@mazjq/zhiro 0.0.0-stage → 0.2.1

package/dist/chunk-4UWMDKMH.js

@@ -0,0 +1,2605 @@
+import {
+  json,
+  now,
+  tx
+} from "./chunk-4EPNDXRS.js";
+import {
+  updateConfig
+} from "./chunk-4JJEOOW2.js";
+import {
+  __name
+} from "./chunk-2SSYR7VW.js";
+
+// apps/daemon/src/data/http.ts
+function httpContext(log = () => {
+}, signal) {
+  const ua = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/128.0 Safari/537.36";
+  const withTimeout = /* @__PURE__ */ __name(async (url, opts = {}) => {
+    const ctl = new AbortController();
+    const t = setTimeout(() => ctl.abort(), opts.timeoutMs ?? 2e4);
+    signal?.addEventListener("abort", () => ctl.abort(), { once: true });
+    try {
+      const res = await fetch(url, { ...opts, headers: { "user-agent": ua, ...opts.headers ?? {} }, signal: ctl.signal }).catch((e) => {
+        throw new Error(`${ctl.signal.aborted ? "TIMEOUT" : "SITE_DOWN"}: ${e?.cause?.code ?? e?.message ?? e} ${url}`);
+      });
+      if (!res.ok) throw new Error(`${res.status === 401 ? "LOGIN_REQUIRED" : res.status === 403 || res.status === 429 ? "BLOCKED" : "SITE_DOWN"}: HTTP ${res.status} ${url}`);
+      if (opts.expectJson !== false && /fetchJson/.test(opts.__kind ?? "") && !/json/.test(res.headers.get("content-type") ?? "")) throw new Error(`SCHEMA_CHANGED: 没返回 JSON ${url}`);
+      return res;
+    } finally {
+      clearTimeout(t);
+    }
+  }, "withTimeout");
+  return {
+    fetchJson: /* @__PURE__ */ __name((url, opts) => withTimeout(url, { ...opts ?? {}, __kind: "fetchJson" }).then((r) => r.json()), "fetchJson"),
+    fetchText: /* @__PURE__ */ __name((url, opts) => withTimeout(url, opts).then((r) => r.text()), "fetchText"),
+    sleep: /* @__PURE__ */ __name((ms) => new Promise((r) => setTimeout(r, ms)), "sleep"),
+    signal,
+    log
+  };
+}
+__name(httpContext, "httpContext");
+var PRIVATE_HOST = /^(localhost|127\.|10\.|192\.168\.|169\.254\.|0\.|\[::1\]|\[fe80|\[fc|\[fd|172\.(1[6-9]|2\d|3[01])\.)/i;
+var publicUrl = /* @__PURE__ */ __name((u) => {
+  try {
+    const x = new URL(u);
+    return /^https?:$/.test(x.protocol) && !PRIVATE_HOST.test(x.hostname) && !/^\d+$/.test(x.hostname);
+  } catch {
+    return false;
+  }
+}, "publicUrl");
+var isLocalPath = /* @__PURE__ */ __name((s) => /^([a-zA-Z]:[\\/]|\\\\|\/)/.test(s.trim()), "isLocalPath");
+
+// apps/daemon/src/data/readers/career-ops/mokahr.mjs
+import { createDecipheriv } from "crypto";
+
+// apps/daemon/src/data/readers/career-ops/_html-entities.mjs
+var NAMED_ENTITIES = {
+  amp: "&",
+  lt: "<",
+  gt: ">",
+  quot: '"',
+  apos: "'",
+  nbsp: " ",
+  // French / Portuguese / Spanish / German / Nordic letters, lower and upper.
+  agrave: "à",
+  aacute: "á",
+  acirc: "â",
+  atilde: "ã",
+  auml: "ä",
+  aring: "å",
+  aelig: "æ",
+  ccedil: "ç",
+  egrave: "è",
+  eacute: "é",
+  ecirc: "ê",
+  euml: "ë",
+  igrave: "ì",
+  iacute: "í",
+  icirc: "î",
+  iuml: "ï",
+  ntilde: "ñ",
+  ograve: "ò",
+  oacute: "ó",
+  ocirc: "ô",
+  otilde: "õ",
+  ouml: "ö",
+  oslash: "ø",
+  ugrave: "ù",
+  uacute: "ú",
+  ucirc: "û",
+  uuml: "ü",
+  yacute: "ý",
+  yuml: "ÿ",
+  szlig: "ß",
+  Agrave: "À",
+  Aacute: "Á",
+  Acirc: "Â",
+  Atilde: "Ã",
+  Auml: "Ä",
+  Aring: "Å",
+  AElig: "Æ",
+  Ccedil: "Ç",
+  Egrave: "È",
+  Eacute: "É",
+  Ecirc: "Ê",
+  Euml: "Ë",
+  Igrave: "Ì",
+  Iacute: "Í",
+  Icirc: "Î",
+  Iuml: "Ï",
+  Ntilde: "Ñ",
+  Ograve: "Ò",
+  Oacute: "Ó",
+  Ocirc: "Ô",
+  Otilde: "Õ",
+  Ouml: "Ö",
+  Oslash: "Ø",
+  Ugrave: "Ù",
+  Uacute: "Ú",
+  Ucirc: "Û",
+  Uuml: "Ü",
+  Yacute: "Ý",
+  // Punctuation these same pages emit around titles.
+  deg: "°",
+  hellip: "…",
+  laquo: "«",
+  raquo: "»",
+  ndash: "–",
+  mdash: "—",
+  lsquo: "‘",
+  rsquo: "’",
+  ldquo: "“",
+  rdquo: "”",
+  middot: "·",
+  euro: "€"
+};
+var CASE_INSENSITIVE_NAMES = /* @__PURE__ */ new Set(["amp", "lt", "gt", "quot", "apos", "nbsp"]);
+function isEmittableCodePoint(code) {
+  return code === 9 || code === 10 || code === 13 || code >= 32 && code <= 55295 || code >= 57344 && code <= 65533 || code >= 65536 && code <= 1114111;
+}
+__name(isEmittableCodePoint, "isEmittableCodePoint");
+function decodeEntities(s) {
+  return s.replace(/&(#[xX][0-9a-fA-F]+|#[0-9]+|[a-zA-Z]+);/g, (m, body) => {
+    if (body[0] === "#") {
+      const isHex = body[1] === "x" || body[1] === "X";
+      const code = parseInt(body.slice(isHex ? 2 : 1), isHex ? 16 : 10);
+      return isEmittableCodePoint(code) ? String.fromCodePoint(code) : m;
+    }
+    if (Object.hasOwn(NAMED_ENTITIES, body)) return NAMED_ENTITIES[body];
+    const lower = body.toLowerCase();
+    return CASE_INSENSITIVE_NAMES.has(lower) ? NAMED_ENTITIES[lower] : m;
+  });
+}
+__name(decodeEntities, "decodeEntities");
+
+// apps/daemon/src/data/readers/career-ops/_html-to-text.mjs
+var DESCRIPTION_CAP = 4e3;
+var HTML_TAG_RE = /<(?:[^>"']|"[^"]*"|'[^']*')+>/g;
+var HTML_MEDIA_RE = /<(script|style)\b(?:[^>"']|"[^"]*"|'[^']*')*>[\s\S]*?<\/\1\s*>/gi;
+function stripMarkup(content) {
+  return content.replace(HTML_MEDIA_RE, " ").replace(HTML_TAG_RE, " ");
+}
+__name(stripMarkup, "stripMarkup");
+function htmlToText(content) {
+  if (typeof content !== "string" || !content) return "";
+  const decoded = decodeEntities(stripMarkup(content));
+  const decodedTwice = decodeEntities(stripMarkup(decoded));
+  return stripMarkup(decodedTwice).replace(/<(?=\/?[a-z!?])/gi, "").replace(/\s+/g, " ").trim().slice(0, DESCRIPTION_CAP);
+}
+__name(htmlToText, "htmlToText");
+
+// apps/daemon/src/data/readers/career-ops/mokahr.mjs
+var API = "https://app.mokahr.com/api/outer/ats-apply/website/jobs/v2";
+var DETAIL_HOST = "app.mokahr.com";
+var AES_IV = Buffer.from("de7c21ed8d6f50fe", "utf8");
+var MAX_LIMIT = 50;
+var DEFAULT_KEYWORDS = [""];
+var DEFAULT_MAX_PAGES = 10;
+var INTER_PAGE_DELAY_MS = 400;
+var TENANT_PATH_RE = /^\/(?:social-recruitment|campus-recruitment|apply)\/([^/]+)\/(\d+)\/?$/;
+var ROBOTS_EXCLUDED_PATHS = /* @__PURE__ */ new Set([
+  "/social-recruitment/lingjuninvest/46355",
+  "/social-recruitment/shopee/74378"
+]);
+function parseTenantUrl(url) {
+  let u;
+  try {
+    u = new URL(url);
+  } catch {
+    return null;
+  }
+  if (u.protocol !== "https:" || u.hostname !== DETAIL_HOST) return null;
+  const m = TENANT_PATH_RE.exec(u.pathname);
+  if (!m) return null;
+  const siteId = Number(m[2]);
+  if (!Number.isSafeInteger(siteId) || siteId <= 0) return null;
+  const pathname = u.pathname.replace(/\/$/, "");
+  if (ROBOTS_EXCLUDED_PATHS.has(pathname)) return null;
+  return { orgId: m[1], siteId, baseUrl: `${u.origin}${pathname}` };
+}
+__name(parseTenantUrl, "parseTenantUrl");
+function decryptMokaHrEnvelope(envelope) {
+  if (!envelope?.data || !envelope?.necromancer) {
+    throw new Error("mokahr: response missing data/necromancer — not the expected envelope shape");
+  }
+  const key = Buffer.from(envelope.necromancer, "utf8");
+  if (key.length !== 16) {
+    throw new Error(`mokahr: necromancer key is ${key.length} bytes, expected 16 (aes-128-cbc)`);
+  }
+  const ciphertext = Buffer.from(envelope.data, "base64");
+  const decipher = createDecipheriv("aes-128-cbc", key, AES_IV);
+  const plain2 = Buffer.concat([decipher.update(ciphertext), decipher.final()]);
+  return JSON.parse(plain2.toString("utf8"));
+}
+__name(decryptMokaHrEnvelope, "decryptMokaHrEnvelope");
+function parseMokaHrJobs(decrypted, companyName, tenantBaseUrl) {
+  const list = decrypted?.data?.jobs;
+  if (!Array.isArray(list)) return [];
+  const jobs = [];
+  for (const j of list) {
+    const title = j?.title;
+    const id = j?.id;
+    if (!title || id == null) continue;
+    const encodedId = encodeURIComponent(String(id));
+    const cities = Array.isArray(j.locations) ? j.locations.map((l) => [l?.provinceName, l?.cityName].filter(Boolean).join(" ")).filter(Boolean).join("/") : "";
+    const createdAt = typeof j.createdAt === "string" ? j.createdAt.trim() : "";
+    const ts = /(?:Z|[+-]\d{2}:\d{2})$/i.test(createdAt) ? Date.parse(createdAt) : NaN;
+    jobs.push({
+      title,
+      // The tenant frontend links each posting through its hash router. The
+      // scanner/tracker normalizers recognize this exact Moka route and promote
+      // its ID into their internal comparison keys before dropping the fragment.
+      url: `${tenantBaseUrl}#/job/${encodedId}`,
+      company: companyName,
+      location: cities,
+      description: [
+        j.commitment && `类型: ${j.commitment}`,
+        j.department?.name && `部门: ${j.department.name}`,
+        htmlToText(j.jobDescription)
+      ].filter(Boolean).join("\n").slice(0, 4e3),
+      postedAt: Number.isFinite(ts) ? ts : void 0
+    });
+  }
+  return jobs;
+}
+__name(parseMokaHrJobs, "parseMokaHrJobs");
+var mokahr_default = {
+  id: "mokahr",
+  detect(entry) {
+    const url = entry.careers_url;
+    if (typeof url !== "string") return null;
+    if (!parseTenantUrl(url)) return null;
+    return { url };
+  },
+  async fetch(entry, ctx) {
+    const tenant = parseTenantUrl(entry.careers_url);
+    if (!tenant) {
+      throw new Error("mokahr: careers_url must be an allowed HTTPS app.mokahr.com tenant URL with a positive site ID");
+    }
+    const keywords = Array.isArray(entry.keywords) && entry.keywords.length ? entry.keywords : DEFAULT_KEYWORDS;
+    const entryMaxPages = Number(entry.max_pages) > 0 ? Number(entry.max_pages) : DEFAULT_MAX_PAGES;
+    const maxPages = Math.min(entryMaxPages, Number(ctx?.maxPages) > 0 ? Number(ctx.maxPages) : Infinity);
+    const seen = /* @__PURE__ */ new Map();
+    const sleep = /* @__PURE__ */ __name((ms) => typeof ctx?.sleep === "function" ? ctx.sleep(ms) : new Promise((r) => setTimeout(r, ms)), "sleep");
+    let firstRequest = true;
+    let succeededOnce = false;
+    for (const keyword of keywords) {
+      for (let page2 = 1; page2 <= maxPages; page2++) {
+        if (firstRequest) firstRequest = false;
+        else await sleep(INTER_PAGE_DELAY_MS);
+        const offset = (page2 - 1) * MAX_LIMIT;
+        let envelope;
+        try {
+          envelope = /** @type {any} */
+          await ctx.fetchJson(API, {
+            method: "POST",
+            headers: { "content-type": "application/json" },
+            body: JSON.stringify({
+              siteId: tenant.siteId,
+              orgId: tenant.orgId,
+              locale: "zh-CN",
+              limit: MAX_LIMIT,
+              offset,
+              ...keyword ? { keyword } : {}
+            }),
+            redirect: "error"
+          });
+        } catch (err) {
+          if (!succeededOnce) throw err;
+          console.error(`  ⚠ mokahr: keyword "${keyword}" page ${page2} failed (${err.message}) — keeping the ${seen.size} jobs collected so far`);
+          return [...seen.values()];
+        }
+        let decrypted;
+        try {
+          decrypted = decryptMokaHrEnvelope(envelope);
+          if (decrypted?.success === false) {
+            throw new Error(`API error: ${decrypted.msg || decrypted.code || "success=false"}`);
+          }
+        } catch (err) {
+          if (!succeededOnce) throw err;
+          console.error(`  ⚠ mokahr: keyword "${keyword}" page ${page2} failed (${err.message}) — keeping the ${seen.size} jobs collected so far`);
+          return [...seen.values()];
+        }
+        succeededOnce = true;
+        const rawJobs = Array.isArray(decrypted?.data?.jobs) ? decrypted.data.jobs : [];
+        if (rawJobs.length === 0) break;
+        const jobs = parseMokaHrJobs(decrypted, entry.name || tenant.orgId, tenant.baseUrl);
+        for (const job of jobs) {
+          if (!seen.has(job.url)) seen.set(job.url, job);
+        }
+        if (rawJobs.length < MAX_LIMIT) break;
+      }
+    }
+    return [...seen.values()];
+  }
+};
+
+// apps/daemon/src/data/readers/moka.ts
+var provider = /* @__PURE__ */ __name(async () => mokahr_default, "provider");
+var TENANT = /^https:\/\/(?:app\.mokahr\.com\/(campus-recruitment|social-recruitment|apply)|[^/]+\/(campus-recruitment|social-recruitment))\/([^/?#]+)\/(\d+)/;
+function mokaTenant(url) {
+  const m = TENANT.exec(url);
+  return m ? { orgId: m[3], siteId: m[4], canonical: `https://app.mokahr.com/${m[1] ?? m[2]}/${m[3]}/${m[4]}` } : null;
+}
+__name(mokaTenant, "mokaTenant");
+function toListing(j, cfg) {
+  const t = mokaTenant(cfg.url ?? "");
+  const idMatch = /#\/job\/([^/?#]+)/.exec(j.url);
+  const desc = j.description ?? "";
+  const dept = /部门:\s*([^\n]+)/.exec(desc)?.[1]?.trim();
+  const commitment = /类型:\s*([^\n]+)/.exec(desc)?.[1]?.trim();
+  const degree = /(博士|硕士及以上|硕士|本科及以上|本科|大专)/.exec(desc)?.[1];
+  return {
+    externalId: idMatch ? decodeURIComponent(idMatch[1]) : j.url,
+    url: j.url,
+    title: j.title,
+    company: cfg.name || j.company || t.orgId,
+    orgUnit: dept,
+    locations: j.location ? j.location.split("/").map((s) => s.trim()).filter(Boolean) : [],
+    description: desc,
+    employmentType: commitment,
+    datePosted: j.postedAt ? new Date(j.postedAt).toISOString() : void 0,
+    degreeReq: degree,
+    cohort: cfg.cohort,
+    batch: cfg.batch ?? (/campus/.test(cfg.url ?? "") ? "校招" : "社招"),
+    kind: cfg.kind ?? (/campus/.test(cfg.url ?? "") ? degree === "博士" ? "博士专项" : "校招" : "社招"),
+    publisher: `moka:${t.orgId}`,
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9,
+    raw: j
+  };
+}
+__name(toListing, "toListing");
+var moka = {
+  id: "moka",
+  name: "Moka",
+  input: "app.mokahr.com/campus-recruitment/<公司>/<站点>（公司自有域名上的同样路径也认）",
+  mode: "pull",
+  describe: "Moka 招聘站（app.mokahr.com 校招/社招租户），走站点前端自己的接口，零 token",
+  detect: /* @__PURE__ */ __name((url) => !!mokaTenant(url), "detect"),
+  siteKey: /* @__PURE__ */ __name((url) => {
+    const t = mokaTenant(url);
+    return `${t.orgId}/${t.siteId}`;
+  }, "siteKey"),
+  async fetch(cfg, ctx) {
+    const tenant = cfg.url ? mokaTenant(cfg.url) : null;
+    if (!tenant) throw new Error(`BAD_URL: Moka：认不出租户 ${cfg.url}`);
+    const p = await provider();
+    const jobs = await p.fetch(
+      { name: cfg.name ?? "", careers_url: tenant.canonical, keywords: cfg.keywords?.length ? cfg.keywords : [""], max_pages: cfg.maxPages ?? 10 },
+      { fetchJson: ctx.fetchJson, fetchText: ctx.fetchText, sleep: ctx.sleep, maxPages: ctx.maxPages, transport: "http" }
+    );
+    ctx.log(`moka ${cfg.name ?? cfg.url}: ${jobs.length} jobs`);
+    return jobs.map((j) => toListing(j, cfg));
+  }
+};
+
+// apps/daemon/src/data/readers/beisen.ts
+var HOST = /^https:\/\/([a-z0-9-]+)\.zhiye\.com(?:\/|$)/i;
+function beisenTenant(url) {
+  const m = HOST.exec(url);
+  return m ? m[1] : null;
+}
+__name(beisenTenant, "beisenTenant");
+var ZERO_DATE = /^0001-01-01/;
+var okDate = /* @__PURE__ */ __name((s) => s && !ZERO_DATE.test(s) ? new Date(s).toISOString() : void 0, "okDate");
+var clean = /* @__PURE__ */ __name((s) => (s ?? "").replace(/<[^>]+>/g, " ").replace(/&nbsp;/g, " ").replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/[ \t]+/g, " ").replace(/\n{3,}/g, "\n\n").trim(), "clean");
+function apiRowToListing(r, tenant, cfg) {
+  const desc = [r.Duty && `岗位职责
+${clean(r.Duty)}`, r.Require && `任职要求
+${clean(r.Require)}`].filter(Boolean).join("\n\n");
+  const degree = r.Degree || /(博士|硕士及以上|硕士|本科及以上|本科|大专)/.exec(desc)?.[1] || void 0;
+  return {
+    externalId: r.Id,
+    url: `https://${tenant}.zhiye.com/campus/jobs/${r.Id}`,
+    title: clean(r.JobAdName),
+    company: cfg.name || tenant,
+    orgUnit: clean(r.Org) || void 0,
+    locations: (r.LocNames ?? []).map((s) => s.replace(/·/g, " ").trim()).filter(Boolean),
+    description: desc,
+    employmentType: r.Kind || void 0,
+    datePosted: okDate(r.PostDate),
+    validThrough: okDate(r.EndTime),
+    degreeReq: degree,
+    cohort: cfg.cohort,
+    batch: cfg.batch ?? "校招",
+    kind: cfg.kind ?? (degree === "博士" ? "博士专项" : "校招"),
+    publisher: `beisen:${tenant}`,
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9,
+    raw: r
+  };
+}
+__name(apiRowToListing, "apiRowToListing");
+function parseCmsList(html) {
+  const out = [];
+  const re = /<a\s+title="([^"]*)"\s+href="\/zpdetail\/(\d+)(?:\?[^"]*)?"[^>]*>[\s\S]*?<\/a>\s*<\/td>([\s\S]*?)<\/tr>/g;
+  for (const m of html.matchAll(re)) {
+    const cells = [...m[3].matchAll(/<td(?:\s+title="([^"]*)")?[^>]*>([\s\S]*?)<\/td>/g)].map((c) => clean(c[1] ?? c[2]));
+    out.push({ id: m[2], title: clean(m[1]), dept: cells[0] ?? "", loc: cells[1] ?? "", date: cells[2] ?? "" });
+  }
+  return out;
+}
+__name(parseCmsList, "parseCmsList");
+function parseCmsDetail(html) {
+  const body = html.replace(/<script[\s\S]*?<\/script>/g, "").replace(/<style[\s\S]*?<\/style>/g, "");
+  const i = body.search(/岗位职责|职位描述|工作职责|任职要求|岗位要求/);
+  const slice = i >= 0 ? body.slice(i - 100, i + 6e3) : body;
+  return clean(slice).slice(0, 4e3);
+}
+__name(parseCmsDetail, "parseCmsDetail");
+async function tryApi(tenant, cfg, ctx) {
+  const categories = cfg.categories ?? ["2"];
+  const rows = [];
+  const pageSize = 50;
+  for (let page2 = 0; page2 < (cfg.maxPages ?? 10); page2++) {
+    let res;
+    try {
+      res = await ctx.fetchJson(`https://${tenant}.zhiye.com/api/Jobad/GetJobAdPageList`, {
+        method: "POST",
+        headers: { "content-type": "application/json" },
+        body: JSON.stringify({ PageIndex: page2, PageSize: pageSize, Category: categories, KeyWords: cfg.keywords?.[0] ?? "", SpecialType: 0, PortalId: "", DisplayFields: ["Category", "Kind", "LocId", "Org"] }),
+        redirect: "manual"
+      });
+    } catch {
+      return page2 === 0 ? null : rows;
+    }
+    if (!res || res.Code !== 200 || !Array.isArray(res.Data)) return page2 === 0 ? null : rows;
+    rows.push(...res.Data);
+    if (rows.length >= (res.Count ?? 0) || res.Data.length < pageSize) break;
+    await ctx.sleep(400);
+  }
+  return rows;
+}
+__name(tryApi, "tryApi");
+async function cmsFallback(tenant, cfg, ctx) {
+  const base = `https://${tenant}.zhiye.com`;
+  const listPath = cfg.listPath ?? "/campus";
+  const seen = /* @__PURE__ */ new Map();
+  for (let page2 = 0; page2 < (cfg.maxPages ?? 10); page2++) {
+    const html = await ctx.fetchText(page2 === 0 ? `${base}${listPath}` : `${base}${listPath}?PageIndex=${page2}`);
+    const rows = parseCmsList(html);
+    if (rows.length === 0 || rows.every((r) => seen.has(r.id))) break;
+    for (const r of rows) seen.set(r.id, r);
+    await ctx.sleep(400);
+  }
+  const out = [];
+  const detailCap = cfg.detailCap ?? 60;
+  let fetched = 0;
+  for (const r of seen.values()) {
+    let description = "";
+    if (fetched < detailCap) {
+      try {
+        description = parseCmsDetail(await ctx.fetchText(`${base}/zpdetail/${r.id}`));
+        fetched++;
+        await ctx.sleep(300);
+      } catch {
+      }
+    }
+    const degree = /(博士|硕士及以上|硕士|本科及以上|本科|大专)/.exec(description)?.[1];
+    out.push({
+      externalId: r.id,
+      url: `${base}/zpdetail/${r.id}`,
+      title: r.title,
+      company: cfg.name || tenant,
+      orgUnit: r.dept || void 0,
+      locations: r.loc.split(/[,，]/).map((s) => s.trim()).filter(Boolean),
+      description,
+      datePosted: /^\d{4}-\d{2}-\d{2}$/.test(r.date) ? new Date(r.date).toISOString() : void 0,
+      degreeReq: degree,
+      cohort: cfg.cohort,
+      batch: cfg.batch ?? "校招",
+      kind: cfg.kind ?? (degree === "博士" ? "博士专项" : "校招"),
+      publisher: `beisen:${tenant}`,
+      grade: "官方",
+      isOriginal: true,
+      confidence: 0.85,
+      raw: r
+    });
+  }
+  return out;
+}
+__name(cmsFallback, "cmsFallback");
+var beisen = {
+  id: "beisen",
+  name: "北森",
+  input: "<公司>.zhiye.com",
+  mode: "pull",
+  describe: "北森招聘门户（<租户>.zhiye.com）：新版走 GetJobAdPageList 接口，老版解析 /campus 列表页；零 token",
+  detect: /* @__PURE__ */ __name((url) => !!beisenTenant(url), "detect"),
+  siteKey: /* @__PURE__ */ __name((url) => beisenTenant(url), "siteKey"),
+  async fetch(cfg, ctx) {
+    const tenant = beisenTenant(cfg.url ?? "");
+    if (!tenant) throw new Error(`BAD_URL: 北森：认不出租户 ${cfg.url}`);
+    const rows = await tryApi(tenant, cfg, ctx);
+    if (rows) {
+      ctx.log(`beisen ${tenant}: api ${rows.length} rows`);
+      return rows.map((r) => apiRowToListing(r, tenant, cfg));
+    }
+    const listings = await cmsFallback(tenant, cfg, ctx);
+    ctx.log(`beisen ${tenant}: cms ${listings.length} rows`);
+    return listings;
+  }
+};
+
+// apps/daemon/src/data/readers/feishu-base.ts
+import { gunzipSync } from "node:zlib";
+
+// apps/daemon/src/data/catalog.ts
+import { ulid } from "ulid";
+import { createHash } from "node:crypto";
+var genericCompany = /* @__PURE__ */ __name((name) => /^(本|我|该|某|贵|总)?(公司|司|单位|集团|企业|院|所)$|^(unknown|n\/a|null|none|未知|无|暂无)$/i.test(name.trim()), "genericCompany");
+function companyKey(name) {
+  return name.normalize("NFKC").replace(/\s+/g, "").replace(/(有限责任公司|有限公司)$/, "").toLowerCase();
+}
+__name(companyKey, "companyKey");
+function sha(s) {
+  return createHash("sha256").update(s).digest("hex");
+}
+__name(sha, "sha");
+function campaignKey(sourceId, publisher, company, cohort, batch) {
+  return sha(`${sourceId}|${publisher}|${company}|${cohort}|${batch}`);
+}
+__name(campaignKey, "campaignKey");
+var findCompany = /* @__PURE__ */ __name((db, name) => db.prepare("select id from company where name = ?").get(name) ?? db.prepare("select id from company where norm = ? or aliases like ?").get(companyKey(name), `%${JSON.stringify(name)}%`), "findCompany");
+function upsertCompany(db, name, type = "", industry = "") {
+  const found = findCompany(db, name);
+  if (found) {
+    if (type || industry) db.prepare("update company set type = coalesce(nullif(?, ''), type), industry = coalesce(nullif(?, ''), industry) where id = ?").run(type, industry, found.id);
+    return found.id;
+  }
+  const id = ulid();
+  db.prepare("insert into company(id, name, norm, type, industry, created_at) values (?,?,?,?,?,?)").run(id, name, companyKey(name), type, industry, now());
+  return id;
+}
+__name(upsertCompany, "upsertCompany");
+function upsertCampaign(db, c) {
+  const found = db.prepare("select id from campaign where external_key = ?").get(c.externalKey);
+  const t = now();
+  if (found) {
+    db.prepare(`update campaign set kind=?, cohort=?, batch=?, cities=?, degree_req=?, major_req=?, opens_at=?, closes_at=?, deadline_kind=?, deadline_text=?,
+      apply_url=?, notice_url=?, source_grade=?, source_kind=?, has_written_test=?, notes=?, last_verified_at=coalesce(?, last_verified_at), updated_at=? where id=?`).run(
+      c.kind,
+      c.cohort,
+      c.batch,
+      json.str(c.cities),
+      c.degreeReq,
+      c.majorReq,
+      c.opensAt,
+      c.closesAt,
+      c.deadlineKind,
+      c.deadlineText,
+      c.applyUrl,
+      c.noticeUrl,
+      c.sourceGrade,
+      c.sourceKind,
+      c.hasWrittenTest,
+      c.notes,
+      c.lastVerifiedAt ?? null,
+      t,
+      found.id
+    );
+    return { id: found.id, created: false };
+  }
+  const id = ulid();
+  db.prepare(`insert into campaign(id, company_id, external_key, kind, cohort, batch, cities, degree_req, major_req, opens_at, closes_at, deadline_kind, deadline_text,
+    apply_url, notice_url, source_grade, source_kind, has_written_test, notes, last_verified_at, created_at, updated_at) values (?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?,?)`).run(
+    id,
+    c.companyId,
+    c.externalKey,
+    c.kind,
+    c.cohort,
+    c.batch,
+    json.str(c.cities),
+    c.degreeReq,
+    c.majorReq,
+    c.opensAt,
+    c.closesAt,
+    c.deadlineKind,
+    c.deadlineText,
+    c.applyUrl,
+    c.noticeUrl,
+    c.sourceGrade,
+    c.sourceKind,
+    c.hasWrittenTest,
+    c.notes,
+    c.lastVerifiedAt ?? null,
+    t,
+    t
+  );
+  return { id, created: true };
+}
+__name(upsertCampaign, "upsertCampaign");
+var findPostByKey = /* @__PURE__ */ __name((db, campaignId, key) => db.prepare("select id from position where campaign_id = ? and source_key = ?").get(campaignId, key), "findPostByKey");
+var findPostByTitle = /* @__PURE__ */ __name((db, campaignId, title, keyed) => db.prepare(`select id from position where campaign_id = ? and title = ? ${keyed ? "and source_key = ''" : "order by source_key = '' desc"} limit 1`).get(campaignId, title), "findPostByTitle");
+var UPDATE_POST = `update position set title = ?, jd_text = coalesce(nullif(?, ''), jd_text), major_req = coalesce(nullif(?, ''), major_req), degree_req = coalesce(nullif(?, ''), degree_req), closes_at = coalesce(?, closes_at),
+  cities = case when ? != '[]' then ? else cities end, source_key = case when ? != '' then ? else source_key end, status = 'open', last_seen_at = ?, updated_at = ? where id = ?`;
+var INSERT_POST = "insert into position(id, campaign_id, title, jd_text, cities, degree_req, major_req, closes_at, status, direction_tags, last_seen_at, created_at, updated_at, source_key) values (?,?,?,?,?,?,?,?,?,?,?,?,?,?)";
+var updatePost = /* @__PURE__ */ __name((db, id, title, extra, t) => {
+  const key = extra.sourceKey ?? "", cities = json.str(extra.cities ?? []);
+  db.prepare(UPDATE_POST).run(title, extra.jdText ?? "", extra.majorReq ?? "", extra.degreeReq ?? "", extra.closesAt ?? null, cities, cities, key, key, t, t, id);
+}, "updatePost");
+var insertPost = /* @__PURE__ */ __name((db, campaignId, title, extra, t) => {
+  const id = ulid();
+  db.prepare(INSERT_POST).run(id, campaignId, title, extra.jdText ?? "", json.str(extra.cities ?? []), extra.degreeReq ?? "", extra.majorReq ?? "", extra.closesAt ?? null, "open", json.str(extra.directionTags ?? []), t, t, t, extra.sourceKey ?? "");
+  return id;
+}, "insertPost");
+function upsertPosition(db, campaignId, title, extra = {}) {
+  const key = extra.sourceKey ?? "";
+  const found = (key ? findPostByKey(db, campaignId, key) : void 0) ?? findPostByTitle(db, campaignId, title, !!key);
+  const t = now();
+  if (found) {
+    updatePost(db, found.id, title, extra, t);
+    return { id: found.id, created: false };
+  }
+  return { id: insertPost(db, campaignId, title, extra, t), created: true };
+}
+__name(upsertPosition, "upsertPosition");
+var CAMPAIGN_COLS = "id, external_key, kind, cohort, batch, cities, degree_req, major_req, opens_at, closes_at, deadline_kind, deadline_text, apply_url, notice_url, source_grade, source_kind, has_written_test, notes, last_verified_at";
+var POST_COLS = "id, campaign_id, title, source_key, jd_text, major_req, degree_req, closes_at, cities, status";
+var inChunks = /* @__PURE__ */ __name((xs, f) => {
+  for (let i = 0; i < xs.length; i += 500) {
+    const part = xs.slice(i, i + 500);
+    f(part, part.map(() => "?").join(","));
+  }
+}, "inChunks");
+var sameCampaign = /* @__PURE__ */ __name((r, c) => r.kind === c.kind && r.cohort === c.cohort && r.batch === c.batch && r.cities === json.str(c.cities) && r.degree_req === c.degreeReq && r.major_req === c.majorReq && r.opens_at === (c.opensAt ?? null) && r.closes_at === (c.closesAt ?? null) && r.deadline_kind === c.deadlineKind && r.deadline_text === c.deadlineText && r.apply_url === c.applyUrl && r.notice_url === c.noticeUrl && r.source_grade === c.sourceGrade && r.source_kind === c.sourceKind && r.has_written_test === c.hasWrittenTest && r.notes === c.notes, "sameCampaign");
+var Posts = class {
+  static {
+    __name(this, "Posts");
+  }
+  rows = [];
+  byTitle = /* @__PURE__ */ new Map();
+  byKey = /* @__PURE__ */ new Map();
+  add(r) {
+    this.rows.push(r);
+    this.put(r);
+  }
+  put(r) {
+    (this.byTitle.get(r.title) ?? this.byTitle.set(r.title, []).get(r.title)).push(r);
+    (this.byKey.get(r.source_key) ?? this.byKey.set(r.source_key, []).get(r.source_key)).push(r);
+  }
+  drop(r) {
+    for (const xs of [this.byTitle.get(r.title), this.byKey.get(r.source_key)]) xs.splice(xs.indexOf(r), 1);
+  }
+  titled(t) {
+    return this.byTitle.get(t) ?? [];
+  }
+  keyed(k) {
+    return this.byKey.get(k) ?? [];
+  }
+  /** the row as the database now holds it (title and site id included, so the lookups follow) */
+  set(r, next) {
+    this.drop(r);
+    Object.assign(r, next);
+    this.put(r);
+  }
+};
+var CatalogBatch = class {
+  constructor(db, externalKeys) {
+    this.db = db;
+    this.st = {
+      name: db.prepare("select id from company where name = ?"),
+      company: db.prepare("select type, industry from company where id = ?"),
+      campaign: db.prepare(`select ${CAMPAIGN_COLS} from campaign where external_key = ?`),
+      postsOf: db.prepare(`select ${POST_COLS} from position where campaign_id = ?`),
+      post: db.prepare(`select ${POST_COLS} from position where id = ?`)
+    };
+    inChunks([...new Set(externalKeys)], (part, q) => {
+      for (const r of db.prepare(`select ${CAMPAIGN_COLS} from campaign where external_key in (${q})`).all(...part)) this.campaigns.set(r.external_key, r);
+    });
+    const ids = [...this.campaigns.values()].map((c) => c.id);
+    for (const id of ids) this.posts.set(id, new Posts());
+    inChunks(ids, (part, q) => {
+      for (const r of db.prepare(`select ${POST_COLS} from position where campaign_id in (${q})`).all(...part)) this.posts.get(r.campaign_id).add(r);
+    });
+  }
+  db;
+  static {
+    __name(this, "CatalogBatch");
+  }
+  /** company name → id; one found by its norm or an alias may resolve elsewhere once a company is added */
+  companies = /* @__PURE__ */ new Map();
+  byNorm = /* @__PURE__ */ new Set();
+  companyRows = /* @__PURE__ */ new Map();
+  campaigns = /* @__PURE__ */ new Map();
+  // external_key → row
+  targets = /* @__PURE__ */ new Map();
+  // campaign id → what its last row says
+  posts = /* @__PURE__ */ new Map();
+  // campaign id → its posts
+  seen = /* @__PURE__ */ new Set();
+  // posts listed again exactly as they are
+  /** companies, campaigns and posts already there that this batch changed (a new one is not counted) */
+  updated = 0;
+  added = /* @__PURE__ */ new Set();
+  // ids this batch created: what a later row changes in them is part of being new
+  edit(id) {
+    if (!this.added.has(id)) this.updated++;
+  }
+  st;
+  /** The campaigns this batch knows of so far (read up front, or added). */
+  campaignIds() {
+    return [...this.campaigns.values()].map((c) => c.id);
+  }
+  company(name, type = "", industry = "") {
+    let id = this.companies.get(name);
+    if (id === void 0) {
+      const exact = this.st.name.get(name);
+      const found = exact ?? findCompany(this.db, name);
+      if (!found) {
+        id = upsertCompany(this.db, name, type, industry);
+        for (const n of this.byNorm) this.companies.delete(n);
+        this.byNorm.clear();
+        this.companies.set(name, id);
+        this.companyRows.set(id, { type, industry });
+        this.added.add(id);
+        return id;
+      }
+      id = found.id;
+      this.companies.set(name, id);
+      if (!exact) this.byNorm.add(name);
+    }
+    if (type || industry) {
+      const cur = this.companyRows.get(id) ?? { ...this.st.company.get(id) };
+      this.companyRows.set(id, cur);
+      const t = type || cur.type, i = industry || cur.industry;
+      if (t !== cur.type || i !== cur.industry) {
+        this.db.prepare("update company set type = ?, industry = ? where id = ?").run(t, i, id);
+        cur.type = t;
+        cur.industry = i;
+        this.edit(id);
+      }
+    }
+    return id;
+  }
+  /** The campaign's id. Its fields are written at `finish`, from its last row — as the full rewrite leaves them. */
+  campaign(c) {
+    let cur = this.campaigns.get(c.externalKey);
+    if (!cur) {
+      cur = this.st.campaign.get(c.externalKey);
+      if (!cur) {
+        const { id } = upsertCampaign(this.db, c);
+        this.campaigns.set(c.externalKey, this.db.prepare(`select ${CAMPAIGN_COLS} from campaign where id = ?`).get(id));
+        this.posts.set(id, new Posts());
+        this.added.add(id);
+        return { id, created: true };
+      }
+      this.campaigns.set(c.externalKey, cur);
+    }
+    this.targets.set(cur.id, c);
+    return { id: cur.id, created: false };
+  }
+  position(campaignId, title, extra = {}) {
+    let posts = this.posts.get(campaignId);
+    if (!posts) {
+      this.posts.set(campaignId, posts = new Posts());
+      for (const r of this.st.postsOf.all(campaignId)) posts.add(r);
+    }
+    const key = extra.sourceKey ?? "";
+    const asked = /* @__PURE__ */ __name((r) => r ? posts.rows.find((x) => x.id === r.id) : void 0, "asked");
+    let found;
+    if (key) {
+      const m = posts.keyed(key);
+      found = m.length > 1 ? asked(findPostByKey(this.db, campaignId, key)) : m[0];
+    }
+    if (!found) {
+      const m = posts.titled(title);
+      if (key) found = m.find((r) => r.source_key === "");
+      else found = m.find((r) => r.source_key === "") ?? (m.length > 1 ? asked(findPostByTitle(this.db, campaignId, title, false)) : m[0]);
+    }
+    if (!found) {
+      const id = insertPost(this.db, campaignId, title, extra, now());
+      posts.add(this.st.post.get(id));
+      this.added.add(id);
+      return { id, created: true };
+    }
+    const cities = json.str(extra.cities ?? []);
+    const next = {
+      title,
+      jd_text: extra.jdText || found.jd_text,
+      major_req: extra.majorReq || found.major_req,
+      degree_req: extra.degreeReq || found.degree_req,
+      closes_at: extra.closesAt ?? found.closes_at,
+      cities: cities !== "[]" ? cities : found.cities,
+      source_key: key !== "" ? key : found.source_key,
+      status: "open"
+    };
+    if (Object.keys(next).some((k) => next[k] !== found[k])) {
+      updatePost(this.db, found.id, title, extra, now());
+      posts.set(found, next);
+      this.seen.delete(found.id);
+      this.edit(found.id);
+    } else this.seen.add(found.id);
+    return { id: found.id, created: false };
+  }
+  /** Write each campaign from its last row; mark the posts listed again unchanged as seen. */
+  finish() {
+    const verified = /* @__PURE__ */ new Map();
+    for (const cur of this.campaigns.values()) {
+      const c = this.targets.get(cur.id);
+      if (!c) continue;
+      if (!sameCampaign(cur, c)) {
+        upsertCampaign(this.db, c);
+        this.edit(cur.id);
+      } else if (c.lastVerifiedAt != null && c.lastVerifiedAt !== cur.last_verified_at) (verified.get(c.lastVerifiedAt) ?? verified.set(c.lastVerifiedAt, []).get(c.lastVerifiedAt)).push(cur.id);
+    }
+    for (const [stamp, ids] of verified) this.db.prepare("update campaign set last_verified_at = ? where id in (select value from json_each(?))").run(stamp, JSON.stringify(ids));
+    if (this.seen.size) this.db.prepare("update position set last_seen_at = ? where id in (select value from json_each(?))").run(now(), JSON.stringify([...this.seen]));
+  }
+};
+
+// apps/daemon/src/data/sheet.ts
+function parseCsv(text3) {
+  const rows = [];
+  let row = [];
+  let field = "";
+  let inQ = false;
+  const s = text3.replace(/^/, "");
+  for (let i = 0; i < s.length; i++) {
+    const c = s[i];
+    if (inQ) {
+      if (c === '"') {
+        if (s[i + 1] === '"') {
+          field += '"';
+          i++;
+        } else inQ = false;
+      } else field += c;
+    } else if (c === '"') inQ = true;
+    else if (c === ",") {
+      row.push(field);
+      field = "";
+    } else if (c === "\n" || c === "\r") {
+      if (c === "\r" && s[i + 1] === "\n") i++;
+      row.push(field);
+      field = "";
+      rows.push(row);
+      row = [];
+    } else field += c;
+  }
+  if (field.length || row.length) {
+    row.push(field);
+    rows.push(row);
+  }
+  return rows.filter((r) => r.some((f) => f.trim() !== ""));
+}
+__name(parseCsv, "parseCsv");
+function parseDeadline(text3) {
+  const t = (text3 ?? "").trim();
+  if (!t) return { closesAt: null, kind: "unknown" };
+  if (/招满|长期|滚动|持续|尽快|随时/.test(t)) return { closesAt: null, kind: /滚动|持续|尽快|随时/.test(t) ? "rolling" : "until_filled" };
+  const m = t.match(/(\d{4})[-/.年](\d{1,2})[-/.月](\d{1,2})/);
+  if (m) return { closesAt: `${m[1]}-${m[2].padStart(2, "0")}-${m[3].padStart(2, "0")}`, kind: "date" };
+  const md = t.match(/(\d{1,2})月(\d{1,2})日?/);
+  if (md) {
+    const y = (/* @__PURE__ */ new Date()).getFullYear();
+    return { closesAt: `${y}-${md[1].padStart(2, "0")}-${md[2].padStart(2, "0")}`, kind: "date" };
+  }
+  return { closesAt: null, kind: "unknown" };
+}
+__name(parseDeadline, "parseDeadline");
+function splitList(s) {
+  return (s ?? "").split(/[，,、;；|\s]+/).map((x) => x.trim()).filter(Boolean);
+}
+__name(splitList, "splitList");
+function gradeFromSource(source, noticeUrl) {
+  if (/官网|官方|校园招聘网|careers|mokahr|beisen|zhiye/i.test(source + noticeUrl)) return "官方";
+  if (/就业|高校|大学|学院/.test(source)) return "校方转载";
+  if (/mp\.weixin|公众号/.test(source + noticeUrl)) return "校方转载";
+  return "聚合";
+}
+__name(gradeFromSource, "gradeFromSource");
+var FIELDS = {
+  company: "公司",
+  positions: "岗位",
+  deadline: "截止时间",
+  applyUrl: "投递链接",
+  noticeUrl: "公告链接",
+  cities: "工作地点",
+  degree: "学历要求",
+  major: "专业要求",
+  cohort: "届次",
+  batch: "批次",
+  companyType: "企业类型",
+  industry: "行业",
+  source: "信息来源"
+};
+var BUILTIN_COLUMNS = {
+  company: "公司",
+  positions: "招聘岗位",
+  deadline: "截止时间",
+  applyUrl: "简历投递链接",
+  noticeUrl: "公告链接",
+  cities: "工作地点",
+  degree: "学历要求",
+  major: "专业要求",
+  cohort: "招聘届次",
+  batch: "批次",
+  companyType: "企业类型",
+  industry: "行业类别",
+  source: "公告来源"
+};
+var ALIASES = {
+  company: ["公司", "公司名称", "企业", "企业名称", "单位", "单位名称", "用人单位", "招聘单位", "company", "employer"],
+  positions: ["招聘岗位", "岗位", "岗位名称", "职位", "职位名称", "招聘职位", "岗位类别", "position", "title", "job"],
+  deadline: ["截止时间", "截止日期", "截止", "网申截止", "投递截止", "报名截止", "ddl", "deadline"],
+  applyUrl: ["简历投递链接", "投递链接", "网申链接", "网申地址", "投递地址", "申请链接", "投递方式", "apply"],
+  noticeUrl: ["公告链接", "招聘公告", "原文链接", "详情链接", "公告", "链接", "notice", "url", "link"],
+  cities: ["工作地点", "地点", "城市", "工作城市", "location", "city"],
+  degree: ["学历要求", "学历", "degree"],
+  major: ["专业要求", "专业", "major"],
+  cohort: ["招聘届次", "届次", "届别", "毕业年份", "cohort"],
+  batch: ["批次", "招聘批次", "招聘类型", "batch"],
+  companyType: ["企业类型", "公司类型", "单位性质", "企业性质"],
+  industry: ["行业类别", "行业", "所属行业", "industry"],
+  source: ["公告来源", "信息来源", "来源", "source"]
+};
+var plain = /* @__PURE__ */ __name((s) => s.normalize("NFKC").replace(/[\s:：()（）_\-/]/g, "").toLowerCase(), "plain");
+function guessColumns(headers) {
+  const out = {}, taken = /* @__PURE__ */ new Set();
+  for (const exact of [true, false]) for (const f of Object.keys(ALIASES)) {
+    if (out[f]) continue;
+    const h = headers.find((h2) => !taken.has(h2) && ALIASES[f].some((a) => exact ? plain(h2) === plain(a) : plain(h2).includes(plain(a))));
+    if (h) {
+      out[f] = h;
+      taken.add(h);
+    }
+  }
+  return out;
+}
+__name(guessColumns, "guessColumns");
+function splitCell(s) {
+  const cjk = /[\u3400-\u9fff]/.test(s ?? "");
+  return (s ?? "").split(cjk ? /[，,、;；|\s]+/ : /[，,、;；|\n\r]+/).map((x) => x.trim()).filter(Boolean);
+}
+__name(splitCell, "splitCell");
+function tableToListings(table, columns, cfg, today = (/* @__PURE__ */ new Date()).toISOString().slice(0, 10), publisher = "汇总表") {
+  const get = /* @__PURE__ */ __name((row, f, want) => columns[f] ? row.get(columns[f], want) : "", "get");
+  const wantCohort = (cfg.cohort ?? "").replace(/\D/g, "");
+  const split = columns === BUILTIN_COLUMNS ? splitList : splitCell;
+  const out = [];
+  for (const row of table.rows) {
+    const company = get(row, "company").trim();
+    const cohort = get(row, "cohort");
+    if (!company || wantCohort && columns.cohort && !cohort.includes(wantCohort)) continue;
+    const deadlineText = get(row, "deadline");
+    const dl = parseDeadline(deadlineText);
+    if (dl.closesAt && dl.closesAt < today) continue;
+    const batch = get(row, "batch"), sourceText = get(row, "source");
+    const noticeUrl = get(row, "noticeUrl", "link") || get(row, "noticeUrl");
+    const applyUrl = get(row, "applyUrl", "link") || get(row, "applyUrl");
+    const titles = split(get(row, "positions")).slice(0, 40);
+    const g = gradeFromSource(sourceText, noticeUrl);
+    const rowId = row.id || sha(`${company}|${cohort}|${batch}`).slice(0, 16);
+    for (const title of titles.length ? titles : [batch || "校招"]) {
+      out.push({
+        externalId: `${rowId}|${title}`,
+        url: applyUrl || noticeUrl,
+        noticeUrl,
+        title,
+        company,
+        companyType: get(row, "companyType"),
+        industry: get(row, "industry"),
+        locations: split(get(row, "cities")),
+        deadlineText,
+        degreeReq: get(row, "degree"),
+        majorReq: get(row, "major"),
+        cohort,
+        batch,
+        kind: /博士/.test(batch + title) ? "博士专项" : /实习/.test(batch) ? "实习" : "校招",
+        publisher: cfg.name || publisher,
+        grade: g === "官方" ? "聚合" : g,
+        isOriginal: false,
+        confidence: 0.7
+      });
+    }
+  }
+  return out;
+}
+__name(tableToListings, "tableToListings");
+function listingsFromTable(table, cfg, publisher) {
+  const columns = cfg.columns ?? BUILTIN_COLUMNS;
+  if (!columns.company || !table.headers.includes(columns.company)) throw new Error(`NO_COMPANY_COLUMN: 表里没有「${columns.company ?? "公司"}」这一列（现有的列：${table.headers.slice(0, 12).join("、")}）`);
+  return tableToListings(table, columns, cfg, void 0, publisher);
+}
+__name(listingsFromTable, "listingsFromTable");
+function csvToTable(text3) {
+  const [head = [], ...rows] = parseCsv(text3);
+  const headers = head.map((h) => h.trim());
+  const at = new Map(headers.map((h, i) => [h, i]));
+  return { headers, rows: rows.map((r) => ({ id: "", get: /* @__PURE__ */ __name((h) => (r[at.get(h) ?? -1] ?? "").trim(), "get") })) };
+}
+__name(csvToTable, "csvToTable");
+
+// apps/daemon/src/data/readers/feishu-base.ts
+var UA = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/140.0.0.0 Safari/537.36";
+var PAGE = 3e3;
+function parseBaseUrl(url) {
+  try {
+    const u = new URL(url);
+    const m = /^\/base\/([A-Za-z0-9]+)/.exec(u.pathname);
+    const table = u.searchParams.get("table");
+    if (!/\.(feishu\.cn|larksuite\.com)$/.test(u.hostname) || !m || !table) return null;
+    return { origin: u.origin, base: m[1], table, view: u.searchParams.get("view") ?? "" };
+  } catch {
+    return null;
+  }
+}
+__name(parseBaseUrl, "parseBaseUrl");
+async function anonymousCookies(url, signal) {
+  const jar = /* @__PURE__ */ new Map();
+  let next = url;
+  for (let i = 0; i < 12; i++) {
+    const res = await fetch(next, { redirect: "manual", signal, headers: { "user-agent": UA, "accept-language": "zh-CN,zh", cookie: [...jar].map(([k, v]) => `${k}=${v}`).join("; ") } });
+    for (const c of res.headers.getSetCookie()) {
+      const [kv] = c.split(";");
+      const eq = kv.indexOf("=");
+      jar.set(kv.slice(0, eq).trim(), kv.slice(eq + 1).trim());
+    }
+    const loc = res.headers.get("location");
+    if (res.status < 300 || res.status >= 400 || !loc) break;
+    next = new URL(loc, next).href;
+  }
+  if (!jar.has("session")) throw new Error("NOT_PUBLIC: 这张表没有开「互联网上获得链接的人可阅读」，或链接已失效");
+  return [...jar].map(([k, v]) => `${k}=${v}`).join("; ");
+}
+__name(anonymousCookies, "anonymousCookies");
+var unzip = /* @__PURE__ */ __name((s) => JSON.parse(gunzipSync(Buffer.from(s, "base64")).toString("utf8")), "unzip");
+function toTable(fieldMap, recordMap) {
+  const byName = Object.fromEntries(Object.entries(fieldMap).map(([id, f]) => [f.name.trim(), { id, ...f }]));
+  const cell = /* @__PURE__ */ __name((rec, name, want = "text") => {
+    const f = byName[name];
+    const v = f && rec[f.id]?.value;
+    if (v == null) return "";
+    const opt = /* @__PURE__ */ __name((id) => f.property?.options?.find((o) => o.id === id)?.name ?? "", "opt");
+    if (typeof v === "number") return new Date(v + 8 * 36e5).toISOString().slice(0, 10);
+    if (typeof v === "string") return opt(v) || v;
+    if (Array.isArray(v)) return v.map((x) => typeof x === "string" ? opt(x) : want === "link" ? x?.link ?? "" : x?.text ?? "").filter(Boolean).join(typeof v[0] === "string" ? "、" : "");
+    return "";
+  }, "cell");
+  return { headers: Object.keys(byName), rows: Object.entries(recordMap).map(([id, rec]) => ({ id, get: /* @__PURE__ */ __name((h, want) => cell(rec, h, want), "get") })) };
+}
+__name(toTable, "toTable");
+var feishuBase = {
+  id: "feishu-base",
+  name: "飞书多维表格",
+  input: "公开的 feishu.cn/base/…?table=… 链接",
+  aggregate: true,
+  mode: "pull",
+  describe: "飞书多维表格（公开链接，匿名读取）：按订阅里存的列对应读成岗位，没有对应时按内置 17 列格式；按届次过滤，跳过已截止的行",
+  detect: /* @__PURE__ */ __name((url) => !!parseBaseUrl(url), "detect"),
+  siteKey: /* @__PURE__ */ __name((url) => {
+    const u = parseBaseUrl(url);
+    return `${u.base}/${u.table}`;
+  }, "siteKey"),
+  async table(config, ctx) {
+    const { fieldMap, records } = await readBase(config.url ?? "", (m) => ctx.log(`feishu-base ${config.id}: ${m}`), ctx.signal);
+    return toTable(fieldMap, records);
+  },
+  async fetch(config, ctx) {
+    return listingsFromTable(await this.table(config, ctx), config, "飞书汇总表");
+  }
+};
+async function readBase(url, log = () => {
+}, signal) {
+  const u = parseBaseUrl(url);
+  if (!u) throw new Error(`BAD_URL: 需要 https://<tenant>.feishu.cn/base/<token>?table=…，收到 ${url}`);
+  const cookie = await anonymousCookies(url, signal);
+  const api2 = `${u.origin}/space/api/v1/bitable/${u.base}`;
+  const get = /* @__PURE__ */ __name(async (path) => {
+    const res = await fetch(api2 + path, { signal: signal ?? AbortSignal.timeout(9e4), headers: { cookie, "user-agent": UA, referer: `${u.origin}/base/${u.base}` } });
+    const j = await res.json();
+    if (j.code !== 0) throw new Error(`FAILED: 飞书接口 ${j.code}：${j.msg}`);
+    return j.data;
+  }, "get");
+  const q = `tableID=${u.table}&viewID=${u.view}`;
+  const head = unzip((await get(`/clientvars?${q}&recordLimit=1&ondemandLimit=1&needBase=false&viewLazyLoad=true&openType=0`)).table);
+  const total = head.meta?.recordsNum ?? 0;
+  const records = {};
+  for (let offset = 0; offset < Math.max(total, 1); offset += PAGE) {
+    const page2 = unzip((await get(`/records?${q}&offset=${offset}&limit=${PAGE}`)).records).recordMap ?? {};
+    Object.assign(records, page2);
+    log(`${Object.keys(records).length}/${total}`);
+    if (Object.keys(page2).length < PAGE) break;
+    await new Promise((r) => setTimeout(r, 800));
+  }
+  if (total && Object.keys(records).length < total * 0.9) throw new Error(`FAILED: 只拿到 ${Object.keys(records).length}/${total} 行`);
+  return { fieldMap: head.fieldMap, records };
+}
+__name(readBase, "readBase");
+
+// apps/daemon/src/data/readers/feishu-jobs.ts
+var HOST2 = /^https:\/\/([a-z0-9-]+)\.jobs\.feishu\.cn(?:\/([^/?#]*))?/i;
+var SOCIAL_PATH = /^(experienced|social|society|shezhao)/i;
+function feishuTenant(url) {
+  const m = HOST2.exec(url);
+  if (!m || m[2] && SOCIAL_PATH.test(m[2])) return null;
+  const path = m[2] && !/^(position|api)$/i.test(m[2]) ? m[2] : "index";
+  const project = (() => {
+    try {
+      return new URL(url).searchParams.get("project") ?? void 0;
+    } catch {
+      return void 0;
+    }
+  })();
+  return { tenant: m[1], path, project };
+}
+__name(feishuTenant, "feishuTenant");
+var clean2 = /* @__PURE__ */ __name((s) => (s ?? "").replace(/\r/g, "").replace(/[ \t]+\n/g, "\n").replace(/\n{3,}/g, "\n\n").trim(), "clean");
+function postToListing(p, t, cfg) {
+  const desc = [clean2(p.description) && `岗位职责
+${clean2(p.description)}`, clean2(p.requirement) && `任职要求
+${clean2(p.requirement)}`].filter(Boolean).join("\n\n");
+  const degree = /(博士|硕士及以上|硕士|本科及以上|本科|大专)/.exec(desc)?.[1];
+  const family = p.recruit_type?.parent?.name ?? p.recruit_type?.name ?? "";
+  const subject = p.job_subject?.name?.zh_cn ?? p.job_subject?.name?.i18n ?? "";
+  const cohortIn = /(\d{2,4})\s*届/.exec(subject)?.[1];
+  const cities = (p.city_list ?? []).map((c) => c.name ?? "").filter(Boolean);
+  return {
+    externalId: p.id,
+    url: `https://${t.tenant}.jobs.feishu.cn/${t.path}/position/${p.id}/detail/`,
+    title: clean2(p.title),
+    company: cfg.name || t.tenant,
+    orgUnit: p.job_function?.name || void 0,
+    locations: cities.length ? cities : p.city_info?.name ? [p.city_info.name] : [],
+    description: desc,
+    employmentType: p.recruit_type?.name || void 0,
+    datePosted: p.publish_time ? new Date(p.publish_time).toISOString() : void 0,
+    degreeReq: degree,
+    cohort: cfg.cohort ?? (cohortIn ? `${cohortIn.length === 2 ? "20" + cohortIn : cohortIn}届` : void 0),
+    batch: cfg.batch ?? (subject || (family.includes("校招") ? "校招" : "社招")),
+    kind: cfg.kind ?? (family.includes("校招") || /实习/.test(p.recruit_type?.name ?? "") ? degree === "博士" ? "博士专项" : "校招" : "社招"),
+    publisher: `feishu-jobs:${t.tenant}`,
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9,
+    raw: p
+  };
+}
+__name(postToListing, "postToListing");
+var UA_HEADERS = { "content-type": "application/json", "portal-channel": "official", "portal-platform": "pc", "accept-language": "zh-CN" };
+var feishuJobs = {
+  id: "feishu-jobs",
+  name: "飞书招聘",
+  input: "<公司>.jobs.feishu.cn（校招路径）",
+  mode: "pull",
+  describe: "飞书招聘门户（<租户>.jobs.feishu.cn）：走门户前端自己的搜索接口，零 token",
+  detect: /* @__PURE__ */ __name((url) => !!feishuTenant(url), "detect"),
+  // "/", "/index" and "/campus" of one tenant are one portal unless a 招聘项目 narrows it
+  siteKey: /* @__PURE__ */ __name((url) => {
+    const t = feishuTenant(url);
+    return `${t.tenant}${t.project ? "/" + t.project : ""}`;
+  }, "siteKey"),
+  async fetch(cfg, ctx) {
+    const t = feishuTenant(cfg.url ?? "");
+    if (!t) throw new Error(`BAD_URL: 不是飞书招聘的校招站：${cfg.url}`);
+    const base = `https://${t.tenant}.jobs.feishu.cn/api/v1`;
+    const csrf = await ctx.fetchJson(`${base}/csrf/token`, { method: "POST", headers: { ...UA_HEADERS, "website-path": t.path }, body: "{}" });
+    const token = csrf?.data?.token ?? "";
+    if (!token) throw new Error("SCHEMA_CHANGED: 飞书招聘：没拿到 csrf token");
+    const headers = { ...UA_HEADERS, "website-path": t.path, "x-csrf-token": token, cookie: `atsx-csrf-token=${encodeURIComponent(token)}` };
+    const out = [];
+    const seen = /* @__PURE__ */ new Set();
+    const limit = 100;
+    for (let page2 = 0; page2 < (cfg.maxPages ?? 10); page2++) {
+      const body = {
+        keyword: cfg.keywords?.[0] ?? "",
+        limit,
+        offset: page2 * limit,
+        job_category_id_list: [],
+        tag_id_list: [],
+        location_code_list: [],
+        subject_id_list: t.project ? [t.project] : [],
+        recruitment_id_list: [],
+        portal_type: 3,
+        job_function_id_list: [],
+        storefront_id_list: [],
+        portal_entrance: 1
+      };
+      const res = await ctx.fetchJson(`${base}/search/job/posts`, { method: "POST", headers, body: JSON.stringify(body) });
+      if (res?.code !== 0) throw new Error(`SCHEMA_CHANGED: 飞书招聘：${res?.message ?? "接口返回异常"}`);
+      const posts = res.data?.job_post_list ?? [];
+      for (const p of posts) {
+        if (seen.has(p.id)) continue;
+        seen.add(p.id);
+        const l = postToListing(p, t, cfg);
+        if (l.kind === "社招" && cfg.kind !== "社招") continue;
+        out.push(l);
+      }
+      ctx.log(`feishu-jobs ${t.tenant}/${t.path}: page ${page2 + 1} +${posts.length} (${out.length}/${res.data?.count ?? "?"})`);
+      if (posts.length < limit || out.length >= (res.data?.count ?? 0)) break;
+      await ctx.sleep(400);
+    }
+    return out;
+  }
+};
+
+// apps/daemon/src/data/types.ts
+var recentPulls = /* @__PURE__ */ new Map();
+async function pullOnce(key, pull) {
+  const hit = recentPulls.get(key);
+  if (hit && Date.now() - hit.at < 6 * 36e5) return hit.listings;
+  const listings = await pull();
+  recentPulls.set(key, { at: Date.now(), listings });
+  return listings;
+}
+__name(pullOnce, "pullOnce");
+
+// apps/daemon/src/data/readers/zhilian.ts
+var API2 = "https://fe.zhaopin.com/grace/api/dsc/search-job-list";
+var JOB = /(C[CZ]L?\d+J\d+)/;
+function zhilianUrl(url) {
+  let u;
+  try {
+    u = new URL(url);
+  } catch {
+    return null;
+  }
+  if (!/(^|\.)zhaopin\.com$/.test(u.hostname) || u.hostname === "fe.zhaopin.com") return null;
+  const job = JOB.exec(u.pathname)?.[1];
+  if (!job && /^(www\.|m\.)?zhaopin\.com$|^(xiaoyuan|m|webapp)\.zhaopin\.com$/.test(u.hostname) && !/\/company\/KA|\/\d{4}\//.test(u.pathname)) return null;
+  return { host: u.hostname, job };
+}
+__name(zhilianUrl, "zhilianUrl");
+var ID = /(?:xiaozhaoId|companyXzId|companyOutId|rootCompanyId|orgNumbers?)["']?[:=]\[?["']?(\d{5,7})\b|globalData\.companyId,[\w$]+="?(\d{5,7})\b/g;
+var idsIn = /* @__PURE__ */ __name((text3) => [...new Set([...text3.matchAll(ID)].map((m) => m[1] || m[2]))], "idsIn");
+async function count(ctx, id) {
+  const r = await ctx.fetchJson(API2, { method: "POST", headers: { "content-type": "application/json" }, body: JSON.stringify({ orgNumbers: [id], jobSource: 2, pageIndex: 1, pageSize: 1 }) }).catch(() => null);
+  return r?.data?.pageInfo?.totalNum ?? 0;
+}
+__name(count, "count");
+async function page(ctx, url) {
+  for (let i = 0; ; i++) {
+    const t = await ctx.fetchText(url);
+    if (!/EO_Bot_Ssid/.test(t) || t.length > 5e3 || i >= 3) return t;
+    await ctx.sleep(1500 * (i + 1));
+  }
+}
+__name(page, "page");
+var roots = /* @__PURE__ */ new Map();
+async function rootIdsOf(url, ctx) {
+  const z = zhilianUrl(url);
+  const key = z?.job && /^(xiaoyuan|m|www|webapp)\./.test(z.host) ? z.job : new URL(url).origin + (url.includes("webapp.zhaopin.com") ? new URL(url).pathname.split("/").slice(0, 4).join("/") : "");
+  if (roots.has(key)) return roots.get(key);
+  let ids = [];
+  if (z?.job) ids = [/"companyRootId":(\d+)/.exec(await page(ctx, `https://xiaoyuan.zhaopin.com/job/${z.job}`))?.[1] ?? ""].filter(Boolean);
+  else {
+    const u = new URL(url), base = u.origin + (u.host.startsWith("webapp.") ? u.pathname.split("/").slice(0, 4).join("/") : "") + "/";
+    const pages = [url], seen = new Set(pages), js = [];
+    let all = "";
+    for (let i = 0; i < pages.length && i < 8; i++) {
+      const html = await page(ctx, pages[i]).catch(() => "");
+      all += html;
+      for (const m of html.matchAll(/(?:src|href)="([^"]+\.js)"/g)) {
+        const s = new URL(m[1], pages[i]).href;
+        if (/\/assets\/|\/static\//.test(s) && !seen.has(s)) {
+          seen.add(s);
+          js.push(s);
+        }
+      }
+      for (let j = 0; j < js.length && j < 40; j++) {
+        const src = js[j];
+        if (src === "") continue;
+        js[j] = "";
+        const t = await ctx.fetchText(src).catch(() => "");
+        all += t;
+        for (const m of t.matchAll(/["'`](\.{1,2}\/[\w./-]+?\.js)(?=["'`])/g)) {
+          const s = new URL(m[1], src).href;
+          if (!seen.has(s)) {
+            seen.add(s);
+            js.push(s);
+          }
+        }
+      }
+      for (const m of all.matchAll(/nativeHref:"(\w+)"/g)) {
+        const p = base + (m[1] === "index" ? "" : `${m[1]}/index.html`);
+        if (!seen.has(p)) {
+          seen.add(p);
+          pages.push(p);
+        }
+      }
+    }
+    ids = idsIn(all);
+  }
+  const live = [];
+  for (const id of ids) if (await count(ctx, id) > 0) live.push(id);
+  if (!live.length) throw new Error(`ZERO_POSTS: 智联：没找到有岗位的校招机构号 ${url}`);
+  roots.set(key, live);
+  return live;
+}
+__name(rootIdsOf, "rootIdsOf");
+function itemToListing(it, root, cfg) {
+  const j = it.job ?? {}, c = it.company ?? {};
+  if (!j.jobNumber || !j.title) return null;
+  const detail = (j.detail ?? "").replace(/<br\s*\/?>/gi, "\n").replace(/<[^>]+>/g, "").replace(/&nbsp;/g, " ").replace(/\n{3,}/g, "\n\n").trim();
+  return {
+    externalId: j.jobNumber,
+    url: j.url || `https://xiaoyuan.zhaopin.com/job/${j.jobNumber}`,
+    title: j.title.trim(),
+    company: c.campusOrgName?.trim() || cfg.name || `智联 ${root}`,
+    orgUnit: c.campusDepartment?.name || void 0,
+    locations: j.cityName ? [j.cityName] : [],
+    description: detail,
+    datePosted: j.modifiedTime ? new Date(j.modifiedTime).toISOString() : void 0,
+    degreeReq: j.minEducationName || void 0,
+    majorReq: j.needMajor || void 0,
+    cohort: cfg.cohort ?? "2027届",
+    batch: cfg.batch ?? "校招",
+    kind: cfg.kind ?? (j.minEducationName === "博士" ? "博士专项" : "校招"),
+    publisher: `zhilian:${root}`,
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9
+  };
+}
+__name(itemToListing, "itemToListing");
+var zhilian = {
+  id: "zhilian",
+  name: "智联校园",
+  input: "xiaoyuan.zhaopin.com 岗位/公司页、集团专属站 <集团>.zhaopin.com",
+  mode: "pull",
+  describe: "智联校园（xiaoyuan.zhaopin.com、集团专属站 <集团>.zhaopin.com）：走站点前端的搜索接口，按集团校招机构号拉全部单位的岗位，零 token",
+  detect: /* @__PURE__ */ __name((url) => !!zhilianUrl(url), "detect"),
+  // ponytail: the root id needs a request, so two job links of one group are two subscriptions (same campaigns, a scan wasted)
+  siteKey: /* @__PURE__ */ __name((url) => {
+    const z = zhilianUrl(url);
+    return z.job && /^(xiaoyuan|m|www|webapp)\./.test(z.host) ? `job/${z.job}` : z.host;
+  }, "siteKey"),
+  async fetch(cfg, ctx) {
+    const ids = cfg.rootId ? [String(cfg.rootId)] : await rootIdsOf(cfg.url ?? "", ctx);
+    const root = ids.join(",");
+    return pullOnce(`zhilian:${root}:${cfg.keywords?.[0] ?? ""}`, async () => {
+      const out = [];
+      const seen = /* @__PURE__ */ new Set();
+      for (let page2 = 1; page2 <= (cfg.maxPages ?? 20); page2++) {
+        const r = await ctx.fetchJson(API2, { method: "POST", headers: { "content-type": "application/json" }, body: JSON.stringify({ orgNumbers: ids, jobSource: 2, pageIndex: page2, pageSize: 200, keyword: cfg.keywords?.[0] ?? "" }) });
+        if (r?.code !== 200) throw new Error(`SCHEMA_CHANGED: 智联：${r?.message ?? "接口返回异常"}`);
+        for (const it of r.data?.jobList ?? []) {
+          const l = itemToListing(it, root, cfg);
+          if (l && !seen.has(l.externalId)) {
+            seen.add(l.externalId);
+            out.push(l);
+          }
+        }
+        ctx.log(`zhilian ${root}: page ${page2} (${out.length}/${r.data?.pageInfo?.totalNum ?? "?"})`);
+        if (page2 >= (r.data?.pageInfo?.totalPage ?? 1)) break;
+        await ctx.sleep(400);
+      }
+      return out;
+    });
+  }
+};
+
+// apps/daemon/src/data/readers/iguopin.ts
+var G = "https://gp-api.iguopin.com";
+var H = { "content-type": "application/json", Device: "pc", Version: "5.0.0", Subsite: "iguopin" };
+var CAMPUS = "115xW5oQ";
+function iguopinUrl(url) {
+  let u;
+  try {
+    u = new URL(url);
+  } catch {
+    return null;
+  }
+  if (!/(^|\.)iguopin\.com$/.test(u.hostname)) return null;
+  const sub = u.hostname.split(".")[0], id = u.searchParams.get("id") ?? void 0;
+  if (sub !== "www" && sub !== "iguopin" && !/^(m|api\d*|gp-api)$/.test(sub)) return { tenant: sub, path: u.pathname, kind: "tenant" };
+  if (/^\/job\/detail/.test(u.pathname) && id) return { path: u.pathname, id, kind: "job" };
+  if (/^\/company/.test(u.pathname) && id) return { path: u.pathname, id, kind: "company" };
+  return null;
+}
+__name(iguopinUrl, "iguopinUrl");
+async function api(ctx, path, body) {
+  const r = await ctx.fetchJson(G + path, body ? { method: "POST", headers: H, body: JSON.stringify(body) } : { headers: H });
+  if (r?.code !== 200) throw new Error(`SCHEMA_CHANGED: 国聘 ${path}：${r?.code} ${r?.msg ?? ""}`);
+  return r.data;
+}
+__name(api, "api");
+function pickNav(nav, path) {
+  const jobs = nav.filter((n) => n?.props?.projectId || n?.props?.nature);
+  return jobs.find((n) => n.route === path) ?? jobs.find((n) => /校/.test(n.title ?? "")) ?? jobs[0];
+}
+__name(pickNav, "pickNav");
+async function queryOf(url, ctx) {
+  const t = iguopinUrl(url);
+  if (!t) throw new Error(`BAD_URL: 不是国聘链接：${url}`);
+  if (t.kind === "company") return { projectIds: [], companyId: t.id, nature: [CAMPUS] };
+  if (t.kind === "job") {
+    const j = await api(ctx, `/api/jobs/v1/info?id=${t.id}`);
+    return { projectIds: [], companyId: String(j.company_id), nature: [CAMPUS] };
+  }
+  const d = await api(ctx, `/api/activity/exclusive/v1/info?domain=${t.tenant}`);
+  const nav = pickNav(JSON.parse(d.content ?? "{}")?.params?.nav ?? [], t.path);
+  const projectIds = String(nav?.props?.projectId ?? "").split(",").filter(Boolean);
+  return { projectIds, companyId: String(d.company_id), nature: String(nav?.props?.nature ?? CAMPUS).split(",").filter(Boolean) };
+}
+__name(queryOf, "queryOf");
+function postToListing2(j, cfg) {
+  if (!j?.job_id || !j.job_name) return null;
+  const end = String(j.end_time ?? "").slice(0, 10);
+  return {
+    externalId: String(j.job_id),
+    url: `https://www.iguopin.com/job/detail?id=${j.job_id}`,
+    title: String(j.job_name).trim(),
+    company: String(j.company_name ?? "").trim() || cfg.name || "国聘",
+    orgUnit: j.department_cn || void 0,
+    industry: j.company_info?.industry_cn || void 0,
+    locations: [...new Set((j.district_list ?? []).map((d) => String(d.area_cn ?? "").split("-")[0]).filter(Boolean))],
+    description: String(j.contents ?? "").trim(),
+    datePosted: j.start_time ? String(j.start_time).replace(" ", "T") + "+08:00" : void 0,
+    validThrough: /^\d{4}-\d{2}-\d{2}$/.test(end) ? end : void 0,
+    deadlineText: j.end_time ? String(j.end_time) : void 0,
+    degreeReq: j.education_cn || void 0,
+    majorReq: Array.isArray(j.major_cn) ? j.major_cn.join("、") : void 0,
+    cohort: cfg.cohort ?? "2027届",
+    batch: cfg.batch ?? "校招",
+    kind: cfg.kind ?? (j.education_cn === "博士" ? "博士专项" : /实习/.test(j.nature_cn ?? "") ? "实习" : "校招"),
+    publisher: "iguopin",
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9
+  };
+}
+__name(postToListing2, "postToListing");
+async function discover(cfg, ctx) {
+  const out = /* @__PURE__ */ new Map();
+  for (const keyword of (cfg.keywords ?? []).slice(0, 8)) {
+    for (let page2 = 1; page2 <= (cfg.maxPages ?? 10); page2++) {
+      const d = await api(ctx, "/api/jobs/v1/list", { page: page2, page_size: 200, nature: [CAMPUS], keyword, source: "s_job_list" });
+      const list = d?.list ?? [];
+      for (const j of list) {
+        const l = postToListing2(j, { ...cfg, name: "" });
+        if (l && !out.has(l.externalId)) out.set(l.externalId, l);
+      }
+      ctx.log(`iguopin 发现「${keyword}」: page ${page2} +${list.length} (${out.size})`);
+      if (list.length < 200) break;
+      await ctx.sleep(400);
+    }
+  }
+  return [...out.values()];
+}
+__name(discover, "discover");
+var iguopin = {
+  id: "iguopin",
+  name: "国聘",
+  input: "<集团>.iguopin.com 专场、www.iguopin.com/company?id=、岗位详情页",
+  mode: "pull",
+  describe: "国聘（<集团>.iguopin.com 专场、www.iguopin.com/company?id=）：走站点前端接口，按招聘项目或公司树拉校招岗位（含截止），零 token",
+  detect: /* @__PURE__ */ __name((url) => !!iguopinUrl(url), "detect"),
+  siteKey: /* @__PURE__ */ __name((url) => {
+    const t = iguopinUrl(url);
+    return t.kind === "tenant" ? t.tenant : `${t.kind}/${t.id}`;
+  }, "siteKey"),
+  async fetch(cfg, ctx) {
+    if (cfg.discover) return discover(cfg, ctx);
+    const q = await queryOf(cfg.url ?? "", ctx);
+    return pullOnce(`iguopin:${JSON.stringify(q)}`, async () => {
+      const out = [];
+      const seen = /* @__PURE__ */ new Set();
+      for (const mode of q.projectIds.length ? ["project", "company"] : ["company"]) {
+        for (let page2 = 1; page2 <= (cfg.maxPages ?? 20); page2++) {
+          const d = mode === "project" ? await api(ctx, "/api/jobs/v1/project-job", { page: page2, page_size: 200, project_id: q.projectIds }) : await api(ctx, "/api/jobs/v1/list", { page: page2, page_size: 200, company_id_with_sub: q.companyId, nature: q.nature, source: "s_job_list" });
+          const list = d?.list ?? [];
+          for (const j of list) {
+            const l = postToListing2(j, cfg);
+            if (l && !seen.has(l.externalId)) {
+              seen.add(l.externalId);
+              out.push(l);
+            }
+          }
+          ctx.log(`iguopin ${mode}: page ${page2} +${list.length} (${out.length})`);
+          if (list.length < 200) break;
+          await ctx.sleep(400);
+        }
+        if (out.length) break;
+      }
+      return out;
+    });
+  }
+};
+
+// apps/daemon/src/data/readers/chnenergy.ts
+var BASE = "https://zhaopin.chnenergy.com.cn";
+var FORM = { "content-type": "application/x-www-form-urlencoded" };
+var CAMPUS2 = /毕业生|校园|校招|管培|专项|博士后|应届/;
+var strip = /* @__PURE__ */ __name((s) => s.replace(/<!--[\s\S]*?-->/g, "").replace(/<[^>]+>/g, "").replace(/&nbsp;/g, " ").replace(/\s+/g, " ").trim(), "strip");
+function parseAnnouncements(html) {
+  return [...html.matchAll(/showgg\?id=([\w-]+)"[^>]*>([\s\S]*?)<\/a>[\s\S]*?(\d{4})年(\d{2})月(\d{2})日/g)].map((m) => ({ id: m[1], title: strip(m[2]), date: `${m[3]}-${m[4]}-${m[5]}` }));
+}
+__name(parseAnnouncements, "parseAnnouncements");
+function parseStations(html) {
+  const pages = Number(/共<b>\d+<\/b>条，共<b>(\d+)<\/b>页/.exec(html)?.[1] ?? 0);
+  const rows = [...html.matchAll(/showgw\?id=([\w-]+)"[^>]*title="([^"]*)"[\s\S]*?<small>([^<]*)<\/small>[\s\S]*?<p class="list-group-item-text">([\s\S]*?)<\/p>/g)].map((m) => {
+    const [unit, major, degree, city, headcount] = strip(m[4].replace(/<span title="([^"]*)">[^<]*<\/span>/g, "$1")).split(/\s*\|\s*/);
+    return { id: m[1], title: strip(m[2]), category: strip(m[3]), unit: unit ?? "", major: major ?? "", degree: degree ?? "", city: city ?? "", headcount: headcount ?? "" };
+  });
+  return { pages, rows };
+}
+__name(parseStations, "parseStations");
+async function retry(ctx, f) {
+  for (let i = 0; ; i++) {
+    try {
+      return await f();
+    } catch (e) {
+      if (i >= 3) throw e;
+      await ctx.sleep(2e3 * (i + 1));
+    }
+  }
+}
+__name(retry, "retry");
+async function post(ctx, path, body) {
+  return retry(ctx, () => ctx.fetchText(BASE + path, { method: "POST", headers: FORM, body: new URLSearchParams(body).toString() }));
+}
+__name(post, "post");
+var chnenergy = {
+  id: "chnenergy",
+  name: "国家能源集团招聘网",
+  input: "zhaopin.chnenergy.com.cn 任一页面（整站校招公告）",
+  notices: true,
+  mode: "pull",
+  describe: "国家能源集团招聘网（zhaopin.chnenergy.com.cn）：整站当前校招公告（统招、管培生、专项、博士后）的全部岗位，零 token",
+  detect: /* @__PURE__ */ __name((url) => /^https?:\/\/zhaopin\.chnenergy\.com\.cn/i.test(url), "detect"),
+  siteKey: /* @__PURE__ */ __name(() => "zhaopin.chnenergy.com.cn", "siteKey"),
+  async fetch(cfg, ctx) {
+    const anns = [];
+    for (let p = 0; p < 5; p++) {
+      const got = parseAnnouncements(await post(ctx, "/annc/annclist", { pagenum: String(p), ggtype: "1" }));
+      anns.push(...got.filter((a) => !anns.some((x) => x.id === a.id)));
+      if (got.length < 10) break;
+      await ctx.sleep(300);
+    }
+    const year = Number((cfg.cohort ?? "2027").replace(/\D/g, "").slice(0, 4)) || 2027;
+    const campus = anns.filter((a) => CAMPUS2.test(a.title) && !/社会招聘/.test(a.title) && (a.title.includes(String(year)) || a.date >= `${year - 1}-07-01`));
+    const out = [];
+    for (const a of campus) {
+      let deadline;
+      for (let page2 = 0, pages = 1; page2 < pages && page2 < (cfg.maxPages ?? 400); page2++) {
+        const r = parseStations(await post(ctx, "/annc/showggStationList", { id: a.id, pagenum: String(page2) }));
+        pages = r.pages;
+        if (!deadline && r.rows[0]) {
+          const d = /截止时间：<strong>(\d{4})年(\d{2})月(\d{2})日/.exec(await retry(ctx, () => ctx.fetchText(`${BASE}/annc/showgw?id=${r.rows[0].id}`)));
+          deadline = d ? `${d[1]}-${d[2]}-${d[3]}` : "";
+          if (deadline && deadline < a.date) {
+            ctx.log(`chnenergy ${a.title.slice(0, 24)}: 岗位列表截止 ${deadline} 早于公告发布 ${a.date}，判为旧列表，跳过`);
+            break;
+          }
+        }
+        for (const s of r.rows) out.push({
+          // one title is hired by dozens of plants (「电气检修」): the pool keys positions by title, so the unit goes in it
+          externalId: s.id,
+          url: `${BASE}/annc/showgw?id=${s.id}`,
+          noticeUrl: `${BASE}/annc/showgg?id=${a.id}`,
+          title: s.unit ? `${s.title}（${s.unit}）` : s.title,
+          company: "国家能源集团",
+          orgUnit: s.unit || void 0,
+          locations: s.city ? s.city.split(/[,，、]/).filter(Boolean) : [],
+          description: `${s.category}岗；${s.unit}；专业：${s.major}；${s.degree}；${s.headcount}`,
+          majorReq: s.major,
+          degreeReq: s.degree,
+          validThrough: deadline || void 0,
+          cohort: cfg.cohort ?? `${year}届`,
+          batch: a.title.replace(/^国家能源投资集团有限责任公司|^国家能源集团/, "").slice(0, 40),
+          kind: /博士后/.test(a.title) ? "博士后" : /管培/.test(a.title) ? "管培生" : s.degree.startsWith("博士") ? "博士专项" : "校招",
+          publisher: "chnenergy",
+          grade: "官方",
+          isOriginal: true,
+          confidence: 0.9
+        });
+        ctx.log(`chnenergy ${a.title.slice(0, 20)}: page ${page2 + 1}/${pages} (${out.length})`);
+        await ctx.sleep(200);
+      }
+    }
+    return out;
+  }
+};
+
+// apps/daemon/src/data/readers/chng.ts
+import { createCipheriv } from "node:crypto";
+var BASE2 = "https://zhaopin.chng.com.cn";
+var EDU = { "3": "专科", "4": "本科", "5": "硕士研究生", "6": "博士研究生" };
+var ROUTE = ["CampusRecruit", "InternRecruit", "ManagementTraineeRecruit"];
+async function keys(ctx) {
+  const html = await ctx.fetchText(BASE2 + "/");
+  for (const m of html.matchAll(/src="([^"]+\.js)"/g)) {
+    const js = await ctx.fetchText(new URL(m[1], BASE2 + "/").toString());
+    const s = [...js.matchAll(/Utf8\.parse\("([^"]+)"\)/g)].map((x) => x[1]);
+    const key = s.find((x) => x.length === 32), iv = s.find((x) => x.length === 16);
+    if (key && iv) return { key: Buffer.from(key), iv: Buffer.from(iv) };
+  }
+  throw new Error("SCHEMA_CHANGED: 华能：页面脚本里没找到接口密钥");
+}
+__name(keys, "keys");
+var text = /* @__PURE__ */ __name((h) => (h ?? "").replace(/<\/p>|<br\s*\/?>/gi, "\n").replace(/<[^>]+>/g, "").replace(/&nbsp;/g, " ").replace(/\n{3,}/g, "\n\n").trim(), "text");
+function positionToListing(p, a, cohort) {
+  if (!p?.id || !p.positionName) return null;
+  const end = String(p.endDate ?? a.endTime ?? "").slice(0, 10);
+  return {
+    externalId: String(p.id),
+    url: `${BASE2}/${ROUTE[a.taskType] ?? "CampusRecruit"}?positionId=${p.id}`,
+    noticeUrl: `${BASE2}/${ROUTE[a.taskType] ?? "CampusRecruit"}?aId=${a.id}`,
+    title: String(p.positionName).trim(),
+    company: String(p.recruitCompanyName || a.recruitCompanyName || "中国华能").trim(),
+    locations: p.address ? String(p.address).split(/[,，、\s]+/).filter(Boolean) : [],
+    description: text(p.jobReq),
+    validThrough: /^\d{4}-\d{2}-\d{2}$/.test(end) ? end : void 0,
+    deadlineText: p.endDate ?? void 0,
+    degreeReq: EDU[String(p.eduType)] ?? void 0,
+    majorReq: (p.majors ?? []).map((m) => m.majorName).filter(Boolean).join("、") || void 0,
+    cohort,
+    batch: a.taskType === 2 ? "管培生" : a.taskType === 1 ? "实习" : "校招",
+    kind: a.taskType === 1 ? "实习" : String(p.eduType) === "6" ? "博士专项" : "校招",
+    publisher: "chng",
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9
+  };
+}
+__name(positionToListing, "positionToListing");
+var chng = {
+  id: "chng",
+  name: "华能招聘网",
+  input: "zhaopin.chng.com.cn 任一页面（整站校招/管培公告）",
+  notices: true,
+  mode: "pull",
+  describe: "中国华能招聘网（zhaopin.chng.com.cn）：整站当前校招/管培/实习公告的全部岗位（含专业与截止），零 token",
+  detect: /* @__PURE__ */ __name((url) => /^https?:\/\/zhaopin\.chng\.com\.cn/i.test(url), "detect"),
+  siteKey: /* @__PURE__ */ __name(() => "zhaopin.chng.com.cn", "siteKey"),
+  async fetch(cfg, ctx) {
+    const { key, iv } = await keys(ctx);
+    const enc = /* @__PURE__ */ __name((s) => {
+      const c = createCipheriv("aes-256-cbc", key, iv);
+      return c.update(s, "utf8", "hex") + c.final("hex");
+    }, "enc");
+    const get = /* @__PURE__ */ __name(async (rest, params) => {
+      const r = await ctx.fetchJson(`${BASE2}/app-api/recruit/${enc(`${rest}?${new URLSearchParams(Object.entries(params).map(([k, v]) => [k, String(v)])).toString()}`)}`, { headers: { referer: BASE2 + "/" } });
+      if (r?.code !== 0) throw new Error(`SCHEMA_CHANGED: 华能 ${rest}：${r?.msg ?? "接口返回异常"}`);
+      return r.data;
+    }, "get");
+    const year = (cfg.cohort ?? "2027").replace(/\D/g, "").slice(0, 4) || "2027";
+    const anns = [];
+    for (let p = 1; p <= 10; p++) {
+      const d = await get("announcement/index", { pageNo: p, pageSize: 100, recruitType: 1 });
+      anns.push(...d?.list ?? []);
+      if (!d?.list?.length || anns.length >= (d.total ?? 0)) break;
+      await ctx.sleep(300);
+    }
+    const out = [];
+    for (const a of anns.filter((x) => !x.recruitYear || String(x.recruitYear) >= year)) {
+      const full = await get("announcement/select", { aId: a.id });
+      for (const p of full?.positions ?? []) {
+        const l = positionToListing(p, { ...a, ...full }, cfg.cohort ?? `${year}届`);
+        if (l) out.push(l);
+      }
+      ctx.log(`chng ${String(a.title ?? "").slice(0, 20)}: ${out.length}`);
+      await ctx.sleep(300);
+    }
+    return out;
+  }
+};
+
+// apps/daemon/src/data/readers/dayee.ts
+function dayeeSuite(url) {
+  const m = /^https?:\/\/[^/]+\/(SU[0-9a-f]{24})(?:[/?#]|$)/i.exec(url);
+  if (!m) return null;
+  let q;
+  try {
+    q = new URL(url).searchParams;
+  } catch {
+    q = new URLSearchParams();
+  }
+  return { su: m[1], recruitType: /interns?\.html/.test(url) ? "12" : q.get("recruitType") ?? "1", postId: q.get("postId") ?? void 0 };
+}
+__name(dayeeSuite, "dayeeSuite");
+var named = /* @__PURE__ */ __name((x) => x && !genericCompany(String(x)) ? String(x) : "", "named");
+var details = /* @__PURE__ */ new Map();
+function rowToListing(p, d, su, cfg, today = (/* @__PURE__ */ new Date()).toISOString().slice(0, 10)) {
+  if (!p?.postId || !p.postName) return null;
+  const end = String(p.endDate ?? "").slice(0, 10);
+  if (/^\d{4}-\d{2}-\d{2}$/.test(end) && end < today) return null;
+  const year = Number((cfg.cohort ?? "2027").replace(/\D/g, "").slice(0, 4)) || 2027;
+  const projYear = Number(/(20\d\d)/.exec(p.projectName ?? "")?.[1] ?? 0);
+  if (projYear && projYear < year) return null;
+  const ownDomain = !!cfg.url && !/\/\/wecruit\.hotjob\.cn\//i.test(cfg.url);
+  const jd = [d?.workContent && `岗位职责
+${String(d.workContent).trim()}`, d?.serviceCondition && `任职要求
+${String(d.serviceCondition).trim()}`].filter(Boolean).join("\n\n");
+  const places = (d?.workPlaceList ?? []).map((w) => w.name).concat(p.workPlaceStr ? String(p.workPlaceStr).split(/[,，;；]/) : []);
+  return {
+    externalId: String(p.postId),
+    url: `https://wecruit.hotjob.cn/${su}/pb/posDetail.html?postId=${p.postId}&postType=campus`,
+    title: String(p.postName).trim(),
+    // on the shared host a group lists its member companies in `company` (中国电科's 研究所); a tenant on its own domain is
+    // one company and the column holds its departments (荣耀: 「Marketing与销售服务部」)
+    // … and some tenants fill it with the literal word 「公司」 (three sites did): that is no name, the subscription's is
+    company: String(ownDomain && cfg.name || named(p.company) || named(d?.orgName) || cfg.name || "").trim() || "大易招聘",
+    orgUnit: (ownDomain && cfg.name ? p.company : p.department) || void 0,
+    locations: [...new Set(places.map((x) => x.split("-")[0].trim()).filter(Boolean))],
+    description: jd || void 0,
+    validThrough: /^\d{4}-\d{2}-\d{2}$/.test(end) ? end : void 0,
+    datePosted: p.publishDate ? String(p.publishDate).slice(0, 10) : void 0,
+    degreeReq: p.educationStr || d?.education || void 0,
+    majorReq: p.subject || void 0,
+    cohort: cfg.cohort ?? `${year}届`,
+    batch: cfg.batch ?? (p.projectName || "校招"),
+    kind: cfg.kind ?? (/实习/.test(p.projectName ?? "") ? "实习" : "校招"),
+    publisher: `dayee:${su}`,
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9
+  };
+}
+__name(rowToListing, "rowToListing");
+var dayee = {
+  id: "dayee",
+  name: "大易",
+  input: "wecruit.hotjob.cn/SU…（学校招聘、岗位详情页；公司自有域名上的同样路径也认）",
+  mode: "pull",
+  describe: "大易新版校招站（wecruit.hotjob.cn/SU…）：走页面自己的接口，列表 + 每个岗位的职责与要求，零 token",
+  detect: /* @__PURE__ */ __name((url) => !!dayeeSuite(url), "detect"),
+  siteKey: /* @__PURE__ */ __name((url) => dayeeSuite(url).su, "siteKey"),
+  async fetch(cfg, ctx) {
+    const t = dayeeSuite(cfg.url ?? "");
+    if (!t) throw new Error(`BAD_URL: 不是大易新版校招站：${cfg.url}`);
+    const call = /* @__PURE__ */ __name(async (op, form) => {
+      const r = await ctx.fetchJson(`https://wecruit.hotjob.cn/wecruit/positionInfo/${op}/${t.su}?iSaJAx=isAjax&request_locale=zh_CN&t=${Date.now()}`, {
+        method: "POST",
+        headers: { "content-type": "application/x-www-form-urlencoded" },
+        body: new URLSearchParams(Object.entries(form).map(([k, v]) => [k, String(v)])).toString()
+      });
+      if (String(r?.state) !== "200") throw new Error(`SCHEMA_CHANGED: 大易 ${op}：${r?.Exception?.message ?? JSON.stringify(r).slice(0, 120)}`);
+      await ctx.sleep(250);
+      return r.data;
+    }, "call");
+    return pullOnce(`dayee:${t.su}:${t.recruitType}`, async () => {
+      const base = { isFrompb: "true", recruitType: t.recruitType };
+      const rows = /* @__PURE__ */ new Map();
+      const add = /* @__PURE__ */ __name((d) => {
+        for (const p of d?.pageForm?.pageData ?? []) rows.set(p.postId, p);
+      }, "add");
+      const first = await call("listPosition", { ...base, pageSize: 10, currentPage: 1 });
+      add(first);
+      const total = Number(first?.pageForm?.dataCount ?? first?.positonNum ?? 0);
+      for (let size = 1; size < 50 && size < total; size *= 2) add(await call("listPosition", { ...base, pageSize: size, currentPage: 2 }));
+      for (let page2 = 2; (page2 - 1) * 50 < total && page2 <= (cfg.maxPages ?? 40); page2++) add(await call("listPosition", { ...base, pageSize: 50, currentPage: page2 }));
+      ctx.log(`dayee ${t.su}: ${rows.size}/${total} rows`);
+      const out = [];
+      for (const p of rows.values()) {
+        if (!rowToListing(p, null, t.su, cfg)) continue;
+        if (!details.has(p.postId)) details.set(p.postId, await call("listPositionDetail", { postId: p.postId, recruitType: t.recruitType }).catch(() => null));
+        const l = rowToListing(p, details.get(p.postId), t.su, cfg);
+        if (l) out.push(l);
+      }
+      return out;
+    });
+  }
+};
+
+// apps/daemon/src/data/readers/job51.ts
+import { createHash as createHash2 } from "node:crypto";
+var GUID = /[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}/i;
+var md5 = /* @__PURE__ */ __name((s) => createHash2("md5").update(s, "utf8").digest("hex"), "md5");
+function job51Url(url) {
+  let u;
+  try {
+    u = new URL(url);
+  } catch {
+    return false;
+  }
+  return /^(campus|ideal|xym|xyz)\.51job\.com$/.test(u.hostname);
+}
+__name(job51Url, "job51Url");
+function literalToJson(s) {
+  let o = "", i = 0;
+  while (i < s.length) {
+    const c = s[i];
+    if (c === "'" || c === '"') {
+      let v = "";
+      i++;
+      while (i < s.length && s[i] !== c) {
+        if (s[i] === "\\") {
+          const n = s[i + 1];
+          v += { n: "\n", r: "\r", t: "	" }[n] ?? n;
+          i += 2;
+        } else v += s[i++];
+      }
+      i++;
+      o += JSON.stringify(v);
+    } else if (/[A-Za-z_$]/.test(c)) {
+      let w = "";
+      while (i < s.length && /[\w$]/.test(s[i])) w += s[i++];
+      o += /^(true|false|null)$/.test(w) ? w : JSON.stringify(w);
+    } else {
+      o += c;
+      i++;
+    }
+  }
+  return o;
+}
+__name(literalToJson, "literalToJson");
+function siteLinks(text3, base) {
+  const root = new URL(".", base).href, out = [];
+  for (const m of text3.matchAll(/["']((?:\.\/)?[\w./-]+\.(?:html?|js))(?:\?[^"']*)?["']/g)) {
+    const p = new URL(m[1], base).href;
+    if (p.startsWith(root) && !out.includes(p) && !/jquery|swiper|jweixin|\.min\.js/.test(p)) out.push(p);
+  }
+  return out;
+}
+__name(siteLinks, "siteLinks");
+async function guidOf(url, ctx) {
+  const direct = /[?&]id=([0-9a-f-]{36})/i.exec(url) ?? /CtmID=([0-9a-f-]{36})/i.exec(url);
+  if (direct) return direct[1];
+  let ctmid = /ctmid=(\d+)/i.exec(url)?.[1];
+  if (!ctmid) {
+    const base = /\/$|\.html?$/.test(url) ? url : url + "/";
+    const pages = [base, new URL("js/job.js", base).href, new URL("js/jobs.js", base).href];
+    for (let i = 0; i < pages.length && i < 16 && !ctmid; i++) {
+      const t = await ctx.fetchText(pages[i]).catch(() => "");
+      ctmid = /ctmid["']?\s*[:=]\s*["']?(\d{5,})/i.exec(t)?.[1];
+      for (const p of siteLinks(t, base)) if (!pages.includes(p)) p.endsWith(".js") ? pages.splice(i + 1, 0, p) : pages.push(p);
+    }
+  }
+  if (!ctmid) throw new Error(`SCHEMA_CHANGED: 51job：页面里没找到企业编号 ${url}`);
+  const r = await fetch(`https://xyz.51job.com/External/Apply.aspx?CtmID=${ctmid}`, { redirect: "manual", headers: { "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) Chrome/128.0 Safari/537.36" }, signal: AbortSignal.timeout(2e4) }).catch(() => null);
+  const g = GUID.exec(r?.headers.get("location") ?? "")?.[0] ?? (await directory(ctx)).get(ctmid);
+  if (!g) throw new Error(`SCHEMA_CHANGED: 51job：企业编号 ${ctmid} 没有换到招聘号`);
+  return g;
+}
+__name(guidOf, "guidOf");
+var dir = null;
+async function directory(ctx) {
+  if (dir && Date.now() - dir.at < 864e5) return dir.map;
+  const t = await ctx.fetchText("https://xym.51job.com/jobs/jobs_ctmlist.aspx").catch(() => "");
+  const map = new Map([...t.matchAll(/\{ctmid:'([0-9a-f-]{36})',mname:'[^']*',ehireid:'(\d+)'/g)].map((m) => [m[2], m[1]]));
+  dir = { at: Date.now(), map };
+  return map;
+}
+__name(directory, "directory");
+var details2 = /* @__PURE__ */ new Map();
+function jobToListing(j, d, cfg, employer) {
+  if (!j?.JobID || !j.CjobName) return null;
+  const units = String(d?.OrgName ?? "").split("/").filter(Boolean);
+  const jd = String(d?.CJobInfo ?? "").replace(/<br\s*\/?>|<\/p>/gi, "\n").replace(/<[^>]+>/g, "").replace(/&nbsp;/g, " ").replace(/\n{3,}/g, "\n\n").trim();
+  return {
+    externalId: String(j.JobID),
+    url: d?.Url ? `https:${String(d.Url).replace(/^https?:/, "")}` : `https://xym.51job.com/wechat/vuectmjobs/#/index?id=${j.JobID}`,
+    title: String(j.CjobName).trim(),
+    company: units[1] || units[0] || employer || cfg.name || "51job",
+    orgUnit: units.slice(2).join("/") || void 0,
+    locations: j.JobAreaname ? [String(j.JobAreaname)] : [],
+    description: jd || void 0,
+    degreeReq: j.DegreeFrom || d?.DegreeName || void 0,
+    cohort: cfg.cohort ?? "2027届",
+    batch: cfg.batch ?? "校招",
+    kind: cfg.kind ?? (/实习/.test(String(j.Term ?? "")) ? "实习" : "校招"),
+    publisher: "job51",
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.85
+  };
+}
+__name(jobToListing, "jobToListing");
+var job51 = {
+  id: "job51",
+  name: "前程无忧",
+  input: "campus / ideal / xym / xyz.51job.com（无截止日期）",
+  mode: "pull",
+  describe: "前程无忧校招页（campus / ideal / xym / xyz.51job.com）：按雇主号走微信职位页的接口，列表 + 岗位详情，零 token；无截止日期",
+  detect: job51Url,
+  siteKey: /* @__PURE__ */ __name((url) => GUID.exec(url)?.[0] ?? /ctmid=(\d+)/i.exec(url)?.[1] ?? new URL(url).host + new URL(url).pathname.split("/").slice(0, 2).join("/"), "siteKey"),
+  async fetch(cfg, ctx) {
+    const id = await guidOf(cfg.url ?? "", ctx);
+    const token = md5(md5(id));
+    const api2 = /* @__PURE__ */ __name(async (func, body) => {
+      const t = await ctx.fetchText(`https://xym.51job.com/VueData/Jobs.ashx?func=${func}`, { method: "POST", headers: { "content-type": "application/x-www-form-urlencoded" }, body: new URLSearchParams(Object.entries(body).map(([k, v]) => [k, String(v)])).toString() });
+      const r = JSON.parse(literalToJson(t));
+      if (r?.status !== "ok" || typeof r.data !== "object") throw new Error(`SCHEMA_CHANGED: 51job ${func}：${t.slice(0, 80)}`);
+      await ctx.sleep(200);
+      return r.data;
+    }, "api");
+    return pullOnce(`job51:${id}`, async () => {
+      const info = await api2("getctminfo", { CtmID: id, Token: token }).catch(() => ({}));
+      const jobs = [];
+      for (let p = 1; p <= (cfg.maxPages ?? 200); p++) {
+        const d = await api2("getjobslistbyarea", { CtmID: id, PageIndex: p, areaname: "", keyword: "", funtypecode: "", token });
+        jobs.push(...d.records ?? []);
+        if (p >= Number(d.pagecount ?? 1)) break;
+      }
+      ctx.log(`job51 ${info?.ShowName ?? id}: ${jobs.length} jobs`);
+      const out = [];
+      for (const j of jobs) {
+        if (!details2.has(j.JobID)) details2.set(j.JobID, await api2("getJobInfo", { CtmID: id, JobID: j.JobID, Token: token }).catch(() => null));
+        const l = jobToListing(j, details2.get(j.JobID), cfg, String(info?.ShowName ?? "").replace(/20\d\d.*$/, ""));
+        if (l) out.push(l);
+      }
+      return out;
+    });
+  }
+};
+
+// apps/daemon/src/data/readers/own-sites.ts
+var clean3 = /* @__PURE__ */ __name((s) => (s ?? "").replace(/\r/g, "").replace(/\n{3,}/g, "\n\n").trim(), "clean");
+var degreeIn = /* @__PURE__ */ __name((s) => /(博士|硕士及以上|硕士|本科及以上|本科|大专)/.exec(s)?.[1], "degreeIn");
+var JSON_POST = { method: "POST", headers: { "content-type": "application/json", accept: "application/json" } };
+var KS = "https://campus.kuaishou.cn/recruit/campus/e/api/v1";
+function ksListing(p, projects, cfg) {
+  const desc = [clean3(p.description) && `岗位职责
+${clean3(p.description)}`, clean3(p.positionDemand) && `任职要求
+${clean3(p.positionDemand)}`].filter(Boolean).join("\n\n");
+  const batch = projects.get(p.recruitSubProjectCode ?? "") ?? "校招", intern = p.positionNatureCode === "intern" || /实习/.test(batch);
+  return {
+    externalId: String(p.id),
+    url: `https://campus.kuaishou.cn/#/campus/job-info/${p.id}`,
+    title: clean3(p.name),
+    company: cfg.name || "快手",
+    locations: (p.workLocationDicts ?? []).map((w) => w.name).filter(Boolean),
+    description: desc,
+    degreeReq: degreeIn(desc),
+    datePosted: p.releaseTime ? (/* @__PURE__ */ new Date(p.releaseTime.replace(" ", "T") + "+08:00")).toISOString() : void 0,
+    cohort: cfg.cohort ?? (/(\d{4})/.exec(batch)?.[1] ? `${/(\d{4})/.exec(batch)[1]}届` : void 0),
+    batch,
+    kind: cfg.kind ?? (intern ? "实习" : "校招"),
+    publisher: "kuaishou",
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9,
+    raw: p
+  };
+}
+__name(ksListing, "ksListing");
+var kuaishou = {
+  id: "kuaishou",
+  name: "快手校招",
+  input: "campus.kuaishou.cn",
+  mode: "pull",
+  describe: "快手校招官网（campus.kuaishou.cn）：当届应届生与实习生项目的全部岗位",
+  detect: /* @__PURE__ */ __name((url) => /^https:\/\/campus\.kuaishou\.cn(\/|$)/i.test(url), "detect"),
+  siteKey: /* @__PURE__ */ __name(() => "campus.kuaishou.cn", "siteKey"),
+  async fetch(cfg, ctx) {
+    const dict = await ctx.fetchJson(`${KS}/dictionary/batch?types=recruitSubProject`);
+    const all = dict?.result?.recruitSubProject ?? [];
+    if (dict?.code !== 0 || !all.length) throw new Error("SCHEMA_CHANGED: 快手校招没返回招聘项目列表");
+    const year = /(\d{4})/.exec(cfg.cohort ?? "")?.[1] ?? String((/* @__PURE__ */ new Date()).getFullYear() + ((/* @__PURE__ */ new Date()).getMonth() >= 6 ? 1 : 0));
+    const live = all.filter((x) => x.name.includes(year));
+    if (!live.length) throw new Error(`ZERO_POSTS: 快手校招还没有 ${year} 届的招聘项目`);
+    const projects = new Map(all.map((x) => [x.code, x.name]));
+    const out = [];
+    for (let page2 = 1; page2 <= (cfg.maxPages ?? 20); page2++) {
+      const res = await ctx.fetchJson(`${KS}/open/positions/simple`, { ...JSON_POST, body: JSON.stringify({ recruitSubProjectCodes: live.map((x) => x.code), pageSize: 100, pageNum: page2 }) });
+      if (res?.code !== 0 || !Array.isArray(res.result?.list)) throw new Error(`SCHEMA_CHANGED: 快手校招岗位接口返回异常（${res?.message ?? "无 list"}）`);
+      out.push(...res.result.list.map((p) => ksListing(p, projects, cfg)));
+      ctx.log(`kuaishou: page ${page2} (${out.length}/${res.result.total})`);
+      if (res.result.list.length < 100 || out.length >= res.result.total) break;
+      await ctx.sleep(400);
+    }
+    return out;
+  }
+};
+var XHS = "https://job.xiaohongshu.com/websiterecruit/position/pageQueryPosition";
+function xhsListing(p, cfg) {
+  const desc = [clean3(p.duty) && `岗位职责
+${clean3(p.duty)}`, clean3(p.qualification) && `任职要求
+${clean3(p.qualification)}`].filter(Boolean).join("\n\n");
+  const batch = p.jobProjectName && p.jobProjectName !== "其他项目" ? p.jobProjectName : /实习/.test(p.positionName) ? "实习" : "校招";
+  return {
+    externalId: String(p.positionId),
+    url: `https://job.xiaohongshu.com/campus/position/${p.positionId}`,
+    title: clean3(p.positionName),
+    company: cfg.name || "小红书",
+    orgUnit: p.jobType || void 0,
+    locations: (p.workplace ?? "").split(/[，,、]/).map((s) => s.trim().replace(/市$/, "")).filter(Boolean),
+    description: desc,
+    degreeReq: degreeIn(desc),
+    datePosted: p.publishTime ? (/* @__PURE__ */ new Date(p.publishTime + "T00:00:00+08:00")).toISOString() : void 0,
+    cohort: cfg.cohort ?? (/(\d{4})/.exec(batch)?.[1] ? `${/(\d{4})/.exec(batch)[1]}届` : void 0),
+    batch,
+    kind: cfg.kind ?? (/实习/.test(p.positionName + batch) ? "实习" : "校招"),
+    publisher: "xiaohongshu",
+    grade: "官方",
+    isOriginal: true,
+    confidence: 0.9,
+    raw: p
+  };
+}
+__name(xhsListing, "xhsListing");
+var xiaohongshu = {
+  id: "xiaohongshu",
+  name: "小红书校招",
+  input: "job.xiaohongshu.com/campus",
+  mode: "pull",
+  describe: "小红书校招官网（job.xiaohongshu.com/campus）：校园招聘、REDstar 与实习的全部岗位",
+  detect: /* @__PURE__ */ __name((url) => /^https:\/\/job\.xiaohongshu\.com\/campus(\/|$|\?)/i.test(url), "detect"),
+  siteKey: /* @__PURE__ */ __name(() => "job.xiaohongshu.com/campus", "siteKey"),
+  async fetch(cfg, ctx) {
+    const out = [];
+    for (let page2 = 1; page2 <= (cfg.maxPages ?? 20); page2++) {
+      const res = await ctx.fetchJson(XHS, { method: "POST", headers: { ...JSON_POST.headers, origin: "https://job.xiaohongshu.com", referer: "https://job.xiaohongshu.com/campus/position" }, body: JSON.stringify({ positionName: "", pageNum: page2, pageSize: 50, recruitType: "campus" }) });
+      if (!Array.isArray(res?.data?.list)) throw new Error(`SCHEMA_CHANGED: 小红书校招岗位接口返回异常（${res?.alertMsg ?? "无 list"}）`);
+      out.push(...res.data.list.filter((p) => !p.recruitStatus || p.recruitStatus === "in_recruitment").map((p) => xhsListing(p, cfg)));
+      ctx.log(`xiaohongshu: page ${page2}/${res.data.totalPage} (${out.length}/${res.data.total})`);
+      if (page2 >= (res.data.totalPage ?? 1) || !res.data.list.length) break;
+      await ctx.sleep(400);
+    }
+    return out;
+  }
+};
+
+// apps/daemon/src/data/readers/csv.ts
+import { existsSync, readFileSync } from "node:fs";
+import { resolve } from "node:path";
+var UA2 = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/140.0.0.0 Safari/537.36";
+var isCsvPath = /* @__PURE__ */ __name((s) => isLocalPath(s) && /\.csv$/i.test(s.trim()), "isCsvPath");
+function isCsvLink(s) {
+  try {
+    const u = new URL(s);
+    if (!/^https?:$/.test(u.protocol)) return false;
+    return /\.csv$/i.test(u.pathname) || ["output", "format", "exportFormat"].some((k) => u.searchParams.get(k)?.toLowerCase() === "csv") || /out:csv/i.test(u.searchParams.get("tqx") ?? "");
+  } catch {
+    return false;
+  }
+}
+__name(isCsvLink, "isCsvLink");
+function decode(bytes) {
+  const utf8 = new TextDecoder("utf-8").decode(bytes);
+  return utf8.includes("�") ? new TextDecoder("gbk").decode(bytes) : utf8;
+}
+__name(decode, "decode");
+async function read(where, signal) {
+  if (isLocalPath(where)) {
+    if (!existsSync(where)) throw new Error(`FILE_NOT_FOUND: 找不到这个文件：${where}`);
+    return readFileSync(where);
+  }
+  if (!publicUrl(where)) throw new Error(`BLOCKED: 只读公网上的链接，不读本机或内网地址：${where}`);
+  const res = await fetch(where, { signal: signal ?? AbortSignal.timeout(6e4), headers: { "user-agent": UA2 } });
+  if (!publicUrl(res.url)) throw new Error(`BLOCKED: 链接跳到了本机或内网地址：${res.url}`);
+  if (res.status === 401 || res.status === 403) throw new Error(`NOT_PUBLIC: 这个链接要登录才能看（HTTP ${res.status}），请换成任何人都能打开的链接`);
+  if (!res.ok) throw new Error(`SITE_DOWN: 这个链接打不开（HTTP ${res.status}）。检查链接是否完整、是否还有效`);
+  if (/text\/html/i.test(res.headers.get("content-type") ?? "")) throw new Error("NOT_PUBLIC: 这个链接打开的是网页而不是 CSV（可能要登录，或者不是导出 / 发布链接）");
+  return new Uint8Array(await res.arrayBuffer());
+}
+__name(read, "read");
+var csv = {
+  id: "csv",
+  name: "CSV 表格",
+  input: "以 .csv 结尾或带 output=csv 的链接；或本机 .csv 文件的完整路径",
+  aggregate: true,
+  mode: "pull",
+  describe: "CSV 表格（链接或本机文件）：按订阅里存的列对应读成岗位；每次按周期重读整张表",
+  detect: /* @__PURE__ */ __name((url) => isCsvPath(url) || isCsvLink(url), "detect"),
+  siteKey: /* @__PURE__ */ __name((url) => isLocalPath(url) ? resolve(url.trim()).toLowerCase() : url.split("#")[0], "siteKey"),
+  async table(config, ctx) {
+    return csvToTable(decode(await read((config.url ?? "").trim(), ctx.signal)));
+  },
+  async fetch(config, ctx) {
+    return listingsFromTable(await this.table(config, ctx), config, "CSV 表格");
+  }
+};
+
+// apps/daemon/src/data/readers/gaoxiaojob.ts
+var M = "https://m.gaoxiaojob.com";
+var W = "https://www.gaoxiaojob.com";
+var UA3 = { "user-agent": "Mozilla/5.0 (iPhone; CPU iPhone OS 17_0 like Mac OS X) AppleWebKit/605.1.15 Mobile/15E148" };
+var DAY = 864e5;
+var LONG = /长期有效|长期招聘|常年|招满为止|招满即止|额满为止/;
+var columnOf = /* @__PURE__ */ __name((url) => /^https?:\/\/(?:www\.|m\.)?gaoxiaojob\.com\/column\/(\d+)\.html/.exec(url.trim())?.[1] ?? null, "columnOf");
+function dayOf(mmdd, today = /* @__PURE__ */ new Date()) {
+  const m = /^(\d{1,2})-(\d{1,2})$/.exec(mmdd.trim());
+  if (!m) return null;
+  const y = today.getUTCFullYear(), mm = m[1].padStart(2, "0"), dd = m[2].padStart(2, "0");
+  const d = `${y}-${mm}-${dd}`;
+  return Date.parse(d) > today.getTime() + DAY ? `${y - 1}-${mm}-${dd}` : d;
+}
+__name(dayOf, "dayOf");
+var text2 = /* @__PURE__ */ __name((html) => html.replace(/<script[\s\S]*?<\/script>|<style[\s\S]*?<\/style>/g, " ").replace(/<br\s*\/?>|<\/p>|<\/div>|<\/li>/g, "\n").replace(/<[^>]+>/g, " ").replace(/&nbsp;/g, " ").replace(/&amp;/g, "&").replace(/&lt;/g, "<").replace(/&gt;/g, ">").replace(/[ \t\r]+/g, " ").replace(/\n\s*/g, "\n"), "text");
+function parseNotice(id, title, html) {
+  const t = text2(html);
+  const flat = t.replace(/\s+/g, " ");
+  const field = /* @__PURE__ */ __name((k) => new RegExp(`${k}[:：]\\s*(.*?)\\s+(?:截止日期|学历要求|所属省份|工作地点|报名方式|栏目分类|需求学科|发布时间)`).exec(flat)?.[1]?.trim() ?? "", "field");
+  const posted = /发布时间[:：]\s*(\d{4}-\d{2}-\d{2})/.exec(flat)?.[1] ?? null;
+  const deadlineText = field("截止日期");
+  const full = /(\d{4})[-/.年](\d{1,2})[-/.月](\d{1,2})/.exec(deadlineText);
+  const closesAt = full ? `${full[1]}-${full[2].padStart(2, "0")}-${full[3].padStart(2, "0")}` : null;
+  const source = /来源链接[:：]\s*(https?:\/\/\S+)/.exec(flat)?.[1] ?? "";
+  const a = t.indexOf("公告详情"), b = t.indexOf("来源于网络", a);
+  const body = (a >= 0 ? t.slice(a + 4, b > a ? b : a + 8e3) : "").trim().slice(0, 8e3);
+  return { id, title, posted, deadlineText, closesAt, place: field("工作地点"), source, body };
+}
+__name(parseNotice, "parseNotice");
+function stillOpen(n, listedOn, today = /* @__PURE__ */ new Date()) {
+  if (n.closesAt) return Date.parse(`${n.closesAt}T23:59:59+08:00`) >= today.getTime();
+  const on = n.posted ?? listedOn;
+  if (!on) return false;
+  const age = (today.getTime() - Date.parse(on)) / DAY;
+  return age <= (LONG.test(n.deadlineText) ? 180 : 90);
+}
+__name(stillOpen, "stillOpen");
+function noticeListings(n, jobs, cfg) {
+  const company = (jobs.find((j) => j.companyName)?.companyName ?? "").replace(/#+$/, "").trim() || unitFrom(n.title);
+  const base = {
+    company,
+    noticeUrl: `${W}/announcement/detail/${n.id}.html`,
+    datePosted: n.posted ?? void 0,
+    validThrough: n.closesAt ?? void 0,
+    deadlineText: n.deadlineText || void 0,
+    cohort: cfg.cohort,
+    batch: n.title.slice(0, 80),
+    // one announcement, one campaign (campaigns group by publisher|company|cohort|batch)
+    publisher: "gaoxiaojob",
+    grade: "聚合",
+    isOriginal: false,
+    confidence: 0.7
+    // a reprint: below the 0.85–0.9 the official readers give
+  };
+  const intern = /* @__PURE__ */ __name((s) => /实习/.test(s) ? "实习" : void 0, "intern");
+  const jd = [n.body, n.source ? `原公告：${n.source}` : ""].filter(Boolean).join("\n\n");
+  if (!jobs.length) return [{ ...base, externalId: `ann:${n.id}`, url: base.noticeUrl, title: n.title, locations: n.place ? [n.place] : [], description: jd, kind: intern(n.title) }];
+  return jobs.map((j) => ({
+    ...base,
+    externalId: `job:${j.jobId}`,
+    url: j.url ? `${W}${j.url}` : base.noticeUrl,
+    title: j.jobName,
+    locations: j.city ? [j.city] : n.place ? [n.place] : [],
+    description: jd,
+    degreeReq: j.education || void 0,
+    majorReq: j.major || void 0,
+    kind: intern(`${j.jobName} ${n.title}`)
+  }));
+}
+__name(noticeListings, "noticeListings");
+var unitFrom = /* @__PURE__ */ __name((title) => /^(.+?)(?:\d{4}|诚聘|招聘|公开|引进|招收)/.exec(title)?.[1]?.trim() || title.slice(0, 30), "unitFrom");
+async function jobsOf(ctx, id) {
+  const out = [];
+  for (let page2 = 1; page2 <= 20; page2++) {
+    const d = await ctx.fetchJson(`${M}/announcement/get-job-list?announcementId=${id}&page=${page2}`, { headers: UA3 });
+    const list = d?.data?.list ?? [];
+    out.push(...list);
+    if (!list.length || out.length >= Number(d?.data?.count ?? 0)) return out;
+    await ctx.sleep(pause());
+  }
+  return out;
+}
+__name(jobsOf, "jobsOf");
+var pause = /* @__PURE__ */ __name(() => 600 + Math.floor(Math.random() * 800), "pause");
+var gaoxiaojob = {
+  id: "gaoxiaojob",
+  name: "高校人才网",
+  input: "www.gaoxiaojob.com/column/<栏目编号>.html（如武汉 48、武汉科研院所 509、武汉企业 511）",
+  mode: "pull",
+  describe: "高校人才网的一个栏目（研究院、实验室、高校团队、事业单位、企业）：读手机版公开列表和每条公告的岗位清单，只收仍有效的；转载来源，记为聚合，零 token",
+  detect: /* @__PURE__ */ __name((url) => !!columnOf(url), "detect"),
+  siteKey: /* @__PURE__ */ __name((url) => `column/${columnOf(url)}`, "siteKey"),
+  async fetch(cfg, ctx) {
+    const col = columnOf(cfg.url ?? "");
+    if (!col) throw new Error("SCHEMA_CHANGED: 不是高校人才网的栏目链接");
+    const maxPages = cfg.maxPages ?? 60;
+    const today = /* @__PURE__ */ new Date();
+    const out = [];
+    let oldPages = 0, page2 = 1;
+    for (; page2 <= maxPages; page2++) {
+      const d = await ctx.fetchJson(`${M}/home/get-announcement-list?columnId=${col}&isEstablishment=0&announcementHeat=0&page=${page2}`, { headers: UA3 });
+      if (d?.code !== 200) throw new Error(`SCHEMA_CHANGED: 列表接口返回 ${d?.code ?? "?"} ${d?.msg ?? ""}`);
+      const items = d?.data ?? [];
+      if (!items.length) break;
+      let fresh = 0;
+      for (const it of items) {
+        const listedOn = dayOf(it.refreshTime ?? "", today);
+        if (listedOn && (today.getTime() - Date.parse(listedOn)) / DAY > 180) continue;
+        fresh++;
+        await ctx.sleep(pause());
+        const n = parseNotice(
+          String(it.id),
+          it.title,
+          await ctx.fetchText(`${W}/announcement/detail/${it.id}.html`)
+          /* the desktop page: its 基本信息 block is what parseNotice reads */
+        );
+        if (!stillOpen(n, listedOn, today)) continue;
+        await ctx.sleep(pause());
+        out.push(...noticeListings(n, await jobsOf(ctx, n.id), cfg));
+      }
+      ctx.log(`gaoxiaojob ${col}: page ${page2} ${items.length} 条，在窗口内 ${fresh}（累计 ${out.length}）`);
+      oldPages = fresh ? 0 : oldPages + 1;
+      if (oldPages >= 2) break;
+      await ctx.sleep(pause());
+    }
+    if (page2 > maxPages) throw new Error(`TIMEOUT: 读到第 ${maxPages} 页还没读完这个栏目，这次不算（读一半会把另一半标成下架）`);
+    return out;
+  }
+};
+
+// apps/daemon/src/data/readers/index.ts
+var capturePush = { id: "capture", name: "任意页面抓取", mode: "push", describe: "任意招聘页/简章：书签一键捕获，或给 URL 由 daemon 抓取（公众号/就业网/官网），模型抽取岗位" };
+var sources = Object.fromEntries([moka, beisen, feishuBase, feishuJobs, zhilian, iguopin, chnenergy, chng, dayee, job51, kuaishou, xiaohongshu, gaoxiaojob, csv, capturePush].map((s) => [s.id, s]));
+function sourceFor(url) {
+  return Object.values(sources).find((s) => s.detect?.(url));
+}
+__name(sourceFor, "sourceFor");
+function siteKeyOf(url) {
+  const src = sourceFor(url);
+  return src ? `${src.id}:${src.siteKey?.(url) ?? url}` : url;
+}
+__name(siteKeyOf, "siteKeyOf");
+var aggregateIds = /* @__PURE__ */ __name(() => [...Object.values(sources).filter((s) => s.aggregate).map((s) => s.id), "sheet"], "aggregateIds");
+var isAggregate = /* @__PURE__ */ __name((sourceId) => aggregateIds().includes(sourceId), "isAggregate");
+var aggregateSql = /* @__PURE__ */ __name(() => aggregateIds().map((id) => `'${id}'`).join(","), "aggregateSql");
+var isNotices = /* @__PURE__ */ __name((sourceId) => !!sources[sourceId]?.notices, "isNotices");
+
+// apps/daemon/src/data/ingest.ts
+import { ulid as ulid2 } from "ulid";
+import { mkdirSync, writeFileSync, readdirSync, rmSync, statSync, readFileSync as readFileSync2 } from "node:fs";
+import { join } from "node:path";
+function campaignUrl(url) {
+  const hash = url.indexOf("#");
+  const base = hash < 0 ? url : url.slice(0, hash), frag = hash < 0 ? "" : url.slice(hash + 1);
+  const moka2 = base.includes("mokahr.com/campus-recruitment/") || base.includes("mokahr.com/social-recruitment/");
+  if (frag.startsWith("/job/") || moka2 && (frag === "" || frag === "/")) return moka2 ? base + "#/jobs" : base;
+  return url;
+}
+__name(campaignUrl, "campaignUrl");
+var sourceKey = /* @__PURE__ */ __name((aggregate, l) => aggregate || /#\d+$/.test(l.externalId) || l.externalId === l.url || l.externalId === l.title ? "" : l.externalId, "sourceKey");
+var ownSnapshots = /* @__PURE__ */ __name((dir2, configId) => readdirSync(dir2).filter((f) => f.slice(25) === `${configId}.json`).sort(), "ownSnapshots");
+var SNAPSHOT_KEEP = 7;
+function keepSnapshot(vault, sourceId, configId, listings) {
+  const dir2 = join(vault, "snapshots", sourceId);
+  mkdirSync(dir2, { recursive: true });
+  const at = now(), body = JSON.stringify({ sourceId, configId, at, count: listings.length, listings }, null, 1);
+  const prior = ownSnapshots(dir2, configId);
+  const last = prior.length ? join(dir2, prior[prior.length - 1]) : "";
+  const tail = /* @__PURE__ */ __name((s) => {
+    const i = s.indexOf('\n "count"');
+    return i > 0 ? s.slice(i) : null;
+  }, "tail");
+  try {
+    if (last && statSync(last).size === Buffer.byteLength(body) && tail(readFileSync2(last, "utf8")) === tail(body)) return last;
+  } catch {
+  }
+  const path = join(dir2, `${at.replace(/[:.]/g, "-")}-${configId}.json`);
+  writeFileSync(path, body, "utf8");
+  for (const old of [...prior, path.slice(dir2.length + 1)].sort().slice(0, -SNAPSHOT_KEEP)) rmSync(join(dir2, old), { force: true });
+  return path;
+}
+__name(keepSnapshot, "keepSnapshot");
+function ingest(store, vault, sourceId, configId, listings, opts = {}) {
+  const db = store.db;
+  const runId = ulid2();
+  const started = now();
+  const snapshotPath = keepSnapshot(vault, sourceId, configId, listings);
+  const groups = /* @__PURE__ */ new Map();
+  const groupOf = listings.map((l) => {
+    const k = `${l.publisher}|${l.company}|${l.cohort ?? ""}|${l.batch ?? ""}`;
+    const g = groups.get(k) ?? groups.set(k, { listings: [] }).get(k);
+    g.listings.push(l);
+    return g;
+  });
+  const filed = /* @__PURE__ */ __name((l) => !!l.title && !!l.company && !genericCompany(l.company), "filed");
+  const ofGroup = /* @__PURE__ */ __name((g, l) => {
+    let w = g.whole;
+    if (!w) {
+      const group = g.listings;
+      const d = group.find((x) => x.validThrough || x.deadlineText) ?? l;
+      const perPost = new Set(group.map((x) => x.validThrough?.slice(0, 10)).filter(Boolean)).size > 1;
+      const dl = perPost ? { closesAt: null, kind: "unknown" } : d.validThrough ? { closesAt: d.validThrough.slice(0, 10), kind: "date" } : parseDeadline(d.deadlineText ?? "");
+      g.whole = w = {
+        externalKey: campaignKey(sourceId, l.publisher, l.company, l.cohort ?? "", l.batch ?? ""),
+        cohort: l.cohort ?? "",
+        batch: l.batch ?? "",
+        cities: Array.from(new Set(group.flatMap((x) => x.locations))).slice(0, 20),
+        degreeReq: group.find((x) => x.degreeReq)?.degreeReq ?? "",
+        majorReq: group.find((x) => x.majorReq)?.majorReq ?? "",
+        opensAt: null,
+        closesAt: dl.closesAt,
+        deadlineKind: dl.kind,
+        deadlineText: perPost ? "各岗位截止不同" : d.deadlineText ?? "",
+        sourceKind: sourceId,
+        hasWrittenTest: "",
+        notes: ""
+      };
+    }
+    return w;
+  }, "ofGroup");
+  const campaigns = /* @__PURE__ */ new Set();
+  const listedRows = /* @__PURE__ */ new Map();
+  let positionsNew = 0;
+  let positionsSeen = 0;
+  let positionsClosed = 0;
+  let updated = 0;
+  tx(db, () => {
+    const batch = new CatalogBatch(db, listings.flatMap((l, i) => filed(l) ? [ofGroup(groupOf[i], l).externalKey] : []));
+    const verifiedAt = now();
+    const aggregate = isAggregate(sourceId);
+    const rows = /* @__PURE__ */ new Map();
+    const rowKey = /* @__PURE__ */ __name((externalId, positionId) => `${externalId}\0${positionId}`, "rowKey");
+    inChunks(batch.campaignIds(), (part, q) => {
+      for (const r of db.prepare(`select id, external_id, position_id, extracted_json from source_listing where campaign_id in (${q}) and +provider = ? order by rowid`).all(...part, sourceId)) {
+        const k = rowKey(r.external_id, r.position_id);
+        if (!rows.has(k)) rows.set(k, { id: r.id, extracted_json: r.extracted_json });
+      }
+    });
+    const findRow = db.prepare("select id, extracted_json from source_listing where provider = ? and external_id = ? and position_id = ?");
+    const updateRow = db.prepare("update source_listing set url = ?, extracted_json = ?, confidence = ?, content_path = ?, fetched_at = ? where id = ?");
+    const insertRow = db.prepare("insert into source_listing(id, position_id, campaign_id, provider, tenant, external_id, url, publisher, grade, is_original, extracted_json, confidence, content_path, fetched_at) values (?,?,?,?,?,?,?,?,?,?,?,?,?,?)");
+    for (const [i, l] of listings.entries()) {
+      if (!filed(l)) continue;
+      const companyId = batch.company(l.company, l.companyType ?? "", l.industry ?? "");
+      const { id: campaignId } = batch.campaign({
+        ...ofGroup(groupOf[i], l),
+        companyId,
+        kind: l.kind ?? "校招",
+        sourceGrade: l.grade,
+        applyUrl: campaignUrl(l.url),
+        noticeUrl: campaignUrl(l.noticeUrl ?? l.url),
+        lastVerifiedAt: verifiedAt
+      });
+      campaigns.add(campaignId);
+      const p = batch.position(campaignId, l.title, { jdText: l.description ?? "", cities: l.locations, degreeReq: l.degreeReq ?? "", majorReq: l.majorReq ?? "", closesAt: l.validThrough ? l.validThrough.slice(0, 10) : null, sourceKey: sourceKey(aggregate, l) });
+      p.created ? positionsNew++ : positionsSeen++;
+      const said = json.str({ ...l, raw: void 0 }), k = rowKey(l.externalId, p.id);
+      listedRows.set(k, p.id);
+      const existing = rows.get(k) ?? findRow.get(sourceId, l.externalId, p.id);
+      if (existing) {
+        if (existing.extracted_json !== said) {
+          updateRow.run(l.url, said, l.confidence, snapshotPath, now(), existing.id);
+          rows.set(k, { id: existing.id, extracted_json: said });
+        }
+      } else {
+        const id = ulid2();
+        insertRow.run(id, p.id, campaignId, sourceId, l.publisher, l.externalId, l.url, l.publisher, l.grade, l.isOriginal ? 1 : 0, said, l.confidence, snapshotPath, now());
+        rows.set(k, { id, extracted_json: said });
+      }
+    }
+    batch.finish();
+    updated = batch.updated;
+    if (opts.closeUnseen && opts.complete !== false && listings.length && campaigns.size) {
+      const ids = [...campaigns];
+      inChunks(ids, (part, q) => {
+        positionsClosed += Number(db.prepare(`update position set status = 'closed', updated_at = ? where campaign_id in (${q})
+          and status = 'open' and last_seen_at < ? and id not in (select position_id from application where state != 'cancelled')`).run(now(), ...part, started).changes);
+      });
+      const gone = db.prepare("select distinct campaign_id id from source_listing where provider = ? and content_path like ? and campaign_id is not null").all(sourceId, `%-${configId}.json`).map((r) => r.id).filter((id) => !campaigns.has(id));
+      for (let i = 0; i < gone.length; i += 200) {
+        const part = gone.slice(i, i + 200);
+        positionsClosed += Number(db.prepare(`update position set status = 'closed', updated_at = ? where campaign_id in (${part.map(() => "?").join(",")})
+          and status = 'open' and id not in (select position_id from application where state != 'cancelled')`).run(now(), ...part).changes);
+      }
+    }
+    db.prepare("insert into scan_run(id, source, config_id, started_at, finished_at, complete, found, new_positions, error, snapshot_path) values (?,?,?,?,?,?,?,?,?,?)").run(runId, sourceId, configId, started, now(), opts.complete === false ? 0 : 1, listings.length, positionsNew, opts.error ?? "", snapshotPath);
+  });
+  return { runId, received: listings.length, campaigns: campaigns.size, positionsNew, positionsSeen, positionsClosed, updated, snapshotPath, listed: { campaignIds: [...campaigns], positionIds: [...listedRows.values()] } };
+}
+__name(ingest, "ingest");
+function recordFailedRun(store, sourceId, configId, error) {
+  store.db.prepare("insert into scan_run(id, source, config_id, started_at, finished_at, complete, found, new_positions, error, snapshot_path) values (?,?,?,?,?,?,?,?,?,?)").run(ulid2(), sourceId, configId, now(), now(), 0, 0, 0, error, "");
+}
+__name(recordFailedRun, "recordFailedRun");
+
+// apps/daemon/src/data/sync.ts
+var SYNC_REASONS = {
+  NO_OFFICIAL_LINK: "还没有这家公司的官网招聘链接",
+  UNSUPPORTED_PLATFORM: "官网用的招聘系统还读不了",
+  TIMEOUT: "官网超时没响应",
+  LOGIN_REQUIRED: "官网要登录才能看",
+  BLOCKED: "被官网拦了（验证码或访问限制）",
+  SCHEMA_CHANGED: "官网改版，读取方式要更新",
+  DECRYPT_FAILED: "官网返回的数据解不开",
+  ZERO_POSTS: "官网读到了，但一个岗位都没有",
+  SITE_DOWN: "官网打不开",
+  FAILED: "读取失败",
+  NOT_PUBLIC: "表格没有公开，读不到",
+  BAD_URL: "链接不是这类招聘站的地址",
+  NO_COMPANY_COLUMN: "没认出哪一列是公司",
+  EMPTY_TABLE: "表格读到了，但没有一行能成岗位",
+  FILE_NOT_FOUND: "找不到这个文件"
+};
+function reasonOf(error) {
+  const code = /^([A-Z_]{4,}):/.exec(error)?.[1];
+  if (code && code in SYNC_REASONS) return code;
+  if (/timeout|timed out|ETIMEDOUT|aborted|超时/i.test(error)) return "TIMEOUT";
+  if (/\b401\b|login|登录/i.test(error)) return "LOGIN_REQUIRED";
+  if (/\b(403|429)\b|captcha|验证码|forbidden/i.test(error)) return "BLOCKED";
+  if (/decrypt|解密/i.test(error)) return "DECRYPT_FAILED";
+  if (/ENOTFOUND|ECONNREFUSED|ECONNRESET|fetch failed|\b5\d\d\b/i.test(error)) return "SITE_DOWN";
+  if (/没找到有岗位|没有岗位/.test(error)) return "ZERO_POSTS";
+  if (/未找到对应的官网|\b404\b/.test(error)) return "SITE_DOWN";
+  if (/JSON|Unexpected token|返回异常|not iterable|undefined|parameter error|没找到企业编号|没有换到/i.test(error)) return "SCHEMA_CHANGED";
+  return "FAILED";
+}
+__name(reasonOf, "reasonOf");
+var running = /* @__PURE__ */ new Map();
+async function runSource(ctx, configId, opts = {}) {
+  const cfg = ctx.sources.find((c) => c.id === configId);
+  if (!cfg) return { ok: false, error: `NOT_FOUND: source config ${configId}` };
+  const src = sources[cfg.source];
+  if (!src?.fetch) return { ok: false, error: `NOT_FOUND: source kind ${cfg.source} is not a pull source` };
+  const progress = { stage: "reading", since: (/* @__PURE__ */ new Date()).toISOString(), done: 0, total: 0 };
+  running.set(cfg.id, progress);
+  ctx.onPull?.(cfg.id, true);
+  const log = /* @__PURE__ */ __name((m) => {
+    console.log(`[scan] ${m}`);
+    const n = /: (\d+)\/(\d+)$/.exec(m);
+    if (n) {
+      progress.done = Number(n[1]);
+      progress.total = Number(n[2]);
+    }
+  }, "log");
+  try {
+    const listings = await src.fetch(cfg, httpContext(log));
+    if (!listings.length && !cfg.discover && !cfg.keywords?.length) throw new Error(src.table ? "EMPTY_TABLE: 表格读到了，但没有一行能成岗位（没有公司、已过截止、或届次对不上的行都不算）" : "ZERO_POSTS: 官网读到了，但一个岗位都没有");
+    Object.assign(progress, { stage: "filing", done: 0, total: listings.length });
+    await new Promise((ok) => setImmediate(ok));
+    const { listed, ...result } = ingest(ctx, ctx.vault, src.id, cfg.id, listings, { closeUnseen: true });
+    if (!opts.quiet && (result.positionsNew || result.positionsClosed)) ctx.onDataChanged?.();
+    else if (result.updated) ctx.onPostsEdited?.();
+    return { ok: true, result, listed };
+  } catch (e) {
+    const error = String(e?.message ?? e);
+    recordFailedRun(ctx, src.id, cfg.id, error);
+    return { ok: false, error };
+  } finally {
+    if (running.get(cfg.id) === progress) running.delete(cfg.id);
+    ctx.onPull?.(cfg.id, false);
+  }
+}
+__name(runSource, "runSource");
+function listSources(ctx) {
+  const kinds = Object.values(sources).map((x) => ({ id: x.id, mode: x.mode, describe: x.describe }));
+  const configured = ctx.sources.map((c) => {
+    const runs = ctx.db.prepare("select started_at, found, new_positions, complete, error from scan_run where config_id = ? order by started_at desc limit 10").all(c.id);
+    const r = runs[0], firstOk = runs.findIndex((x) => !x.error);
+    return {
+      id: c.id,
+      source: c.source,
+      name: c.name ?? "",
+      url: c.url ?? "",
+      enabled: c.enabled !== false,
+      everyHours: c.everyHours ?? 24,
+      auto: c.auto === true,
+      lastRun: r ? { at: r.started_at, found: r.found, newPositions: r.new_positions, complete: !!r.complete, error: r.error } : null,
+      // health from what the runs themselves say: when it last worked, how many times in a row it has not, and why
+      health: { lastOkAt: runs[firstOk]?.started_at ?? null, failures: firstOk < 0 ? runs.length : firstOk, reason: r?.error ? reasonOf(r.error) : "" },
+      ...running.has(c.id) ? { running: { ...running.get(c.id) } } : {}
+    };
+  });
+  return { kinds, configured };
+}
+__name(listSources, "listSources");
+var RETRY_SOON_MS = 60 * 6e4;
+var SOON_TRIES = 3;
+function sourceScan(ctx) {
+  return {
+    kind: "source_scan",
+    order: ["pull"],
+    steps: {
+      pull: /* @__PURE__ */ __name(async ({ subjectId, checkpoint }) => {
+        const cfg = ctx.sources.find((c) => c.id === subjectId);
+        if (!cfg || cfg.enabled === false) return { type: "done" };
+        const r = await runSource(ctx, subjectId);
+        const hours = cfg.everyHours ?? 24;
+        if (hours <= 0) return { type: "done" };
+        const pace = hours * 36e5;
+        if (!r.ok) {
+          const failures = (checkpoint?.failures ?? 0) + 1;
+          return { type: "sleep", until: new Date(Date.now() + (failures <= SOON_TRIES ? Math.min(RETRY_SOON_MS, pace) : pace)).toISOString(), checkpoint: { ...checkpoint, failures, lastError: r.error } };
+        }
+        return { type: "sleep", until: new Date(Date.now() + pace).toISOString(), checkpoint: { lastRunId: r.result.runId } };
+      }, "pull")
+    }
+  };
+}
+__name(sourceScan, "sourceScan");
+function scheduleSources(ctx) {
+  let n = 0;
+  for (const c of ctx.sources) {
+    if (c.enabled === false || (c.everyHours ?? 24) <= 0) continue;
+    ctx.workflows?.ensure("source_scan", c.id, `source_scan:${c.id}`);
+    n++;
+  }
+  return n;
+}
+__name(scheduleSources, "scheduleSources");
+var PREVIEW_KEEP_MS = 3 * 6e4;
+var previewed = null;
+async function previewSource(ctx, p) {
+  const url = p.url.trim();
+  const src = sourceFor(url);
+  if (!src?.fetch) throw Object.assign(new Error(`UNSUPPORTED_URL: 还认不出这个链接是哪种招聘站或表格：${url}`), { code: "UNSUPPORTED_URL" });
+  const subscribed = ctx.sources.find((c) => c.url && siteKeyOf(c.url) === siteKeyOf(url))?.id ?? null;
+  const fields = Object.keys(FIELDS).map((field) => ({ field, label: FIELDS[field], required: field === "company" }));
+  const base = { source: src.id, name: src.name, subscribed, fields };
+  if (!src.table) return { ...base, table: false, headers: [], columns: {}, rows: 0, posts: 0, companies: 0, sample: [], problem: "" };
+  const key = `${src.id}:${url}`;
+  const table = previewed && previewed.key === key && Date.now() - previewed.at < PREVIEW_KEEP_MS ? previewed.table : await src.table({ id: "preview", source: src.id, url }, httpContext());
+  previewed = { key, at: previewed?.key === key ? previewed.at : Date.now(), table };
+  const columns = Object.fromEntries(Object.entries(p.columns ?? guessColumns(table.headers)).filter(([f, h]) => f in FIELDS && table.headers.includes(h)));
+  const listings = columns.company ? tableToListings(table, columns, { cohort: p.cohort }) : [];
+  const problem = !table.rows.length ? "表是空的" : !columns.company ? `没认出哪一列是公司。现有的列：${table.headers.slice(0, 20).join("、")}` : !listings.length ? "没有一行能成岗位：公司为空、已过截止、或届次对不上的行都不算" : "";
+  return {
+    ...base,
+    table: true,
+    headers: table.headers,
+    columns,
+    rows: table.rows.length,
+    posts: listings.length,
+    companies: new Set(listings.map((l) => l.company)).size,
+    sample: listings.slice(0, 3).map((l) => ({ company: l.company, title: l.title, deadline: l.deadlineText ?? "", cities: l.locations, url: l.url })),
+    problem
+  };
+}
+__name(previewSource, "previewSource");
+
+// apps/daemon/src/data/subscriptions.ts
+function write(vault, list) {
+  updateConfig(vault, (doc) => {
+    doc.sources = list;
+  });
+}
+__name(write, "write");
+function saveSources(ctx, list) {
+  write(ctx.vault, list);
+  const kept = new Set(list.map((c) => c.id));
+  for (const c of ctx.sources) if (!kept.has(c.id)) ctx.workflows?.end(`source_scan:${c.id}`);
+  ctx.sources.splice(0, ctx.sources.length, ...list);
+  ctx.onSourcesChanged?.();
+}
+__name(saveSources, "saveSources");
+function addSource(ctx, p) {
+  p = { ...p, url: p.url.trim() };
+  const src = sourceFor(p.url);
+  if (!src) throw Object.assign(new Error(`UNSUPPORTED_URL: no pull source detects ${p.url}`), { code: "UNSUPPORTED_URL" });
+  const key = siteKeyOf(p.url);
+  const same = ctx.sources.find((c) => c.url && siteKeyOf(c.url) === key);
+  if (same) return { id: same.id, source: same.source };
+  const base = `${src.id}-${p.name.replace(/[^\w一-鿿]+/g, "").slice(0, 20) || "src"}`;
+  let id = base;
+  for (let i = 2; ctx.sources.some((c) => c.id === id); i++) id = `${base}-${i}`;
+  const entry = { id, source: src.id, name: p.name, url: p.url, cohort: p.cohort, batch: p.batch, everyHours: p.everyHours ?? 24, enabled: true, ...p.auto ? { auto: true } : {}, ...p.columns && src.table ? { columns: p.columns } : {} };
+  write(ctx.vault, [...ctx.sources, entry]);
+  ctx.sources.push(entry);
+  ctx.onSourcesChanged?.();
+  return { id, source: src.id };
+}
+__name(addSource, "addSource");
+function removeSource(ctx, id) {
+  const cfg = ctx.sources.find((c) => c.id === id);
+  if (!cfg) throw Object.assign(new Error(`NOT_FOUND: 没有这个订阅：${id}`), { code: "NOT_FOUND" });
+  saveSources(ctx, ctx.sources.filter((c) => c.id !== id));
+  return cfg;
+}
+__name(removeSource, "removeSource");
+
+export {
+  genericCompany,
+  companyKey,
+  sha,
+  campaignKey,
+  upsertCompany,
+  upsertCampaign,
+  upsertPosition,
+  parseCsv,
+  parseDeadline,
+  splitList,
+  gradeFromSource,
+  parseBaseUrl,
+  readBase,
+  siteLinks,
+  httpContext,
+  publicUrl,
+  isLocalPath,
+  sources,
+  sourceFor,
+  siteKeyOf,
+  isAggregate,
+  aggregateSql,
+  isNotices,
+  ingest,
+  recordFailedRun,
+  SYNC_REASONS,
+  reasonOf,
+  runSource,
+  listSources,
+  sourceScan,
+  scheduleSources,
+  previewSource,
+  saveSources,
+  addSource,
+  removeSource
+};