fix(ops): keep the path of self-hosted LangSmith and Weave endpoints (#40584)

api/core/ops/utils.py

@@ -105,7 +105,13 @@ def validate_url(url: str, default_url: str, allowed_schemes: tuple = ("https",
     return normalized_url
 
 
-def validate_url_with_path(url: str, default_url: str, required_suffix: str | None = None) -> str:
+def validate_url_with_path(
+    url: str,
+    default_url: str,
+    required_suffix: str | None = None,
+    *,
+    allowed_schemes: tuple[str, ...] = ("https", "http"),
+) -> str:
     """
     Validate URL that may include path components
 
@@ -113,22 +119,26 @@ def validate_url_with_path(url: str, default_url: str, required_suffix: str | No
         url: The URL to validate
         default_url: Default URL to use if input is None or empty
         required_suffix: Optional suffix that URL must end with
+        allowed_schemes: Tuple of allowed URL schemes (default: https, http)
 
     Returns:
-        Validated URL string
+        Validated URL string, returned verbatim so path, query and trailing
+        separators survive — `required_suffix` consumers depend on that
 
     Raises:
         ValueError: If URL format is invalid or doesn't match required suffix
     """
     if not url or url.strip() == "":
         return default_url
+    url = url.strip()
 
     # Parse URL to validate format
     parsed = urlparse(url)
 
     # Check if scheme is allowed
-    if parsed.scheme not in ("https", "http"):
-        raise ValueError("URL must start with https:// or http://")
+    if parsed.scheme not in allowed_schemes:
+        expected = " or ".join(f"{scheme}://" for scheme in allowed_schemes)
+        raise ValueError(f"URL must start with {expected}")
 
     # Check required suffix if specified
     if required_suffix and not url.endswith(required_suffix):

api/providers/trace/trace-langsmith/src/dify_trace_langsmith/config.py

@@ -1,7 +1,7 @@
 from pydantic import ValidationInfo, field_validator
 
 from core.ops.entities.config_entity import BaseTracingConfig
-from core.ops.utils import validate_url
+from core.ops.utils import validate_url_with_path
 
 
 class LangSmithConfig(BaseTracingConfig):
@@ -16,5 +16,5 @@ class LangSmithConfig(BaseTracingConfig):
     @field_validator("endpoint")
     @classmethod
     def endpoint_validator(cls, v, info: ValidationInfo):
-        # LangSmith only allows HTTPS
-        return validate_url(v, "https://api.smith.langchain.com", allowed_schemes=("https",))
+        # LangSmith only allows HTTPS; self-hosted deployments may sit under a path prefix
+        return validate_url_with_path(v, "https://api.smith.langchain.com", allowed_schemes=("https",))

api/providers/trace/trace-langsmith/tests/unit_tests/langsmith_trace/test_langsmith_trace.py

@@ -63,6 +63,16 @@ def test_init(langsmith_config, monkeypatch: pytest.MonkeyPatch):
     assert instance.file_base_url == "http://test.url"
 
 
+def test_init_passes_self_hosted_path_to_client(monkeypatch: pytest.MonkeyPatch):
+    config = LangSmithConfig(api_key="ls-123", project="default", endpoint="https://langsmith.internal/api")
+    mock_client_class = MagicMock()
+    monkeypatch.setattr("dify_trace_langsmith.langsmith_trace.Client", mock_client_class)
+
+    LangSmithDataTrace(config)
+
+    mock_client_class.assert_called_once_with(api_key="ls-123", api_url="https://langsmith.internal/api")
+
+
 def test_trace_dispatch(trace_instance, monkeypatch: pytest.MonkeyPatch):
     methods = [
         "workflow_trace",

api/providers/trace/trace-langsmith/tests/unit_tests/test_config_entity.py

@@ -31,5 +31,15 @@ class TestLangSmithConfig:
 
     def test_endpoint_validation_https_only(self):
         """Test endpoint validation only allows HTTPS"""
-        with pytest.raises(ValidationError, match="URL scheme must be one of"):
+        with pytest.raises(ValidationError, match="URL must start with https://"):
             LangSmithConfig(api_key="key", project="project", endpoint="http://insecure.com")
+
+    def test_endpoint_preserves_path(self):
+        """Self-hosted LangSmith endpoints keep their API path prefix"""
+        config = LangSmithConfig(api_key="key", project="project", endpoint="https://langsmith.internal/api")
+        assert config.endpoint == "https://langsmith.internal/api"
+
+    def test_endpoint_preserves_versioned_path(self):
+        """Self-hosted LangSmith endpoints keep multi-segment API paths"""
+        config = LangSmithConfig(api_key="key", project="project", endpoint="https://langsmith.internal/api/v1")
+        assert config.endpoint == "https://langsmith.internal/api/v1"

api/providers/trace/trace-weave/src/dify_trace_weave/config.py

@@ -1,7 +1,7 @@
 from pydantic import ValidationInfo, field_validator
 
 from core.ops.entities.config_entity import BaseTracingConfig
-from core.ops.utils import validate_url
+from core.ops.utils import validate_url_with_path
 
 
 class WeaveConfig(BaseTracingConfig):
@@ -19,11 +19,11 @@ class WeaveConfig(BaseTracingConfig):
     @classmethod
     def endpoint_validator(cls, v, info: ValidationInfo):
         # Weave only allows HTTPS for endpoint
-        return validate_url(v, "https://trace.wandb.ai", allowed_schemes=("https",))
+        return validate_url_with_path(v, "https://trace.wandb.ai", allowed_schemes=("https",))
 
     @field_validator("host")
     @classmethod
     def host_validator(cls, v, info: ValidationInfo):
         if v is not None and v.strip() != "":
-            return validate_url(v, v, allowed_schemes=("https", "http"))
+            return validate_url_with_path(v, v, allowed_schemes=("https", "http"))
         return v

api/providers/trace/trace-weave/tests/unit_tests/test_config_entity.py

@@ -41,7 +41,7 @@ class TestWeaveConfig:
 
     def test_endpoint_validation_https_only(self):
         """Test endpoint validation only allows HTTPS"""
-        with pytest.raises(ValidationError, match="URL scheme must be one of"):
+        with pytest.raises(ValidationError, match="URL must start with https://"):
             WeaveConfig(api_key="key", project="project", endpoint="http://insecure.wandb.ai")
 
     def test_host_validation_optional(self):
@@ -57,5 +57,20 @@ class TestWeaveConfig:
 
     def test_host_validation_invalid_scheme(self):
         """Test host validation rejects invalid schemes when provided"""
-        with pytest.raises(ValidationError, match="URL scheme must be one of"):
+        with pytest.raises(ValidationError, match="URL must start with https:// or http://"):
             WeaveConfig(api_key="key", project="project", host="ftp://invalid.host.com")
+
+    def test_endpoint_preserves_path(self):
+        """Self-hosted Weave endpoints keep their path prefix"""
+        config = WeaveConfig(api_key="key", project="project", endpoint="https://wandb.internal/api")
+        assert config.endpoint == "https://wandb.internal/api"
+
+    def test_host_preserves_path(self):
+        """Self-hosted W&B hosts keep their path prefix"""
+        config = WeaveConfig(api_key="key", project="project", host="https://wandb.internal/wandb")
+        assert config.host == "https://wandb.internal/wandb"
+
+    def test_host_preserves_http_path(self):
+        """Self-hosted W&B hosts may be plain http and keep their path"""
+        config = WeaveConfig(api_key="key", project="project", host="http://wandb.internal/wandb")
+        assert config.host == "http://wandb.internal/wandb"

api/providers/trace/trace-weave/tests/unit_tests/weave_trace/test_weave_trace.py

@@ -251,6 +251,16 @@ class TestInit:
         )
         assert instance.host == "https://my.wandb.host"
 
+    def test_init_with_host_path(self, mock_wandb, mock_weave):
+        """A self-hosted host keeps its path prefix all the way to wandb.login."""
+        config = _make_weave_config(host="https://wandb.internal/api")
+        instance = WeaveDataTrace(config)
+
+        mock_wandb.login.assert_called_once_with(
+            key="wv-api-key", verify=True, relogin=True, host="https://wandb.internal/api"
+        )
+        assert instance.host == "https://wandb.internal/api"
+
     def test_init_without_entity(self, mock_wandb, mock_weave):
         """Test __init__ initializes weave without entity prefix when entity is None."""
         mock_w, weave_client = mock_weave

api/tests/unit_tests/core/ops/test_config_entity.py

@@ -48,11 +48,17 @@ class TestConfigIntegration:
         aliyun_config = AliyunConfig(
             license_key="test_license", endpoint="https://tracing-analysis-dc-hz.aliyuncs.com/api/v1/traces"
         )
+        langsmith_config = LangSmithConfig(
+            api_key="key", project="project", endpoint="https://langsmith.internal/api/v1"
+        )
+        weave_config = WeaveConfig(api_key="key", project="project", endpoint="https://weave.internal/wandb")
 
         assert arize_config.endpoint == "https://arize.com"
         assert phoenix_with_path_config.endpoint == "https://app.phoenix.arize.com/s/dify-integration"
         assert phoenix_without_path_config.endpoint == "https://app.phoenix.arize.com"
         assert aliyun_config.endpoint == "https://tracing-analysis-dc-hz.aliyuncs.com/api/v1/traces"
+        assert langsmith_config.endpoint == "https://langsmith.internal/api/v1"
+        assert weave_config.endpoint == "https://weave.internal/wandb"
 
     def test_project_default_values(self):
         """Test that project default values are set correctly"""

api/tests/unit_tests/core/ops/test_utils.py

@@ -155,6 +155,30 @@ class TestValidateUrlWithPath:
         with pytest.raises(ValueError, match="URL must start with https:// or http://"):
             validate_url_with_path("example.com", "https://default.com")
 
+    def test_restricted_scheme_accepts_allowed_scheme(self):
+        """Test https-only validation keeps the path of an https URL"""
+        result = validate_url_with_path(
+            "https://langsmith.internal/api", "https://default.com", allowed_schemes=("https",)
+        )
+        assert result == "https://langsmith.internal/api"
+
+    def test_restricted_scheme_rejects_http(self):
+        """Test https-only validation rejects http and names only https in the error"""
+        with pytest.raises(ValueError) as excinfo:
+            validate_url_with_path("http://langsmith.internal/api", "https://default.com", allowed_schemes=("https",))
+        assert str(excinfo.value) == "URL must start with https://"
+
+    def test_default_schemes_keep_original_error_message(self):
+        """Test the two-scheme default keeps the exact message existing providers assert on"""
+        with pytest.raises(ValueError) as excinfo:
+            validate_url_with_path("ftp://example.com", "https://default.com")
+        assert str(excinfo.value) == "URL must start with https:// or http://"
+
+    def test_surrounding_whitespace_is_stripped(self):
+        """Test surrounding whitespace is removed while the path is preserved"""
+        result = validate_url_with_path("  https://example.com/api/v1  ", "https://default.com")
+        assert result == "https://example.com/api/v1"
+
 
 class TestValidateProjectName:
     """Test cases for validate_project_name function"""