Add orchestration ID reuse policy with ERROR, IGNORE, and TERMINATE actions

[Copilot]

Implements orchestration ID reuse policy to control behavior when creating instances with duplicate IDs, matching durabletask-go and durabletask-java functionality.

Changes

Protobuf

• Added CreateOrchestrationAction enum (ERROR, IGNORE, TERMINATE)
• Updated OrchestrationIdReusePolicy to include action field and renamed replaceableStatus to operationStatus

Public API

• CreateOrchestrationAction enum defining three collision behaviors
• OrchestrationIdReusePolicy class with static factory methods for each action
• WithIdReusePolicy() extension method on StartOrchestrationOptions
• IdReusePolicy property on StartOrchestrationOptions

Implementation

• GrpcDurableTaskClient handles new policy with fallback to legacy DedupeStatuses
• ProtoUtils conversion methods between public and protobuf types

Usage

// Throw if running instance exists
var options = new StartOrchestrationOptions()
    .WithIdReusePolicy(OrchestrationIdReusePolicy.Error(
        OrchestrationRuntimeStatus.Running));

// Silently ignore if completed instance exists  
var options = new StartOrchestrationOptions()
    .WithIdReusePolicy(OrchestrationIdReusePolicy.Ignore(
        OrchestrationRuntimeStatus.Completed));

// Atomically terminate and replace running instance
var options = new StartOrchestrationOptions()
    .WithIdReusePolicy(OrchestrationIdReusePolicy.Terminate(
        OrchestrationRuntimeStatus.Running));

Existing DedupeStatuses API remains supported for backward compatibility.

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Support orchestration ID reuse policy</issue_title>
<issue_description>durabletask-go now support orchestration ID reuse policy, customer can reuse the orchestration ID with three different action ERROR, IGNORE, and TERMINATE with a target runtime status set.

ERROR- If there is an existing workflow then the scheduler throws an exception (this is the current behavior).
IGNORE- If there is an existing workflow already scheduled, then the scheduler does nothing.
TERMINATE- Terminates any existing workflows with the same instance ID and then schedules a new instance as one atomic action, similar to on-demand ContinueAsNew.

SDK needs to expose those options to customer orchestration ID reuse support.

Reference:

1. Support reusing orchestration id durabletask-go#46
2. Support orchestration id reuse policy durabletask-java#188</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Support orchestration ID reuse policy #255

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[torosent]

@copilot This PR contains breaking changes in the proto files