fix: preserve identifiers in Apec dedupe

src/job_research/apec/dedupe.py

@@ -8,12 +8,16 @@ def dedupe_apec_listings(listings: list[ApecListing]) -> list[ApecListing]:
 
     for listing in listings:
         source_job_id = listing.source_job_id
-        if listing.url in seen_urls or (source_job_id and source_job_id in seen_source_job_ids):
-            continue
+        is_duplicate = listing.url in seen_urls or (
+            source_job_id is not None and source_job_id in seen_source_job_ids
+        )
 
         seen_urls.add(listing.url)
-        if source_job_id:
+        if source_job_id is not None:
             seen_source_job_ids.add(source_job_id)
+
+        if is_duplicate:
+            continue
         deduped.append(listing)
 
     return deduped

tests/apec/test_dedupe.py

@@ -55,3 +55,28 @@ def test_dedupe_apec_listings_collapses_mixed_key_duplicates() -> None:
     deduped = dedupe_apec_listings([first, second])
 
     assert deduped == [first]
+
+
+def test_dedupe_apec_listings_keeps_secondary_ids_from_skipped_rows() -> None:
+    first = ApecListing(
+        source="apec",
+        source_job_id=None,
+        url="url1",
+        fetched_at="2026-06-01T10:00:00Z",
+    )
+    second = ApecListing(
+        source="apec",
+        source_job_id="job-123",
+        url="url1",
+        fetched_at="2026-06-01T10:01:00Z",
+    )
+    third = ApecListing(
+        source="apec",
+        source_job_id="job-123",
+        url="url2",
+        fetched_at="2026-06-01T10:02:00Z",
+    )
+
+    deduped = dedupe_apec_listings([first, second, third])
+
+    assert deduped == [first]