#!/usr/bin/env bun
import { searches } from "../config.ts";
import { fetchJobs, isEnglishContent } from "../fetch.ts";

const CSV = "./results/jobs.csv";
const BASEL_LABELS = new Set(["Basel (city)", "Basel Metropolitan Area", "Basel-Country"]);

const existingIds = new Set<string>();
try {
  const text = await Bun.file(CSV).text();
  for (const line of text.split("\n").slice(1)) {
    const id = line.match(/\/jobs\/view\/(\d+)\//)?.[1];
    if (id) existingIds.add(id);
  }
} catch { /* first run */ }

type Row = { id: string; title: string; company: string; location: string; link: string; posted: string; isNew: boolean };
const seen = new Set<string>();
const rows: Row[] = [];

for (const config of searches.filter((s) => BASEL_LABELS.has(s.label))) {
  const jobs = await fetchJobs(config);
  console.error(`[${config.label}] ${jobs.length} jobs`);
  for (const job of jobs) {
    if (seen.has(job.id)) continue;
    seen.add(job.id);
    rows.push({
      id: job.id,
      title: job.title,
      company: job.company,
      location: (job as any).location ?? "",
      link: job.link,
      posted: (job as any).postedAt ?? "",
      isNew: !existingIds.has(job.id),
    });
  }
}

const survivors: Row[] = [];
for (const r of rows) {
  if (await isEnglishContent(r.id)) survivors.push(r);
  else console.error(`  DE/FR content dropped: ${r.title} — ${r.company}`);
}

console.error(`\n${survivors.length} survivors (${survivors.filter((r) => r.isNew).length} new)`);
console.log(JSON.stringify(survivors, null, 2));