#
# Copyright 2026 The Dapr Authors
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#     http://www.apache.org/licenses/LICENSE-2.0
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#

import logging
from typing import Any, Dict, Optional

from pydantic import BaseModel, Field

from dapr_agents.tool.workflow.tool_context import WorkflowContextInjectedTool

logger = logging.getLogger(__name__)


def make_mcp_gateway_via_child_workflow_tool(
    *,
    target_app_id: str,
    gateway_workflow_name: str,
    name: str = "McpGatewayCall",
) -> WorkflowContextInjectedTool:
    """
    Create an AgentTool that calls an MCP "gateway workflow" hosted in another Dapr app,
    using the *current* workflow context (passed in by the agent) to schedule a
    multi-app child workflow.

    This uses Dapr's multi-application child workflow routing via `app_id`.

    The called workflow is expected to:
      - Receive input: {"tool": <str>, "arguments": <dict>}
      - Return: a JSON-serializable result (string/dict/etc.) representing the MCP call result.
    """

    class Args(BaseModel):
        tool: str = Field(
            ..., description="Name of the MCP tool to call on the remote side."
        )
        arguments: Dict[str, Any] = Field(
            default_factory=dict,
            description="Arguments to pass to the remote MCP tool.",
        )
        instance_id: Optional[str] = Field(
            default=None,
            description=(
                "Optional child workflow instance id. Use when you need idempotency / "
                "dedupe semantics at the child-workflow level."
            ),
        )

    def _executor(
        ctx: Any,
        tool: str,
        arguments: Dict[str, Any],
        instance_id: Optional[str] = None,
    ) -> Any:
        """
        Schedule a child workflow on the target app id, passing the MCP tool call as input.

        NOTE: This is intentionally a *sync* function that returns a workflow Task.
        In a Dapr workflow orchestrator, you'd typically do:
            result = yield tool_obj(ctx=ctx, tool="X", arguments={...})
        """
        payload = {"tool": tool, "arguments": arguments}

        logger.debug(
            "Scheduling child workflow '%s' on app_id='%s' for MCP tool='%s' args=%s instance_id=%s",
            gateway_workflow_name,
            target_app_id,
            tool,
            arguments,
            instance_id,
        )

        # Dapr multi-app child workflow call: route execution to target_app_id.
        # API: yield ctx.call_child_workflow(workflow='Workflow2', input='my-input', app_id='App2')
        if instance_id:
            return ctx.call_child_workflow(
                workflow=gateway_workflow_name,
                input=payload,
                instance_id=instance_id,
                app_id=target_app_id,
            )

        return ctx.call_child_workflow(
            workflow=gateway_workflow_name,
            input=payload,
            app_id=target_app_id,
        )

    return WorkflowContextInjectedTool(
        name=name,
        description=(
            f"Gateway tool that calls MCP via child workflow '{gateway_workflow_name}' "
            f"hosted by Dapr app '{target_app_id}'."
        ),
        func=_executor,
        args_model=Args,
    )