scaffold

jobsource/sources/__init__.py

@@ -0,0 +1 @@
+"""Job source provider package."""

jobsource/sources/apify_source.py

@@ -0,0 +1,8 @@
+"""Apify ingestion provider (alternative, paid) — implements JobSource.
+
+Scaffold stub -- not implemented yet.
+"""
+# TODO (Stage 1): implement ApifySource per CLAUDE.md "Stage 1 — Ingest".
+# Drop-in alternative to JobSpySource; same JobSource interface.
+# Uses apify-client; actor ID from config (APIFY_ACTOR env var).
+# Map Apify actor output fields → RawJob; same dedup key (LinkedIn jobPostingId).

jobsource/sources/base.py

@@ -0,0 +1,16 @@
+"""JobSource interface: every ingestion provider must implement fetch_recent_jobs().
+
+Scaffold stub -- not implemented yet.
+"""
+# TODO (Stage 1): define the JobSource ABC per CLAUDE.md "Stage 1 — Ingest (deterministic)".
+# Interface:
+#   class JobSource(ABC):
+#       @abstractmethod
+#       def fetch_recent_jobs(
+#           self,
+#           search_terms: list[str],
+#           location: str,
+#           hours_old: int,
+#           results_wanted: int,
+#       ) -> list[RawJob]: ...
+# Implementations: jobspy_source.JobSpySource, apify_source.ApifySource.

jobsource/sources/jobspy_source.py

@@ -0,0 +1,10 @@
+"""JobSpy ingestion provider (default, free) — implements JobSource.
+
+Scaffold stub -- not implemented yet.
+"""
+# TODO (Stage 1): implement JobSpySource per CLAUDE.md "Stage 1 — Ingest".
+# Uses python-jobspy (python_jobspy). Key notes:
+#   - Search LinkedIn via JobSpy; parse LinkedIn numeric jobPostingId from the job URL.
+#   - Map JobSpy result fields → RawJob (company, website from company_url_direct if present).
+#   - Strip tracking query params from linkedin_url; keep only /jobs/view/{id}.
+#   - Log observed fill rate of company_url_direct (see CLAUDE.md Gotchas).