EU Academic Jobs Aggregator

Aggregate public PhD, postdoc, professor, and research positions from configurable European job portals. The defaults cover public pages for EURAXESS, DAAD, EuroScienceJobs, UniversityPositions, and INOMICS; additional portal URLs can be supplied.

The Actor discovers same-domain job links, follows bounded result pagination, and optionally visits detail pages. It reads structured JobPosting metadata when available and includes a labeled-HTML fallback for public EURAXESS details. Requests use timeouts, bounded retries, browser-like public headers, and a configurable delay.

Listings are deduplicated with stable recordId values and preserve the portal hostname, academic level, research field, degree requirement, publication date, deadline, source URL, and extraction timestamp when those fields are public. keywords, country, and academicOnly are applied locally.

Key inputs are startUrls, keywords, country, academicOnly, maxItems, maxPages, includeDetails, timeoutSecs, maxRetries, and requestDelayMs. OUTPUT records platform status separately from resultStatus, along with successfully stored row counts, request receipts, diagnostics, failures, duration, and data availability. Only a complete record with a successful same-portal 2xx request receipt is eligible for the Dataset. Because there is no validated per-search first-party empty-results detector, every unrecognized zero-row result is SUCCEEDED / DEFERRED, never NO_DATA. Partial stored rows remain SUCCEEDED / LIMITED; fatal runtime or storage failures before a stored row are FAILED / FAILED. A confirmed first-party denial or challenge stops requests immediately without retry or route changes, discards any unpersisted records, and reports SUCCEEDED / SKIPPED when no row was stored (LIMITED if a valid prefix had already been stored). Diagnostics, fixtures, dummy records, incomplete candidates, and candidates without matching source receipts are never written as Dataset rows.