Karriere.at jobs search scraper

This Apify Actor extracts complete, source-backed job records locally or in Apify Cloud from the public Karriere.at search and detail pages.

Modes

• search discovers jobs for one query and location, then enriches each result from its official detail page.
• searchMultiple runs bounded searches for each value in queries and deduplicates the detail records.
• single enriches one official https://www.karriere.at/jobs/<numeric-id> URL.
• multiple enriches a bounded list of official detail URLs.
• startUrls accepts official search or detail URLs.

The default input is in ./INPUT.json. Reproducible examples are provided in INPUT-single.json, INPUT-multiple.json, INPUT-search-multiple.json, INPUT-start-urls.json, and INPUT-negative.json.

Extraction contract

Each dataset row is written only after the detail page exposes a matching numeric ID, canonical URL, title, employer, location, and a sufficiently rich description. The mapper preserves published JSON-LD and HTML-derived descriptions, headings, bullets, sections, job facts, candidate criteria, employment, work model, location, salary, dates, company metadata, and explicit application links. Missing optional source values are omitted; the canonical job URL is never used as a fabricated application URL.

Rows are buffered until validation succeeds, deduplicated by official job ID and URL, and written with strict schemas. RUN_SUMMARY, RUN_DIAGNOSTICS, RUN_SKIPS, REQUEST_RECEIPTS, RUN_HEALTH, and RUN_METADATA stay in the key-value store rather than being emitted as job rows.

Apify's platform status is only SUCCEEDED or FAILED; resultStatus separately reports COMPLETE, LIMITED, SKIPPED, DEFERRED, NO_DATA, or FAILED. If genuine records were persisted before a later non-block error, the run remains SUCCEEDED with resultStatus: LIMITED. A fatal zero-row run calls Actor.fail(). A confirmed source barrier stops later requests and discards buffered rows. Diagnostics are KVS-only and never appear in the job dataset.

Local verification

npm install
npm test
npm run lint
npm run check
npm run validate
npx --yes apify-cli validate-schema
npx --yes apify-cli run --purge --input-file INPUT.json

The Actor uses ordinary native HTTPS requests sequentially with bounded timeouts and a 5 MB response cap. It does not use a proxy, fingerprint spoofing, CAPTCHA/WAF bypass, or reader mirror. HTTP 401/403/429/451 or visible source-authored challenge text stops further requests and records the evidence in KVS; no blocked or diagnostic item is written as a job row.