fix: preserve preconfigured logging

packages/data-designer-config/src/data_designer/logging.py

@@ -12,6 +12,8 @@
 
 from pythonjsonlogger.json import JsonFormatter
 
+_logging_configured = False
+
 
 @dataclass
 class LoggerConfig:
@@ -115,6 +117,7 @@ def working() -> str:
 
 
 def configure_logging(config: LoggingConfig | None = None) -> None:
+    global _logging_configured
     config = config or LoggingConfig.default()
 
     root_logger = logging.getLogger()
@@ -137,6 +140,18 @@ def configure_logging(config: LoggingConfig | None = None) -> None:
     for name in config.to_silence:
         quiet_noisy_logger(name)
 
+    _logging_configured = True
+
+
+def is_logging_configured() -> bool:
+    """Return whether ``configure_logging()`` has run in this process.
+
+    Only tracks Data Designer's own ``configure_logging()``. Logging configured
+    through stdlib APIs (e.g. ``logging.basicConfig()`` or manually attached
+    root handlers) is not detected.
+    """
+    return _logging_configured
+
 
 def quiet_noisy_logger(name: str) -> None:
     logger = logging.getLogger(name)

packages/data-designer-config/tests/test_logging.py

@@ -8,12 +8,14 @@
 
 import pytest
 
+import data_designer.logging as logging_mod
 from data_designer.logging import (
     LoggerConfig,
     LoggingConfig,
     OutputConfig,
     RandomEmoji,
     configure_logging,
+    is_logging_configured,
     quiet_noisy_logger,
 )
 
@@ -105,6 +107,16 @@ def test_configure_logging_basic(stub_default_logging_config):
     assert ndd_logger.level == logging.INFO
 
 
+def test_configure_logging_marks_logging_configured(stub_default_logging_config, monkeypatch: pytest.MonkeyPatch):
+    monkeypatch.setattr(logging_mod, "_logging_configured", False)
+
+    assert is_logging_configured() is False
+
+    configure_logging(stub_default_logging_config)
+
+    assert is_logging_configured() is True
+
+
 def test_configure_logging_with_file():
     with tempfile.NamedTemporaryFile(mode="w", delete=False, suffix=".log") as tmp_file:
         tmp_path = Path(tmp_file.name)

packages/data-designer/src/data_designer/interface/data_designer.py

@@ -73,7 +73,7 @@
     DataDesignerProfilingError,
 )
 from data_designer.interface.results import DatasetCreationResults
-from data_designer.logging import LOG_INDENT, RandomEmoji, configure_logging
+from data_designer.logging import LOG_INDENT, RandomEmoji, configure_logging, is_logging_configured
 from data_designer.plugins.plugin import PluginType
 from data_designer.plugins.registry import PluginRegistry
 
@@ -88,12 +88,13 @@
 _interface_runtime_initialized = False
 
 
-def _initialize_interface_runtime() -> None:
+def _initialize_interface_runtime(*, auto_configure_logging: bool = True) -> None:
     """Run one-time runtime initialization for the interface package."""
     global _interface_runtime_initialized
     if _interface_runtime_initialized:
         return
-    configure_logging()
+    if auto_configure_logging and not is_logging_configured():
+        configure_logging()
     resolve_seed_default_model_settings()
     _interface_runtime_initialized = True
 
@@ -141,6 +142,12 @@ class DataDesigner(DataDesignerInterface[DatasetCreationResults]):
         mcp_providers: Optional list of MCP provider configurations to enable tool-calling for
             LLM generation columns. Supports both MCPProvider (remote SSE or Streamable HTTP) and
             LocalStdioMCPProvider (local subprocess).
+        auto_configure_logging: Whether to apply Data Designer's default logging
+            configuration during construction. Set to False when embedding Data
+            Designer in an application that manages its own logging — Data Designer
+            will then leave existing root handlers and log levels untouched.
+            Automatic configuration is also skipped when
+            `data_designer.logging.configure_logging()` was already called.
     """
 
     def __init__(
@@ -153,8 +160,9 @@ def __init__(
         managed_assets_path: Path | str | None = None,
         person_reader: PersonReader | None = None,
         mcp_providers: list[MCPProviderT] | None = None,
+        auto_configure_logging: bool = True,
     ):
-        _initialize_interface_runtime()
+        _initialize_interface_runtime(auto_configure_logging=auto_configure_logging)
         self._secret_resolver = secret_resolver or DEFAULT_SECRET_RESOLVER
         self._artifact_path = Path(artifact_path) if artifact_path is not None else Path.cwd() / "artifacts"
         self._run_config = RunConfig()

packages/data-designer/tests/interface/test_data_designer.py

@@ -16,6 +16,7 @@
 
 import data_designer.interface.data_designer as dd_mod
 import data_designer.lazy_heavy_imports as lazy
+import data_designer.logging as dd_logging
 from data_designer.config.column_configs import (
     ExpressionColumnConfig,
     LLMTextColumnConfig,
@@ -1517,6 +1518,7 @@ def test_validate_raises_error_when_seed_collides(
 def test_initialize_interface_runtime_runs_once(monkeypatch: pytest.MonkeyPatch) -> None:
     """_initialize_interface_runtime only runs initialization once."""
     monkeypatch.setattr(dd_mod, "_interface_runtime_initialized", False)
+    monkeypatch.setattr(dd_logging, "_logging_configured", False)
 
     with (
         patch("data_designer.interface.data_designer.configure_logging") as mock_logging,
@@ -1528,6 +1530,94 @@ def test_initialize_interface_runtime_runs_once(monkeypatch: pytest.MonkeyPatch)
         mock_resolve.assert_called_once()
 
 
+def test_initialize_interface_runtime_respects_preconfigured_logging(monkeypatch: pytest.MonkeyPatch) -> None:
+    monkeypatch.setattr(dd_mod, "_interface_runtime_initialized", False)
+    monkeypatch.setattr(dd_logging, "_logging_configured", True)
+
+    with (
+        patch("data_designer.interface.data_designer.configure_logging") as mock_logging,
+        patch("data_designer.interface.data_designer.resolve_seed_default_model_settings") as mock_resolve,
+    ):
+        dd_mod._initialize_interface_runtime()
+
+        mock_logging.assert_not_called()
+        mock_resolve.assert_called_once()
+
+
+def test_initialize_interface_runtime_skips_logging_when_auto_configure_disabled(
+    monkeypatch: pytest.MonkeyPatch,
+) -> None:
+    monkeypatch.setattr(dd_mod, "_interface_runtime_initialized", False)
+    monkeypatch.setattr(dd_logging, "_logging_configured", False)
+
+    with (
+        patch("data_designer.interface.data_designer.configure_logging") as mock_logging,
+        patch("data_designer.interface.data_designer.resolve_seed_default_model_settings") as mock_resolve,
+    ):
+        dd_mod._initialize_interface_runtime(auto_configure_logging=False)
+
+        mock_logging.assert_not_called()
+        mock_resolve.assert_called_once()
+
+
+def test_init_auto_configures_logging_by_default(
+    monkeypatch: pytest.MonkeyPatch,
+    stub_artifact_path: Path,
+    stub_model_providers: list[ModelProvider],
+) -> None:
+    monkeypatch.setattr(dd_mod, "_interface_runtime_initialized", False)
+    monkeypatch.setattr(dd_logging, "_logging_configured", False)
+
+    with patch("data_designer.interface.data_designer.configure_logging") as mock_logging:
+        DataDesigner(artifact_path=stub_artifact_path, model_providers=stub_model_providers)
+
+        mock_logging.assert_called_once()
+
+
+def test_init_auto_configure_logging_false_preserves_root_handlers(
+    monkeypatch: pytest.MonkeyPatch,
+    stub_artifact_path: Path,
+    stub_model_providers: list[ModelProvider],
+) -> None:
+    monkeypatch.setattr(dd_mod, "_interface_runtime_initialized", False)
+    monkeypatch.setattr(dd_logging, "_logging_configured", False)
+    root_logger = logging.getLogger()
+    sentinel_handler = logging.NullHandler()
+    root_logger.addHandler(sentinel_handler)
+
+    try:
+        DataDesigner(
+            artifact_path=stub_artifact_path,
+            model_providers=stub_model_providers,
+            auto_configure_logging=False,
+        )
+
+        assert sentinel_handler in root_logger.handlers
+        assert dd_logging.is_logging_configured() is False
+    finally:
+        root_logger.removeHandler(sentinel_handler)
+
+
+def test_init_preserves_preconfigured_logging(
+    monkeypatch: pytest.MonkeyPatch,
+    stub_artifact_path: Path,
+    stub_model_providers: list[ModelProvider],
+) -> None:
+    monkeypatch.setattr(dd_mod, "_interface_runtime_initialized", False)
+    monkeypatch.setattr(dd_logging, "_logging_configured", False)
+    data_designer_logger = logging.getLogger("data_designer")
+    previous_level = data_designer_logger.level
+
+    try:
+        dd_logging.configure_logging(dd_logging.LoggingConfig.debug())
+
+        DataDesigner(artifact_path=stub_artifact_path, model_providers=stub_model_providers)
+
+        assert data_designer_logger.level == logging.DEBUG
+    finally:
+        data_designer_logger.setLevel(previous_level)
+
+
 def test_create_dataset_e2e_with_directory_seed_source(
     stub_artifact_path: Path,
     stub_model_providers: list[ModelProvider],