feat(tests): scaffold + integration smoke tests (config, db, models, pipeline)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

tests/integration/test_pipeline_smoke.py

@@ -0,0 +1,100 @@
+"""End-to-end smoke test: ingest → discover → provenance flow (all mocked)."""
+
+from __future__ import annotations
+
+from contextlib import contextmanager
+from typing import Any
+from unittest.mock import MagicMock, patch
+
+import numpy as np
+
+from codex.discover import discovery_leads
+from codex.ingest import ingest_paper
+from codex.models import Citation, Paper
+from codex.provenance import export_bib
+
+
+def _make_conn_cm(mock_conn: MagicMock) -> Any:
+    """Return a context-manager factory that yields *mock_conn*."""
+
+    @contextmanager
+    def _cm() -> Any:
+        yield mock_conn
+
+    return _cm
+
+
+def test_full_pipeline_smoke() -> None:
+    """ingest_paper → discovery_leads → export_bib all execute without errors (all mocked)."""
+    paper = Paper(id="2301.07041", title="T", openalex_id="W123")
+    citations = [Citation(citing_id="2301.07041", cited_id="2301.99999")]
+
+    # --- mock connection setup ---
+    mock_conn = MagicMock()
+    mock_conn.execute = MagicMock()
+    mock_conn.commit = MagicMock()
+
+    # discovery_leads: conn.execute(...).fetchall() → empty list (no leads)
+    discover_execute_result = MagicMock()
+    discover_execute_result.fetchall.return_value = []
+
+    # export_bib: conn.execute(...).fetchall() → one paper row with bibkey
+    bib_row: dict[str, Any] = {
+        "id": "2301.07041",
+        "bibkey": "paper2023",
+        "title": "T",
+        "authors": ["Author One"],
+        "year": 2023,
+    }
+    bib_execute_result = MagicMock()
+    bib_execute_result.fetchall.return_value = [bib_row]
+
+    # The ingest path calls conn.execute multiple times (paper upsert, etc.).
+    # discovery_leads and export_bib each call conn.execute once and use fetchall.
+    # We use side_effect to route calls: first N calls from ingest return a generic
+    # mock; then we supply discover and bib results in order.
+    _call_count: list[int] = [0]
+
+    def _execute_side_effect(*args: Any, **kwargs: Any) -> MagicMock:
+        _call_count[0] += 1
+        sql: str = str(args[0]) if args else ""
+        if "cited_id NOT IN" in sql:
+            return discover_execute_result
+        if "FROM papers" in sql and "bibkey IS NOT NULL" in sql:
+            return bib_execute_result
+        return MagicMock()
+
+    mock_conn.execute.side_effect = _execute_side_effect
+
+    # cursor() context manager for executemany calls inside ingest
+    mock_cursor = MagicMock()
+    mock_conn.cursor.return_value.__enter__ = MagicMock(return_value=mock_cursor)
+    mock_conn.cursor.return_value.__exit__ = MagicMock(return_value=False)
+
+    embedder = MagicMock()
+    embedder.dim = 1024
+    embedder.encode_dense.return_value = np.zeros((1, 1024), dtype=np.float32)
+
+    conn_cm = _make_conn_cm(mock_conn)
+
+    with (
+        patch("codex.ingest.openalex.fetch_paper", return_value=paper),
+        patch("codex.ingest.openalex.fetch_citations", return_value=citations),
+        patch("codex.ingest.get_embedder", return_value=embedder),
+        patch("codex.ingest.get_conn", side_effect=conn_cm),
+        patch("codex.discover.get_conn", side_effect=conn_cm),
+        patch("codex.provenance.get_conn", side_effect=conn_cm),
+    ):
+        # Step 1: ingest
+        result = ingest_paper("2301.07041")
+
+        # Step 2: discover
+        leads = discovery_leads()
+
+        # Step 3: export bib
+        bib = export_bib(["2301.07041"])
+
+    # --- assertions ---
+    assert result.paper_id == "2301.07041"
+    assert isinstance(leads, list)
+    assert "@article" in bib