[MCC-1476022] - Add Dry Publish Function

dataconnect/__init__.py

@@ -11,7 +11,7 @@
     ServerError,
     ValidationError,
 )
-from dataconnect.models import DatasetVersion, PaginatedResponse, Pagination, Study, StudyEnvironment
+from dataconnect.models import DatasetVersion, DryPublishResult, PaginatedResponse, Pagination, Study, StudyEnvironment
 
 __all__ = [
     # Client
@@ -20,6 +20,7 @@
     "Study",
     "StudyEnvironment",
     "DatasetVersion",
+    "DryPublishResult",
     "PaginatedResponse",
     "Pagination",
     # Exceptions — catch these in user application code

dataconnect/client.py

@@ -12,7 +12,7 @@
 
 import pandas as pd
 
-from dataconnect.models import Dataset, DatasetVersion, PaginatedResponse, StudiesResult
+from dataconnect.models import Dataset, DatasetVersion, DryPublishResult, PaginatedResponse, StudiesResult
 from dataconnect.service import DataConnectService, DefaultDataConnectService
 
 _DEFAULT_HOST = "enodia-gateway.platform.imedidata.com"
@@ -87,6 +87,25 @@ def get_datasets(
             page_size=page_size,
         )
 
+    def dry_publish(
+        self,
+        project_token: str,
+        dataset_name: str,
+        key_columns: list[str],
+        source_datasets: list[UUID],
+        data: pd.DataFrame,
+        datetime_formats: dict[str, str] | None = None,
+    ) -> DryPublishResult:
+
+        return self._service.dry_publish(
+            project_token=project_token,
+            dataset_name=dataset_name,
+            key_columns=key_columns,
+            source_datasets=source_datasets,
+            data=data,
+            datetime_formats=datetime_formats,
+        )
+
     # Lifecycle
 
     def close(self) -> None:

dataconnect/models.py

@@ -4,6 +4,8 @@
 from typing import Generic, TypeVar
 from uuid import UUID
 
+import pandas as pd
+
 T = TypeVar("T")
 
 
@@ -61,3 +63,22 @@ class PaginatedResponse(Generic[T]):  # noqa: UP046
     total_records: int
     pagination: Pagination
     items: list[T]
+
+
+@dataclass
+class DryPublishResult:
+    """Result of a dry publish operation, including validation status and details."""
+
+    status: bool
+    is_schema_valid: bool | None = None
+    is_config_valid: bool | None = None
+    is_dataset_valid: bool | None = None
+    errors: list[str] = field(default_factory=list)
+    invalid_datetime_formats: dict[str, str] = field(default_factory=dict)
+    dataset_name: str | None = None
+    dataset_version: int | None = None
+    no_of_columns: int | None = None
+    valid_record_count: int | None = None
+    duplicate_record_count: int | None = None
+    invalid_record_count: int | None = None
+    invalid_records: pd.DataFrame | None = None

dataconnect/service/base.py

@@ -7,7 +7,7 @@
 
 import pandas as pd
 
-from dataconnect.models import Dataset, DatasetVersion, PaginatedResponse, StudiesResult
+from dataconnect.models import Dataset, DatasetVersion, DryPublishResult, PaginatedResponse, StudiesResult
 
 
 class DataConnectService(ABC):
@@ -35,5 +35,16 @@ def fetch_data(
         first_n_rows: int | None = None,
     ) -> pd.DataFrame: ...
 
+    @abstractmethod
+    def dry_publish(
+        self,
+        project_token: str,
+        dataset_name: str,
+        key_columns: list[str],
+        source_datasets: list[UUID],
+        data: pd.DataFrame,
+        datetime_formats: dict[str, str] | None = None,
+    ) -> DryPublishResult: ...
+
     @abstractmethod
     def close(self) -> None: ...

dataconnect/service/default.py

@@ -2,22 +2,24 @@
 
 from __future__ import annotations
 
+import json
 from uuid import UUID
 
 import pandas as pd
 
-from dataconnect.models import Dataset, DatasetVersion, PaginatedResponse, Pagination, StudiesResult
+from dataconnect.models import Dataset, DatasetVersion, DryPublishResult, PaginatedResponse, Pagination, StudiesResult
 from dataconnect.service.base import DataConnectService
 from dataconnect.service.error_handler import translate_error
 from dataconnect.service.mappers import (
+    dry_publish_response_to_domain,
     resource_to_dataset,
     resource_to_dataset_version,
     resource_to_fetched_data,
     resource_to_study,
 )
 from dataconnect.transport.base import Transport
 from dataconnect.transport.errors import TransportError
-from dataconnect.transport.models import DatasetTicket, ResourceQuery
+from dataconnect.transport.models import DatasetTicket, PublishRequest, ResourceQuery
 
 # Server action identifiers
 _ACTION_LIST_STUDIES = "studies.list"
@@ -138,6 +140,71 @@ def get_datasets(
         except TransportError as ex:
             raise translate_error(ex) from ex
 
+    def dry_publish(
+        self,
+        project_token: str,
+        dataset_name: str,
+        key_columns: list[str],
+        source_datasets: list[UUID],
+        data: pd.DataFrame,
+        datetime_formats: dict[str, str] | None = None,
+    ) -> DryPublishResult:
+        """Validate a dataset against the server without committing any changes.
+
+        Encodes the publish configuration as a JSON ``input_config`` string,
+        sets ``is_dry_publish: True`` so the server treats the call as a
+        validation-only run, then delegates to the transport layer.
+
+        The server response is mapped to a :class:`DryPublishResult` via
+        :func:`dry_publish_response_to_domain`.
+
+        Args:
+            project_token: Base64-encoded project token identifying the target
+                study, study environment, and project.
+            dataset_name: Name of the dataset to validate.
+            key_columns: Column names that form the unique key for the dataset.
+            source_datasets: UUIDs of the source datasets the published dataset
+                is derived from.
+            data: The dataset to validate as a ``pd.DataFrame``.
+            datetime_formats: Optional mapping of column name → datetime format
+                string (e.g. ``{"visit_date": "yyyy-MM-dd"}``).  Defaults to
+                an empty dict when omitted.
+
+        Returns:
+            A :class:`DryPublishResult` containing the server's validation
+            outcome, including per-field validity flags, error messages, and
+            an optional ``invalid_records`` DataFrame.
+
+        Raises:
+             DataConnectError: Any :class:`TransportError` from the transport
+             layer is translated by :func:`translate_error` into the public
+             API's :class:`DataConnectError` hierarchy before propagating,
+             for example :class:`ValidationError`,
+             :class:`AuthorizationError`, or :class:`ServerError`."""
+
+        request = PublishRequest(
+            input_config=json.dumps(
+                {
+                    "is_dry_publish": True,
+                    "project_token": project_token,
+                    "dataset_name": dataset_name,
+                    "dataset_description": dataset_name,
+                    "key_columns": key_columns,
+                    "source_datasets": [str(uuid) for uuid in source_datasets],
+                    "datetime_formats": datetime_formats or {},
+                }
+            ),
+            data=data,
+        )
+
+        try:
+            publish_result = self._transport.dry_publish_dataset(request)
+
+            return dry_publish_response_to_domain(publish_result)
+
+        except TransportError as ex:
+            raise translate_error(ex) from ex
+
     def close(self) -> None:
         """Close the underlying transport connection."""
 

dataconnect/service/mappers.py

@@ -15,8 +15,8 @@
 import pyarrow as pa
 
 from dataconnect.exceptions import NotFoundError
-from dataconnect.models import Dataset, DatasetVersion, Study, StudyEnvironment
-from dataconnect.transport.models import DataTable, ResourceInfo
+from dataconnect.models import Dataset, DatasetVersion, DryPublishResult, Study, StudyEnvironment
+from dataconnect.transport.models import DataTable, DryPublishResponse, ResourceInfo
 
 
 def resource_to_study(resource: ResourceInfo) -> Study:
@@ -85,3 +85,42 @@ def resource_to_dataset(resource: ResourceInfo) -> Dataset:
         study_env_uuid=data.get("study_env_uuid", ""),
         dataset_name=data.get("dataset_name", ""),
     )
+
+
+def dry_publish_response_to_domain(result: DryPublishResponse | None) -> DryPublishResult:
+    """Map a transport-layer ``DryPublishResponse`` to a ``DryPublishResult`` domain object.
+
+    ``DryPublishResponse`` carries flat, typed fields returned by the server after a
+    dry-publish call.  The mapping is direct for all shared fields with one
+    exception:
+
+    * ``DryPublishResponse.dataset_valid`` → ``DryPublishResult.is_dataset_valid``
+      (renamed for naming consistency with the other ``is_*_valid`` fields).
+
+    Args:
+        result: The transport-layer result returned by
+            :meth:`Transport.dry_publish_dataset`.  Pass ``None`` to obtain a
+            default :class:`DryPublishResult` with ``status=False`` and all
+            other fields at their zero values.
+
+    Returns:
+        A :class:`DryPublishResult` suitable for returning to the caller.
+    """
+    if result is None:
+        return DryPublishResult(status=False)
+
+    return DryPublishResult(
+        status=result.status,
+        is_schema_valid=result.is_schema_valid,
+        is_config_valid=result.is_config_valid,
+        is_dataset_valid=result.dataset_valid,
+        errors=result.errors,
+        invalid_datetime_formats=result.invalid_datetime_formats,
+        dataset_name=result.dataset_name,
+        dataset_version=result.dataset_version,
+        no_of_columns=result.no_of_columns,
+        valid_record_count=result.valid_record_count,
+        duplicate_record_count=result.duplicate_record_count,
+        invalid_record_count=result.invalid_record_count,
+        invalid_records=result.invalid_records,
+    )

dataconnect/transport/__init__.py

@@ -6,11 +6,23 @@
 """
 
 from dataconnect.transport.base import Transport
-from dataconnect.transport.models import DataRef, ResourceInfo, ResourceQuery
+from dataconnect.transport.models import (
+    DataRef,
+    DatasetTicket,
+    DataTable,
+    DryPublishResponse,
+    PublishRequest,
+    ResourceInfo,
+    ResourceQuery,
+)
 
 __all__ = [
     "Transport",
     "ResourceQuery",
     "ResourceInfo",
     "DataRef",
+    "DatasetTicket",
+    "DataTable",
+    "PublishRequest",
+    "DryPublishResponse",
 ]