fix(scanners): review findings — default audio, TCC errno, extension map

src/mac2nix/models/files.py

@@ -30,6 +30,8 @@ class ConfigFileType(StrEnum):
     JSON = "json"
     PLIST = "plist"
     TOML = "toml"
+    YAML = "yaml"
+    XML = "xml"
     CONF = "conf"
     DATABASE = "database"
     UNKNOWN = "unknown"

src/mac2nix/scanners/_utils.py

@@ -122,7 +122,7 @@ def _parse_xml_dict(element: ElementTree.Element) -> dict[str, Any]:
     result: dict[str, Any] = {}
     children = list(element)
     i = 0
-    while i < len(children) - 1:
+    while i + 1 < len(children):
         if children[i].tag == "key":
             key = children[i].text or ""
             value_elem = children[i + 1]
@@ -165,8 +165,11 @@ def read_launchd_plists() -> list[tuple[Path, str, dict[str, Any]]]:
             continue
         try:
             plist_files = sorted(agent_dir.glob("*.plist"))
-        except PermissionError:
-            logger.warning("Permission denied reading: %s", agent_dir)
+        except PermissionError as exc:
+            if exc.errno == errno.EPERM:
+                logger.debug("Skipping TCC-protected directory: %s", agent_dir)
+            else:
+                logger.warning("Permission denied reading: %s", agent_dir)
             continue
         for plist_path in plist_files:
             data = read_plist_safe(plist_path)

src/mac2nix/scanners/app_config.py

@@ -16,6 +16,9 @@
     ".json": ConfigFileType.JSON,
     ".plist": ConfigFileType.PLIST,
     ".toml": ConfigFileType.TOML,
+    ".yaml": ConfigFileType.YAML,
+    ".yml": ConfigFileType.YAML,
+    ".xml": ConfigFileType.XML,
     ".conf": ConfigFileType.CONF,
     ".cfg": ConfigFileType.CONF,
     ".ini": ConfigFileType.CONF,
@@ -24,8 +27,6 @@
     ".sqlite3": ConfigFileType.DATABASE,
 }
 
-_MAX_FILE_SIZE = 10 * 1024 * 1024  # 10MB
-
 
 @register("app_config")
 class AppConfigScanner(BaseScannerPlugin):
@@ -71,11 +72,6 @@ def _scan_app_dir(self, app_dir: Path, entries: list[AppConfigEntry]) -> None:
         for child in children:
             if not child.is_file():
                 continue
-            try:
-                if child.stat().st_size > _MAX_FILE_SIZE:
-                    continue
-            except OSError:
-                continue
 
             ext = child.suffix.lower()
             file_type = _EXTENSION_MAP.get(ext, ConfigFileType.UNKNOWN)

src/mac2nix/scanners/audio.py

@@ -23,8 +23,7 @@ def is_available(self) -> bool:
         return shutil.which("system_profiler") is not None
 
     def scan(self) -> AudioConfig:
-        input_devices, output_devices = self._get_audio_devices()
-        default_input, default_output = self._get_default_devices(input_devices, output_devices)
+        input_devices, output_devices, default_input, default_output = self._get_audio_devices()
         alert_volume = self._get_alert_volume()
 
         return AudioConfig(
@@ -35,19 +34,23 @@ def scan(self) -> AudioConfig:
             alert_volume=alert_volume,
         )
 
-    def _get_audio_devices(self) -> tuple[list[AudioDevice], list[AudioDevice]]:
+    def _get_audio_devices(
+        self,
+    ) -> tuple[list[AudioDevice], list[AudioDevice], str | None, str | None]:
         result = run_command(["system_profiler", "SPAudioDataType", "-json"], timeout=15)
         if result is None or result.returncode != 0:
-            return [], []
+            return [], [], None, None
 
         try:
             data = json.loads(result.stdout)
         except (json.JSONDecodeError, ValueError):
             logger.warning("Failed to parse system_profiler audio output")
-            return [], []
+            return [], [], None, None
 
         input_devices: list[AudioDevice] = []
         output_devices: list[AudioDevice] = []
+        default_input: str | None = None
+        default_output: str | None = None
 
         for item in data.get("SPAudioDataType", []):
             if not isinstance(item, dict):
@@ -63,10 +66,20 @@ def _get_audio_devices(self) -> tuple[list[AudioDevice], list[AudioDevice]]:
                 is_input, is_output = self._classify_device(device_data)
                 if is_input:
                     input_devices.append(device)
+                    if "coreaudio_default_audio_input_device" in device_data:
+                        default_input = name
                 if is_output:
                     output_devices.append(device)
+                    if "coreaudio_default_audio_output_device" in device_data:
+                        default_output = name
+
+        # Fall back to first device if system_profiler didn't mark a default
+        if default_input is None and input_devices:
+            default_input = input_devices[0].name
+        if default_output is None and output_devices:
+            default_output = output_devices[0].name
 
-        return input_devices, output_devices
+        return input_devices, output_devices, default_input, default_output
 
     @staticmethod
     def _classify_device(device_data: dict[str, object]) -> tuple[bool, bool]:
@@ -77,15 +90,6 @@ def _classify_device(device_data: dict[str, object]) -> tuple[bool, bool]:
             is_output = True
         return is_input, is_output
 
-    def _get_default_devices(
-        self,
-        input_devices: list[AudioDevice],
-        output_devices: list[AudioDevice],
-    ) -> tuple[str | None, str | None]:
-        default_input = input_devices[0].name if input_devices else None
-        default_output = output_devices[0].name if output_devices else None
-        return default_input, default_output
-
     def _get_alert_volume(self) -> float | None:
         result = run_command(["osascript", "-e", "alert volume of (get volume settings)"])
         if result is None or result.returncode != 0:

tests/scanners/test_app_config.py

@@ -134,21 +134,32 @@ def test_group_containers(self, tmp_path: Path) -> None:
         assert result.entries[0].app_name == "group.com.example.app"
         assert result.entries[0].file_type == ConfigFileType.JSON
 
-    def test_large_file_skipped(self, tmp_path: Path) -> None:
+    def test_yaml_extension(self, tmp_path: Path) -> None:
         app_support = _setup_app_support(tmp_path)
-        app_dir = app_support / "BigApp"
+        app_dir = app_support / "YamlApp"
         app_dir.mkdir()
-        large_file = app_dir / "huge.json"
-        large_file.write_text("x")
+        (app_dir / "config.yaml").write_text("key: value")
+        (app_dir / "settings.yml").write_text("other: true")
 
-        with (
-            patch("mac2nix.scanners.app_config.Path.home", return_value=tmp_path),
-            patch("mac2nix.scanners.app_config._MAX_FILE_SIZE", 0),
-        ):
+        with patch("mac2nix.scanners.app_config.Path.home", return_value=tmp_path):
             result = AppConfigScanner().scan()
 
         assert isinstance(result, AppConfigResult)
-        assert result.entries == []
+        assert len(result.entries) == 2
+        assert all(e.file_type == ConfigFileType.YAML for e in result.entries)
+
+    def test_xml_extension(self, tmp_path: Path) -> None:
+        app_support = _setup_app_support(tmp_path)
+        app_dir = app_support / "XmlApp"
+        app_dir.mkdir()
+        (app_dir / "config.xml").write_text("<config/>")
+
+        with patch("mac2nix.scanners.app_config.Path.home", return_value=tmp_path):
+            result = AppConfigScanner().scan()
+
+        assert isinstance(result, AppConfigResult)
+        assert len(result.entries) == 1
+        assert result.entries[0].file_type == ConfigFileType.XML
 
     def test_returns_app_config_result(self, tmp_path: Path) -> None:
         with patch("mac2nix.scanners.app_config.Path.home", return_value=tmp_path):

tests/scanners/test_audio.py

@@ -153,6 +153,42 @@ def side_effect(cmd: list[str], **_kwargs: object) -> subprocess.CompletedProces
         assert result.output_devices[0].name == "Mystery Device"
         assert result.input_devices == []
 
+    def test_default_device_uses_explicit_marker(self, cmd_result) -> None:
+        """Default output should use coreaudio_default_audio_output_device, not first-in-list."""
+        audio_json = {
+            "SPAudioDataType": [
+                {
+                    "_name": "GPU",
+                    "_items": [
+                        {
+                            "_name": "HDMI Output",
+                            "coreaudio_device_uid": "hdmi",
+                            "coreaudio_device_output": "yes",
+                        },
+                        {
+                            "_name": "Built-in Speakers",
+                            "coreaudio_device_uid": "builtin",
+                            "coreaudio_device_output": "yes",
+                            "coreaudio_default_audio_output_device": "spaudio_yes",
+                        },
+                    ],
+                }
+            ]
+        }
+
+        def side_effect(cmd: list[str], **_kwargs: object) -> subprocess.CompletedProcess[str] | None:
+            if "SPAudioDataType" in cmd:
+                return cmd_result(json.dumps(audio_json))
+            return None
+
+        with patch("mac2nix.scanners.audio.run_command", side_effect=side_effect):
+            result = AudioScanner().scan()
+
+        assert isinstance(result, AudioConfig)
+        assert len(result.output_devices) == 2
+        # Default should be the explicitly marked device, not the first one
+        assert result.default_output == "Built-in Speakers"
+
     def test_returns_audio_config(self) -> None:
         with patch("mac2nix.scanners.audio.run_command", return_value=None):
             result = AudioScanner().scan()