style: auto-fix lint and format

Author: github-actions[bot]

autosearch/v2/judge.py

@@ -77,8 +77,14 @@ def score_results(
     now = now or dt.datetime.now(dt.timezone.utc)
     total_results = len(results)
     unique_urls = {item.get("url") for item in results if item.get("url")}
-    queries = {str(item.get("query", "")).strip() for item in results if str(item.get("query", "")).strip()}
-    platforms = [str(item.get("source", "")).lower() for item in results if item.get("source")]
+    queries = {
+        str(item.get("query", "")).strip()
+        for item in results
+        if str(item.get("query", "")).strip()
+    }
+    platforms = [
+        str(item.get("source", "")).lower() for item in results if item.get("source")
+    ]
     counts = Counter(platforms)
     query_words = {word.lower() for query in queries for word in WORD_RE.findall(query)}
 
@@ -92,7 +98,10 @@ def score_results(
     match_count = 0
     for item in results:
         haystack_words = {
-            word.lower() for word in WORD_RE.findall(f"{item.get('title', '')} {item.get('snippet', '')}")
+            word.lower()
+            for word in WORD_RE.findall(
+                f"{item.get('title', '')} {item.get('snippet', '')}"
+            )
         }
         if query_words and haystack_words.intersection(query_words):
             match_count += 1
@@ -121,13 +130,16 @@ def score_results(
     }
     weight_sum = sum(float(weights.get(name, 0.0)) for name in dimensions)
     total = (
-        sum(dimensions[name] * float(weights.get(name, 0.0)) for name in dimensions) / weight_sum
+        sum(dimensions[name] * float(weights.get(name, 0.0)) for name in dimensions)
+        / weight_sum
         if weight_sum > 0
         else 0.0
     )
     return {
         "total": max(0.0, min(total, 1.0)),
-        "dimensions": {name: max(0.0, min(value, 1.0)) for name, value in dimensions.items()},
+        "dimensions": {
+            name: max(0.0, min(value, 1.0)) for name, value in dimensions.items()
+        },
         "meta": {
             "total_results": total_results,
             "unique_urls": len(unique_urls),
@@ -144,7 +156,9 @@ def main() -> int:
     try:
         weights = json.loads(args.weights) if args.weights else load_default_weights()
         results = load_results(Path(args.evidence_file))
-        payload = score_results(results, args.evidence_file, target=args.target, weights=weights)
+        payload = score_results(
+            results, args.evidence_file, target=args.target, weights=weights
+        )
     except Exception as exc:
         print(str(exc), file=sys.stderr)
         return 1

autosearch/v2/tests/test_judge.py

@@ -14,13 +14,23 @@
 
 def write_jsonl(path: Path, rows: list[object]) -> Path:
     path.write_text(
-        "".join(json.dumps(row) + "\n" if isinstance(row, dict) else f"{row}\n" for row in rows),
+        "".join(
+            json.dumps(row) + "\n" if isinstance(row, dict) else f"{row}\n"
+            for row in rows
+        ),
         encoding="utf-8",
     )
     return path
 
 
-def result(url: str, source: str, query: str, title: str = "", snippet: str = "", metadata: dict | None = None) -> dict:
+def result(
+    url: str,
+    source: str,
+    query: str,
+    title: str = "",
+    snippet: str = "",
+    metadata: dict | None = None,
+) -> dict:
     return {
         "url": url,
         "title": title,
@@ -33,7 +43,9 @@ def result(url: str, source: str, query: str, title: str = "", snippet: str = ""
 
 
 def score_file(path: Path, target: int = 30) -> dict:
-    return judge.score_results(judge.load_results(path), str(path), target=target, now=NOW)
+    return judge.score_results(
+        judge.load_results(path), str(path), target=target, now=NOW
+    )
 
 
 def test_empty_input(tmp_path: Path) -> None:
@@ -43,50 +55,91 @@ def test_empty_input(tmp_path: Path) -> None:
 
 
 def test_single_result(tmp_path: Path) -> None:
-    row = result("https://example.com/1", "github", "python judge", title="Python judge")
+    row = result(
+        "https://example.com/1", "github", "python judge", title="Python judge"
+    )
     score = score_file(write_jsonl(tmp_path / "single.jsonl", [row]))
     assert score["dimensions"]["quantity"] == pytest.approx(1 / 30)
     assert score["dimensions"]["diversity"] == 0.0
     assert score["dimensions"]["relevance"] == 1.0
 
 
 def test_multi_platform_diversity(tmp_path: Path) -> None:
-    rows = [result(f"https://example.com/{idx}", source, "topic", title="topic") for idx, source in enumerate(
-        ["github"] * 4 + ["reddit"] * 3 + ["hackernews"] * 3,
-        start=1,
-    )]
+    rows = [
+        result(f"https://example.com/{idx}", source, "topic", title="topic")
+        for idx, source in enumerate(
+            ["github"] * 4 + ["reddit"] * 3 + ["hackernews"] * 3,
+            start=1,
+        )
+    ]
     score = score_file(write_jsonl(tmp_path / "diverse.jsonl", rows))
     assert score["dimensions"]["diversity"] > 0.5
 
 
 def test_single_platform_low_diversity(tmp_path: Path) -> None:
-    rows = [result(f"https://example.com/{idx}", "github", "topic", title="topic") for idx in range(10)]
+    rows = [
+        result(f"https://example.com/{idx}", "github", "topic", title="topic")
+        for idx in range(10)
+    ]
     score = score_file(write_jsonl(tmp_path / "single-platform.jsonl", rows))
     assert score["dimensions"]["diversity"] == 0.0
 
 
 def test_relevance_all_match(tmp_path: Path) -> None:
-    rows = [result(f"https://example.com/{idx}", "github", "agent search", title="Agent search guide") for idx in range(3)]
+    rows = [
+        result(
+            f"https://example.com/{idx}",
+            "github",
+            "agent search",
+            title="Agent search guide",
+        )
+        for idx in range(3)
+    ]
     score = score_file(write_jsonl(tmp_path / "all-match.jsonl", rows))
     assert score["dimensions"]["relevance"] == 1.0
 
 
 def test_relevance_none_match(tmp_path: Path) -> None:
-    rows = [result(f"https://example.com/{idx}", "github", "agent search", title="database tuning", snippet="sql indexes") for idx in range(3)]
+    rows = [
+        result(
+            f"https://example.com/{idx}",
+            "github",
+            "agent search",
+            title="database tuning",
+            snippet="sql indexes",
+        )
+        for idx in range(3)
+    ]
     score = score_file(write_jsonl(tmp_path / "none-match.jsonl", rows))
     assert score["dimensions"]["relevance"] == 0.0
 
 
 def test_freshness_all_recent(tmp_path: Path) -> None:
     recent = (NOW - dt.timedelta(days=30)).isoformat().replace("+00:00", "Z")
-    rows = [result(f"https://example.com/{idx}", "reddit", "topic", metadata={"published_at": recent}) for idx in range(3)]
+    rows = [
+        result(
+            f"https://example.com/{idx}",
+            "reddit",
+            "topic",
+            metadata={"published_at": recent},
+        )
+        for idx in range(3)
+    ]
     score = score_file(write_jsonl(tmp_path / "fresh.jsonl", rows))
     assert score["dimensions"]["freshness"] == 1.0
 
 
 def test_freshness_all_old(tmp_path: Path) -> None:
     old = (NOW - dt.timedelta(days=300)).isoformat().replace("+00:00", "Z")
-    rows = [result(f"https://example.com/{idx}", "reddit", "topic", metadata={"updated_at": old}) for idx in range(3)]
+    rows = [
+        result(
+            f"https://example.com/{idx}",
+            "reddit",
+            "topic",
+            metadata={"updated_at": old},
+        )
+        for idx in range(3)
+    ]
     score = score_file(write_jsonl(tmp_path / "old.jsonl", rows))
     assert score["dimensions"]["freshness"] == 0.0
 
@@ -95,15 +148,25 @@ def test_freshness_unix_timestamp(tmp_path: Path) -> None:
     recent_ts = int((NOW - dt.timedelta(days=30)).timestamp())
     old_ts = int((NOW - dt.timedelta(days=300)).timestamp())
     rows = [
-        result("https://example.com/1", "reddit", "topic", metadata={"created_utc": recent_ts}),
-        result("https://example.com/2", "reddit", "topic", metadata={"created_utc": old_ts}),
+        result(
+            "https://example.com/1",
+            "reddit",
+            "topic",
+            metadata={"created_utc": recent_ts},
+        ),
+        result(
+            "https://example.com/2", "reddit", "topic", metadata={"created_utc": old_ts}
+        ),
     ]
     score = score_file(write_jsonl(tmp_path / "unix.jsonl", rows))
     assert score["dimensions"]["freshness"] == 0.5
 
 
 def test_custom_target(tmp_path: Path) -> None:
-    rows = [result(f"https://example.com/{idx}", "github", f"query {idx}", title="query") for idx in range(10)]
+    rows = [
+        result(f"https://example.com/{idx}", "github", f"query {idx}", title="query")
+        for idx in range(10)
+    ]
     score = score_file(write_jsonl(tmp_path / "target.jsonl", rows), target=10)
     assert score["dimensions"]["quantity"] == 1.0
 
@@ -120,10 +183,19 @@ def test_invalid_json_lines(tmp_path: Path) -> None:
 
 
 def test_cli_exit_codes(tmp_path: Path) -> None:
-    evidence = write_jsonl(tmp_path / "cli.jsonl", [result("https://example.com/1", "github", "judge", title="judge")])
+    evidence = write_jsonl(
+        tmp_path / "cli.jsonl",
+        [result("https://example.com/1", "github", "judge", title="judge")],
+    )
     script = Path(judge.__file__)
-    success = subprocess.run([sys.executable, str(script), str(evidence)], capture_output=True, text=True)
-    missing = subprocess.run([sys.executable, str(script), str(tmp_path / "missing.jsonl")], capture_output=True, text=True)
+    success = subprocess.run(
+        [sys.executable, str(script), str(evidence)], capture_output=True, text=True
+    )
+    missing = subprocess.run(
+        [sys.executable, str(script), str(tmp_path / "missing.jsonl")],
+        capture_output=True,
+        text=True,
+    )
     assert success.returncode == 0
     assert json.loads(success.stdout)["meta"]["total_results"] == 1
     assert missing.returncode == 1