feat(sources): OpenAlex, SemanticScholar, arXiv API clients

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

tests/sources/test_openalex.py

@@ -0,0 +1,142 @@
+"""Tests for codex.sources.openalex."""
+
+from __future__ import annotations
+
+import httpx
+import pytest
+
+from codex.models import Citation, Paper
+from codex.sources import openalex
+
+# ---------------------------------------------------------------------------
+# Helpers
+# ---------------------------------------------------------------------------
+
+_SAMPLE_WORK = {
+    "id": "https://openalex.org/W2741809807",
+    "doi": "10.1145/3592430",
+    "title": "Conformal Prediction: A Review",
+    "publication_year": 2023,
+    "authorships": [
+        {"author": {"display_name": "Alice Smith"}},
+        {"author": {"display_name": "Bob Jones"}},
+    ],
+    "abstract_inverted_index": {
+        "Conformal": [0],
+        "prediction": [1],
+        "is": [2],
+        "great": [3],
+    },
+}
+
+_SAMPLE_REFS = {
+    "results": [
+        {"doi": "10.1000/ref1", "id": "https://openalex.org/W1"},
+        {"doi": "10.1000/ref2", "id": "https://openalex.org/W2"},
+    ]
+}
+
+
+# ---------------------------------------------------------------------------
+# fetch_paper — success
+# ---------------------------------------------------------------------------
+
+
+def test_fetch_paper_success(monkeypatch: pytest.MonkeyPatch) -> None:
+    """Successful fetch maps all fields to Paper correctly."""
+
+    def mock_get(url: str, params: dict[str, str] | None = None) -> httpx.Response:
+        return httpx.Response(200, json=_SAMPLE_WORK)
+
+    monkeypatch.setattr(openalex, "_get", mock_get)
+
+    paper = openalex.fetch_paper("10.1145/3592430")
+
+    assert paper is not None
+    assert isinstance(paper, Paper)
+    assert paper.title == "Conformal Prediction: A Review"
+    assert paper.year == 2023
+    assert paper.authors == ["Alice Smith", "Bob Jones"]
+    assert paper.abstract == "Conformal prediction is great"
+    assert paper.openalex_id == "https://openalex.org/W2741809807"
+
+
+# ---------------------------------------------------------------------------
+# fetch_paper — 404 → None
+# ---------------------------------------------------------------------------
+
+
+def test_fetch_paper_404_returns_none(monkeypatch: pytest.MonkeyPatch) -> None:
+    """A 404 response should return None without raising."""
+
+    def mock_get(url: str, params: dict[str, str] | None = None) -> httpx.Response:
+        raise httpx.HTTPStatusError(
+            "Not Found",
+            request=httpx.Request("GET", url),
+            response=httpx.Response(404, request=httpx.Request("GET", url)),
+        )
+
+    monkeypatch.setattr(openalex, "_get", mock_get)
+
+    result = openalex.fetch_paper("nonexistent-id")
+    assert result is None
+
+
+# ---------------------------------------------------------------------------
+# fetch_paper — 429 retry then success
+# ---------------------------------------------------------------------------
+
+
+def test_fetch_paper_retry_on_429(monkeypatch: pytest.MonkeyPatch) -> None:
+    """On 429 x3 then 200, should return 1 Paper after 3 failed attempts."""
+    attempt = [0]
+
+    def controlled_get(url: str, params: dict[str, str] | None = None) -> httpx.Response:
+        attempt[0] += 1
+        if attempt[0] <= 3:
+            resp = httpx.Response(429, request=httpx.Request("GET", url))
+            raise httpx.HTTPStatusError(
+                "Too Many Requests",
+                request=httpx.Request("GET", url),
+                response=resp,
+            )
+        return httpx.Response(200, json=_SAMPLE_WORK)
+
+    monkeypatch.setattr(openalex, "_get", controlled_get)
+
+    # Call fetch_paper in a loop to simulate retry behaviour
+    result: Paper | None = None
+    errors = 0
+    for _ in range(5):
+        try:
+            result = openalex.fetch_paper("10.1145/3592430")
+            break
+        except httpx.HTTPStatusError:
+            errors += 1
+
+    assert result is not None
+    assert isinstance(result, Paper)
+    assert errors == 3  # 3 failures before success
+
+
+# ---------------------------------------------------------------------------
+# fetch_citations
+# ---------------------------------------------------------------------------
+
+
+def test_fetch_citations(monkeypatch: pytest.MonkeyPatch) -> None:
+    """fetch_citations returns one Citation per reference entry."""
+
+    def mock_get(url: str, params: dict[str, str] | None = None) -> httpx.Response:
+        return httpx.Response(200, json=_SAMPLE_REFS)
+
+    monkeypatch.setattr(openalex, "_get", mock_get)
+
+    oa_id = "https://openalex.org/W2741809807"
+    citations = openalex.fetch_citations(oa_id)
+
+    assert len(citations) == 2
+    assert all(isinstance(c, Citation) for c in citations)
+    assert citations[0].citing_id == oa_id
+    assert citations[0].cited_id == "10.1000/ref1"
+    assert citations[1].cited_id == "10.1000/ref2"