chore: update search scripts, extra sources and post drafts

Adds the one-off scripts accumulated across recent sessions, removes temp scripts whose searches now live in config.ts as role-family sweeps, and updates fetch-other.ts plus the LinkedIn post drafts.
2026-08-27 09:03:05 +02:00
parent 3f5360f44c
commit b4957025b1
31 changed files with 2008 additions and 283 deletions
@@ -0,0 +1,101 @@
+// Luzern sweep: LinkedIn + job-room.ch (canton LU)
+import { KEYWORDS } from "../config.ts";
+import type { SearchConfig } from "../config.ts";
+import { fetchJobs, isEnglishContent, 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/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 strip = (s: string) => s.replace(/<[^>]+>/g, "").replace(/\s+/g, " ").trim();
+
+// ---------- LinkedIn ----------
+console.log("=== LinkedIn: Lucerne ===");
+const config: SearchConfig = {
+  label: "Luzern",
+  keywords: KEYWORDS,
+  location: "Lucerne, Switzerland",
+  workType: ["onsite", "hybrid"],
+  timePosted: "month",
+};
+
+const liJobs = await fetchJobs(config);
+console.log(`${liJobs.length} after title filters\n`);
+
+const liKept: any[] = [];
+for (const j of liJobs) {
+  if (MS_STACK_RE.test(j.title)) continue;
+  process.stdout.write(`  ${j.title} … `);
+  const ok = await isEnglishContent(j.id);
+  console.log(ok ? "ok" : "dropped (DE/FR)");
+  if (ok) liKept.push(j);
+}
+
+// ---------- job-room.ch, canton LU ----------
+console.log("\n=== job-room.ch: canton LU ===");
+const API = "https://www.job-room.ch/jobadservice/api/jobAdvertisements/_search";
+const UA = "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36";
+const KWS = ["software", "engineer", "developer", "devops", "data", "python", "cloud", "platform", "backend", "full stack", "ai"];
+
+const ads = new Map<string, any>();
+for (const kw of KWS) {
+  for (let page = 0; page < 5; page++) {
+    const res = await fetch(`${API}?page=${page}&size=100`, {
+      method: "POST",
+      headers: { "Content-Type": "application/json", Accept: "application/json", "User-Agent": UA },
+      body: JSON.stringify({
+        onlineSince: 30, displayRestricted: false, keywords: [kw],
+        cantonCodes: ["LU"], workloadPercentageMin: 0, workloadPercentageMax: 100,
+      }),
+    });
+    if (!res.ok) break;
+    const hits = (await res.json()) as any[];
+    for (const h of hits) if (h.jobAdvertisement?.id) ads.set(h.jobAdvertisement.id, h.jobAdvertisement);
+    if (hits.length < 100) break;
+    await Bun.sleep(200);
+  }
+  await Bun.sleep(200);
+}
+console.log(`${ads.size} unique ads in LU`);
+
+const jrKept: 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;
+
+  const body = strip(d.description ?? "");
+  if ((body.match(DE_FR_RE) ?? []).length >= 6) continue;
+  if (MS_STACK_RE.test(body)) continue;
+
+  const langs = (c.languageSkills ?? [])
+    .map((l: any) => `${l.languageIsoCode}:${l.spokenLevel ?? "?"}/${l.writtenLevel ?? "?"}`).join(" ");
+
+  jrKept.push({
+    title,
+    company: c.company?.name ?? "",
+    city: c.location?.city ?? "",
+    workload: `${c.employment?.workloadPercentageMin ?? "?"}-${c.employment?.workloadPercentageMax ?? "?"}%`,
+    posted: ad.createdTime?.slice(0, 10) ?? "",
+    langs,
+    url: c.externalUrl ?? `https://www.job-room.ch/job-search/${ad.id}`,
+    snippet: body.slice(0, 200),
+  });
+}
+
+// ---------- output ----------
+console.log(`\n\n######## LinkedIn survivors: ${liKept.length} ########\n`);
+for (const j of liKept) console.log(`### ${j.title} [${j.company}]\n    ${j.location} · ${j.postedAt}\n    ${j.link}\n`);
+
+console.log(`\n######## job-room survivors: ${jrKept.length} ########\n`);
+jrKept.sort((a, b) => b.posted.localeCompare(a.posted));
+for (const j of jrKept) {
+  console.log(`### ${j.title} [${j.company}]`);
+  console.log(`    ${j.city} · ${j.workload} · ${j.posted}`);
+  if (j.langs) console.log(`    langs: ${j.langs}`);
+  console.log(`    ${j.url}`);
+  console.log(`    ${j.snippet}\n`);
+}