  const toDrop = [];
  for (const [key, list] of groups) {
    if (list.length <= 1) continue;
    list.sort((a, b) => score(b) - score(a));
    const keep = list[0];
    console.log(
      "重复 " +
        key +
        " x" +
        list.length +
        " 保留 id=" +
        keep.id +
        " 丢弃=" +
        list
          .slice(1)
          .map((x) => x.id)
          .join(",")
    );
    for (let i = 1; i < list.length; i++) {
      toDrop.push({ row: list[i], reason: "dup:" + key });
    }
  }
  for (const r of orphan) {
    console.log("无 batch_id 孤儿 id=" + r.id);
    toDrop.push({ row: r, reason: "orphan_no_batch" });
  }

  console.log("待清理=" + toDrop.length);
  let ok = 0;
  let fail = 0;
  for (const item of toDrop) {
    if (DRY) {
      ok++;
      continue;
    }
    const res = await softDelete(token, item.row.id, item.reason);
    if (res.code == 200) ok++;
    else {
      fail++;
      console.log("FAIL id=" + item.row.id + " " + JSON.stringify(res).slice(0, 200));
    }
  }

  const left = DRY ? rows.length - toDrop.length : (await listAll(token)).length;
  console.log("完成 purged=" + ok + " failed=" + fail + " 剩余可见=" + left);

  // 复检 (batch_id, pay_bank)
  const after = DRY ? rows.filter((r) => !toDrop.some((d) => d.row.id === r.id)) : await listAll(token);
  const chk = new Map();
  for (const r of after) {
    const key = (r.c_batch_id || "").trim() + "||" + ((r.c_pay_bank || "").trim() || "(空)");
    chk.set(key, (chk.get(key) || 0) + 1);
  }
  const still = [...chk.entries()].filter(([, n]) => n > 1);
  console.log("仍重复组数=" + still.length);
  for (const [k, n] of still.slice(0, 10)) console.log("  " + k + " x" + n);
})().catch((e) => {
  console.error(e);
  process.exit(1);
});