deepset-ai
diff --git a/‎deepset_cloud_sdk/workflows/__init__.py‎
Lines changed: 21 additions & 1 deletion b/‎deepset_cloud_sdk/workflows/__init__.py‎
Lines changed: 21 additions & 1 deletion
diff --git a/‎deepset_cloud_sdk/workflows/pipeline_client/__init__.py‎
Lines changed: 1 addition & 0 deletions b/‎deepset_cloud_sdk/workflows/pipeline_client/__init__.py‎
Lines changed: 1 addition & 0 deletions
diff --git a/‎deepset_cloud_sdk/workflows/pipeline_client/models.py‎
Lines changed: 156 additions & 0 deletions b/‎deepset_cloud_sdk/workflows/pipeline_client/models.py‎
Lines changed: 156 additions & 0 deletions
@@ -1 +1,21 @@
-"""Workflows for the deepset cloud SDK."""
+"""Workflows for deepset AI platform SDK."""
+
+from deepset_cloud_sdk.workflows.pipeline_client.models import (
+    IndexConfig,
+    IndexInputs,
+    IndexOutputs,
+    PipelineConfig,
+    PipelineInputs,
+    PipelineOutputs,
+)
+from deepset_cloud_sdk.workflows.sdk import DeepsetSDK
+
+__all__ = [
+    "PipelineInputs",
+    "IndexInputs",
+    "IndexOutputs",
+    "PipelineOutputs",
+    "IndexConfig",
+    "PipelineConfig",
+    "DeepsetSDK",
+]
@@ -0,0 +1 @@
+"""Package to enable importing pipelines and indexes to deepset AI platform."""
@@ -0,0 +1,156 @@
+"""Models for the pipeline service."""
+from typing import List
+
+from pydantic import BaseModel, Field, model_validator
+
+
+class InputOutputBaseModel(BaseModel):
+    """Base model for input and output configurations.
+
+    This class provides common functionality for input and output models, such as YAML conversion.
+    """
+
+    def to_yaml_dict(self) -> dict:
+        """Convert the model to a YAML-compatible dictionary.
+
+        Clears empty values from the dictionary.
+
+        :return: Dictionary ready for YAML serialization.
+        """
+        fields = self.model_dump(exclude_none=True)
+        # Remove empty values
+        return {k: v for k, v in fields.items() if v}
+
+
+class PipelineInputs(InputOutputBaseModel):
+    """Pipeline input configuration.
+
+    Defines the components that should receive the Query input and any filters that apply to it.
+
+    :param query: List of components that will receive the `query` input.
+        Specify each component in the format: '<component-name>.<run-method-parameter-name>', for example: 'retriever.query'
+    :param filters: Optional list of components that will receive the filters input.
+        Specify each component using the format: '<component-name>.<run-method-parameter-name>', for example: 'retriever.filters'.
+    """
+
+    model_config = {"extra": "allow"}  # Allow additional fields in inputs
+
+    query: List[str] = Field(
+        ...,
+        description=(
+            "List of components and parameters that will receive the `query` input when they are executed. "
+            "Use the format: '<component-name>.<run-method-parameter-name>', for example: 'retriever.query'."
+        ),
+        min_items=1,
+    )
+    filters: List[str] = Field(
+        default_factory=list,
+        description=(
+            "List of components and parameters that will receive the `filters` input when they are executed. "
+            "Use the format: '<component-name>.<run-method-parameter-name>', for example: 'retriever.filters'."
+        ),
+    )
+
+
+class PipelineOutputs(InputOutputBaseModel):
+    """Pipeline output configuration.
+
+    Specify the components that will output `documents`, `answers`, or both.
+    You must include at least one. The outputs of these components become the final output of the pipeline.
+
+    :param documents: Name of the component and parameter that will provide `documents` as output.
+        Use the format '<component-name>.<output-parameter>', for example: 'retriever.documents'.
+    :param answers: Name of the component and parameter that will provide `answers` as output.
+        Use the format '<component-name>.<output-parameter>', for example: 'reader.answers'.
+    """
+
+    model_config = {"extra": "allow"}  # Allow additional fields in outputs
+
+    documents: str | None = Field(
+        default=None,
+        description="Name of the component that will provide `documents` as output. "
+        "Format: '<component-name>.<output-parameter>', for example: 'meta_ranker.documents'",
+    )
+    answers: str | None = Field(
+        default=None,
+        description="Name of the component that will provide `answers` as output. "
+        "Format: '<component-name>.<output-parameter>', for example: 'answers_builder.answers'",
+    )
+
+    @model_validator(mode="after")
+    def validate_documents_xor_answers(self) -> "PipelineOutputs":
+        """Validate that either `documents`, `answers`, or both are defined."""
+        if not self.documents and not self.answers:
+            raise ValueError("Define at least one pipeline output, either 'documents, 'answers' or both.")
+        return self
+
+
+class IndexOutputs(InputOutputBaseModel):
+    """Output configuration for the index.
+
+    Index outputs are optional.
+    """
+
+    model_config = {"extra": "allow"}  # Allow additional fields in outputs
+
+
+class PipelineConfig(BaseModel):
+    """Configuration required to import the pipeline into deepset AI Platform.
+
+    :param name: Name of the pipeline to be imported
+    :param inputs: Pipeline input configuration. Use `PipelineInputs` model to define the inputs.
+    :param outputs: Pipeline output configuration. Use `PipelineOutputs` model to define the outputs.
+    """
+
+    model_config = {"extra": "forbid"}
+
+    name: str = Field(..., description="The name of the pipeline to be imported", min_length=1)
+    inputs: PipelineInputs = Field(
+        default_factory=PipelineInputs,
+        description=("Pipeline input configuration. Use `PipelineInputs` model to define the inputs."),
+    )
+    outputs: PipelineOutputs = Field(
+        default_factory=PipelineOutputs,
+        description=("Pipeline output configuration. Use `PipelineOutputs` model to define the outputs."),
+    )
+
+
+class IndexInputs(InputOutputBaseModel):
+    """Configuration required to import an index into deepset AI Platform.
+
+    Defines the index components that should receive the `Files` input.
+
+    :param files: List of components and parameters that should receive files as input.
+        Specify the components using the format: '<component-name>.<run-method-parameter-name>', for example: 'file_type_router.sources'.
+    """
+
+    model_config = {"extra": "allow"}  # Allow additional fields in inputs
+
+    files: List[str] = Field(
+        default_factory=list,
+        description=(
+            "List of components and parameters that will receive files as input when they're executed. "
+            "Format: '<component-name>.<run-parameter-name>', for example: 'file_type_router.sources'."
+        ),
+    )
+
+
+class IndexConfig(BaseModel):
+    """Index configuration for importing an index to deepset AI platform.
+
+    :param name: Name of the index to be imported.
+    :param inputs: Index input configuration. Use `IndexInputs` model to define the inputs.
+    :param outputs: Index output configuration. Optional. Use `IndexOutputs` model to define the outputs.
+    """
+
+    model_config = {"extra": "forbid"}
+
+    name: str = Field(..., description="Name of the index to be imported.", min_length=1)
+    inputs: IndexInputs = Field(
+        default_factory=IndexInputs,
+        description=("Input configuration for the index. Use `IndexInputs` model to define the inputs."),
+    )
+    outputs: IndexOutputs | None = Field(
+        default_factory=IndexOutputs,
+        description=("Optional output configuration for the index. Use `IndexOutputs` model to define the outputs."),
+    )
Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	`+"""Package to enable importing pipelines and indexes to deepset AI platform."""`