// Zürich sweep via job-room.ch (canton ZH), scored against cv/polo.md
import { NON_ENGLISH_RE, JUNIOR_TITLE_RE, JAVA_TITLE_RE, SOFTWARE_TITLE_RE } from "../fetch.ts";

const MS_STACK_RE = /\.net\b|\bc#|sharepoint|dynamics 365|power platform|\bsccm\b|intune|successfactors|\bsap\b|\bm365\b|servicenow|salesforce|avaloq|abap/i;
const PHARMA_RE = /\b(clinical|pharmaceutical|drug\s+(substance|product)|ha?ematology|oncology|formulation|laborator|chemist|chemistry|biolog|preclinical|regulatory\s+affairs|medical\s+affairs|veterinary|pathologist)\b/i;
const DE_FR_RE = /\b(und|oder|wir|Sie|Ihre|Ihrer|Kenntnisse|Erfahrung|Aufgaben|Anforderungen|Stelle|Unternehmen|Mitarbeiter|Bereich|nous|vous|votre|pour|avec|des|les)\b/g;
const EMBEDDED_RE = /\bembedded\b|\bfirmware\b|\bplc\b|\bsps\b|\bqt\b|\bc\+\+\b|automation\s?engineer/i;
const SUPPORT_RE = /service\s?desk|1st|2nd\s?level|helpdesk|support\s?engineer|technician|operator/i;

const SENIOR_RE = /\b(senior|staff|principal|lead|head|director|architect|vp)\b/i;
const STACK_RE = /\b(python|typescript|javascript|ruby|rails|react|node\.?js|golang|\bgo\b|postgres|kafka)\b/i;
const DOMAIN_RE = /\b(platform|kubernetes|docker|terraform|ci\/cd|devops|aws|data\s+(pipeline|platform|engineer)|llm|genai|microservice|sre)\b/i;

const strip = (s: string) => s.replace(/<[^>]+>/g, "").replace(/\s+/g, " ").trim();

const API = "https://www.job-room.ch/jobadservice/api/jobAdvertisements/_search";
const UA = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36";
const KWS = ["software", "engineer", "developer", "devops", "data", "python", "cloud",
             "platform", "typescript", "backend", "full stack", "ai", "kubernetes", "react"];

const ads = new Map<string, any>();
for (const kw of KWS) {
  process.stdout.write(`${kw} `);
  for (let page = 0; page < 8; page++) {
    const res = await fetch(`${API}?page=${page}&size=100`, {
      method: "POST",
      headers: { "Content-Type": "application/json", Accept: "application/json", "User-Agent": UA },
      body: JSON.stringify({
        onlineSince: 30, displayRestricted: false, keywords: [kw],
        cantonCodes: ["ZH"], workloadPercentageMin: 0, workloadPercentageMax: 100,
      }),
    });
    if (!res.ok) break;
    const hits = (await res.json()) as any[];
    for (const h of hits) if (h.jobAdvertisement?.id) ads.set(h.jobAdvertisement.id, h.jobAdvertisement);
    if (hits.length < 100) break;
    await Bun.sleep(200);
  }
  await Bun.sleep(200);
}
console.log(`\n${ads.size} unique ads in ZH\n`);

const kept: any[] = [];
for (const ad of ads.values()) {
  const c = ad.jobContent;
  const d = c?.jobDescriptions?.[0];
  if (!d) continue;
  const title = strip(d.title);
  if (NON_ENGLISH_RE.test(title) || JUNIOR_TITLE_RE.test(title) || JAVA_TITLE_RE.test(title)) continue;
  if (!SOFTWARE_TITLE_RE.test(title)) continue;
  if (MS_STACK_RE.test(title) || PHARMA_RE.test(title)) continue;
  if (EMBEDDED_RE.test(title) || SUPPORT_RE.test(title)) continue;

  const body = strip(d.description ?? "");
  if ((body.match(DE_FR_RE) ?? []).length >= 6) continue;
  if (MS_STACK_RE.test(body)) continue;

  const hay = `${title} ${body}`;
  let score = 0;
  if (SENIOR_RE.test(title)) score += 3;
  score += (hay.match(STACK_RE) ?? []).length;
  score += (hay.match(DOMAIN_RE) ?? []).length;

  const langs = (c.languageSkills ?? [])
    .map((l: any) => `${l.languageIsoCode}:${l.spokenLevel ?? "?"}`).join(" ");
  // A declared German requirement above B1 is disqualifying
  const hardGerman = /de(-ch)?:(PROFICIENT|ADVANCED)/.test(langs);

  kept.push({
    score, title,
    company: c.company?.name ?? "",
    city: c.location?.city ?? "",
    workload: `${c.employment?.workloadPercentageMin ?? "?"}-${c.employment?.workloadPercentageMax ?? "?"}%`,
    posted: ad.createdTime?.slice(0, 10) ?? "",
    langs, hardGerman,
    url: c.externalUrl ?? `https://www.job-room.ch/job-search/${ad.id}`,
    snippet: body.slice(0, 230),
  });
}

kept.sort((a, b) => b.score - a.score || b.posted.localeCompare(a.posted));
const good = kept.filter((k) => !k.hardGerman);
console.log(`${kept.length} survivors (${kept.length - good.length} dropped for German requirement)\n`);

for (const j of good.slice(0, 30)) {
  console.log(`### [${j.score}] ${j.title} — ${j.company}`);
  console.log(`    ${j.city} · ${j.workload} · ${j.posted}${j.langs ? ` · ${j.langs}` : ""}`);
  console.log(`    ${j.url}`);
  console.log(`    ${j.snippet}\n`);
}