src/app/WriteClient.cpp - third_party/github/project-chip/connectedhomeip - Git at Google

 /*
  *
  *    Copyright (c) 2021 Project CHIP Authors
  *    All rights reserved.
  *
  *    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.
  */

 /**
  *    @file
  *      This file defines the initiator side of a CHIP Write Interaction.
  *
  */

 #include "lib/core/CHIPError.h"
 #include <app/AppBuildConfig.h>
 #include <app/InteractionModelEngine.h>
 #include <app/TimedRequest.h>
 #include <app/WriteClient.h>

 namespace chip {
 namespace app {

 void WriteClient::Close()
 {
     MoveToState(State::AwaitingDestruction);

     // OnDone below can destroy us before we unwind all the way back into the
     // exchange code and it tries to close itself.  Make sure that it doesn't
     // try to notify us that it's closing, since we will be dead.
     //
     // For more details, see #10344.
     if (mpExchangeCtx != nullptr)
     {
         mpExchangeCtx->SetDelegate(nullptr);
     }

     mpExchangeCtx = nullptr;

     if (mpCallback)
     {
         mpCallback->OnDone(this);
     }
 }

 void WriteClient::Abort()
 {
     //
     // If the exchange context hasn't already been gracefully closed
     // (signaled by setting it to null), then we need to forcibly
     // tear it down.
     //
     if (mpExchangeCtx != nullptr)
     {
         // We might be a delegate for this exchange, and we don't want the
         // OnExchangeClosing notification in that case.  Null out the delegate
         // to avoid that.
         //
         // TODO: This makes all sorts of assumptions about what the delegate is
         // (notice the "might" above!) that might not hold in practice.  We
         // really need a better solution here....
         mpExchangeCtx->SetDelegate(nullptr);
         mpExchangeCtx->Abort();
         mpExchangeCtx = nullptr;
     }
 }

 CHIP_ERROR WriteClient::ProcessWriteResponseMessage(System::PacketBufferHandle && payload)
 {
     CHIP_ERROR err = CHIP_NO_ERROR;
     System::PacketBufferTLVReader reader;
     TLV::TLVReader attributeStatusesReader;
     WriteResponseMessage::Parser writeResponse;
     AttributeStatusIBs::Parser attributeStatusesParser;

     reader.Init(std::move(payload));

     err = writeResponse.Init(reader);
     SuccessOrExit(err);

 #if CHIP_CONFIG_IM_ENABLE_SCHEMA_CHECK
     err = writeResponse.CheckSchemaValidity();
     SuccessOrExit(err);
 #endif
     err = writeResponse.GetWriteResponses(&attributeStatusesParser);
     SuccessOrExit(err);

     attributeStatusesParser.GetReader(&attributeStatusesReader);

     while (CHIP_NO_ERROR == (err = attributeStatusesReader.Next()))
     {
         VerifyOrExit(TLV::AnonymousTag() == attributeStatusesReader.GetTag(), err = CHIP_ERROR_INVALID_TLV_TAG);

         AttributeStatusIB::Parser element;

         err = element.Init(attributeStatusesReader);
         SuccessOrExit(err);

         err = ProcessAttributeStatusIB(element);
         SuccessOrExit(err);
     }

     // if we have exhausted this container
     if (CHIP_END_OF_TLV == err)
     {
         err = CHIP_NO_ERROR;
     }
     SuccessOrExit(err);
     ReturnErrorOnFailure(writeResponse.ExitContainer());

 exit:
     return err;
 }

 CHIP_ERROR WriteClient::PrepareAttributeIB(const ConcreteDataAttributePath & aPath)
 {
     AttributeDataIBs::Builder & writeRequests  = mWriteRequestBuilder.GetWriteRequests();
     AttributeDataIB::Builder & attributeDataIB = writeRequests.CreateAttributeDataIBBuilder();
     ReturnErrorOnFailure(writeRequests.GetError());
     if (aPath.mDataVersion.HasValue())
     {
         attributeDataIB.DataVersion(aPath.mDataVersion.Value());
         mHasDataVersion = true;
     }
     ReturnErrorOnFailure(attributeDataIB.GetError());
     AttributePathIB::Builder & path = attributeDataIB.CreatePath();

     // We are using kInvalidEndpointId just for group write requests. This is not the correct use of ConcreteDataAttributePath.
     // TODO: update AttributePathParams or ConcreteDataAttributePath for a class supports both nullable list index and missing
     // endpoint id.
     if (aPath.mEndpointId != kInvalidEndpointId)
     {
         path.Endpoint(aPath.mEndpointId);
     }
     path.Cluster(aPath.mClusterId).Attribute(aPath.mAttributeId);
     if (aPath.IsListItemOperation())
     {
         if (aPath.mListOp == ConcreteDataAttributePath::ListOperation::AppendItem)
         {
             path.ListIndex(DataModel::NullNullable);
         }
         else
         {
             // We do not support other list operations (i.e. update, delete etc) for now.
             return CHIP_ERROR_UNSUPPORTED_CHIP_FEATURE;
         }
     }
     ReturnErrorOnFailure(path.EndOfAttributePathIB().GetError());

     return CHIP_NO_ERROR;
 }

 CHIP_ERROR WriteClient::FinishAttributeIB()
 {
     AttributeDataIB::Builder & attributeDataIB = mWriteRequestBuilder.GetWriteRequests().GetAttributeDataIBBuilder();
     attributeDataIB.EndOfAttributeDataIB();
     ReturnErrorOnFailure(attributeDataIB.GetError());
     MoveToState(State::AddAttribute);
     return CHIP_NO_ERROR;
 }

 TLV::TLVWriter * WriteClient::GetAttributeDataIBTLVWriter()
 {
     return mWriteRequestBuilder.GetWriteRequests().GetAttributeDataIBBuilder().GetWriter();
 }

 CHIP_ERROR WriteClient::FinalizeMessage(bool aHasMoreChunks)
 {
     System::PacketBufferHandle packet;
     VerifyOrReturnError(mState == State::AddAttribute, CHIP_ERROR_INCORRECT_STATE);

     TLV::TLVWriter * writer = mWriteRequestBuilder.GetWriter();
     ReturnErrorCodeIf(writer == nullptr, CHIP_ERROR_INCORRECT_STATE);
     ReturnErrorOnFailure(writer->UnreserveBuffer(kReservedSizeForTLVEncodingOverhead));

     AttributeDataIBs::Builder & attributeDataIBsBuilder = mWriteRequestBuilder.GetWriteRequests().EndOfAttributeDataIBs();
     ReturnErrorOnFailure(attributeDataIBsBuilder.GetError());

     mWriteRequestBuilder.MoreChunkedMessages(aHasMoreChunks).EndOfWriteRequestMessage();
     ReturnErrorOnFailure(mWriteRequestBuilder.GetError());
     ReturnErrorOnFailure(mMessageWriter.Finalize(&packet));
     mChunks.AddToEnd(std::move(packet));
     return CHIP_NO_ERROR;
 }

 CHIP_ERROR WriteClient::EnsureMessage()
 {
     if (mState != State::AddAttribute)
     {
         return StartNewMessage();
     }
     return CHIP_NO_ERROR;
 }

 CHIP_ERROR WriteClient::StartNewMessage()
 {
     uint16_t reservedSize = 0;

     if (mState == State::AddAttribute)
     {
         ReturnErrorOnFailure(FinalizeMessage(true));
     }

     // Do not allow timed request with chunks.
     VerifyOrReturnError(!(mTimedWriteTimeoutMs.HasValue() && !mChunks.IsNull()), CHIP_ERROR_NO_MEMORY);

     System::PacketBufferHandle packet = System::PacketBufferHandle::New(kMaxSecureSduLengthBytes);
     VerifyOrReturnError(!packet.IsNull(), CHIP_ERROR_NO_MEMORY);

     // Always limit the size of the packet to fit within kMaxSecureSduLengthBytes regardless of the available buffer capacity.
     if (packet->AvailableDataLength() > kMaxSecureSduLengthBytes)
     {
         reservedSize = static_cast<uint16_t>(packet->AvailableDataLength() - kMaxSecureSduLengthBytes);
     }

     // ... and we need to reserve some extra space for the MIC field.
     reservedSize = static_cast<uint16_t>(reservedSize + Crypto::CHIP_CRYPTO_AEAD_MIC_LENGTH_BYTES);

     // ... and the overhead for end of AttributeDataIBs (end of container), more chunks flag, end of WriteRequestMessage (another
     // end of container).
     reservedSize = static_cast<uint16_t>(reservedSize + kReservedSizeForTLVEncodingOverhead);

 #if CONFIG_IM_BUILD_FOR_UNIT_TEST
     // ... and for unit tests.
     reservedSize = static_cast<uint16_t>(reservedSize + mReservedSize);
 #endif

     mMessageWriter.Init(std::move(packet));

     ReturnErrorOnFailure(mMessageWriter.ReserveBuffer(reservedSize));

     ReturnErrorOnFailure(mWriteRequestBuilder.Init(&mMessageWriter));
     mWriteRequestBuilder.SuppressResponse(mSuppressResponse);
     mWriteRequestBuilder.TimedRequest(mTimedWriteTimeoutMs.HasValue());
     ReturnErrorOnFailure(mWriteRequestBuilder.GetError());
     mWriteRequestBuilder.CreateWriteRequests();
     ReturnErrorOnFailure(mWriteRequestBuilder.GetError());

     TLV::TLVWriter * writer = mWriteRequestBuilder.GetWriter();
     VerifyOrReturnError(writer != nullptr, CHIP_ERROR_INCORRECT_STATE);

     return CHIP_NO_ERROR;
 }

 CHIP_ERROR WriteClient::TryPutSinglePreencodedAttributeWritePayload(const ConcreteDataAttributePath & attributePath,
                                                                     const TLV::TLVReader & data)
 {
     TLV::TLVReader dataToWrite;
     dataToWrite.Init(data);

     TLV::TLVWriter * writer = nullptr;

     ReturnErrorOnFailure(PrepareAttributeIB(attributePath));
     VerifyOrReturnError((writer = GetAttributeDataIBTLVWriter()) != nullptr, CHIP_ERROR_INCORRECT_STATE);
     ReturnErrorOnFailure(writer->CopyElement(TLV::ContextTag(to_underlying(AttributeDataIB::Tag::kData)), dataToWrite));
     ReturnErrorOnFailure(FinishAttributeIB());
     return CHIP_NO_ERROR;
 }

 CHIP_ERROR WriteClient::PutSinglePreencodedAttributeWritePayload(const chip::app::ConcreteDataAttributePath & attributePath,
                                                                  const TLV::TLVReader & data)
 {
     TLV::TLVWriter backupWriter;

     mWriteRequestBuilder.GetWriteRequests().Checkpoint(backupWriter);

     // First attempt to write this attribute.
     CHIP_ERROR err = TryPutSinglePreencodedAttributeWritePayload(attributePath, data);
     if (err == CHIP_ERROR_NO_MEMORY || err == CHIP_ERROR_BUFFER_TOO_SMALL)
     {
         // If it failed with no memory, then we create a new chunk for it.
         mWriteRequestBuilder.GetWriteRequests().Rollback(backupWriter);
         mWriteRequestBuilder.GetWriteRequests().ResetError();
         ReturnErrorOnFailure(StartNewMessage());
         err = TryPutSinglePreencodedAttributeWritePayload(attributePath, data);
         // Since we have created a new chunk for this element, the encode is expected to succeed.
     }
     return err;
 }

 CHIP_ERROR WriteClient::PutPreencodedAttribute(const ConcreteDataAttributePath & attributePath, const TLV::TLVReader & data)
 {
     ReturnErrorOnFailure(EnsureMessage());

     // ListIndex is missing and the data is an array -- we are writing a whole list.
     if (!attributePath.IsListOperation() && data.GetType() == TLV::TLVType::kTLVType_Array)
     {
         TLV::TLVReader dataReader;
         TLV::TLVReader valueReader;
         CHIP_ERROR err = CHIP_NO_ERROR;

         ConcreteDataAttributePath path = attributePath;

         dataReader.Init(data);
         dataReader.OpenContainer(valueReader);

         // Encode an empty list for the chunking protocol.
         ReturnErrorOnFailure(EncodeSingleAttributeDataIB(path, DataModel::List<uint8_t>()));

         if (err == CHIP_NO_ERROR)
         {
             path.mListOp = ConcreteDataAttributePath::ListOperation::AppendItem;
             while ((err = valueReader.Next()) == CHIP_NO_ERROR)
             {
                 ReturnErrorOnFailure(PutSinglePreencodedAttributeWritePayload(path, valueReader));
             }
         }

         if (err == CHIP_END_OF_TLV)
         {
             err = CHIP_NO_ERROR;
         }
         return err;
     }
     // We are writing a non-list attribute, or we are writing a single element of a list.
     return PutSinglePreencodedAttributeWritePayload(attributePath, data);
 }

 const char * WriteClient::GetStateStr() const
 {
 #if CHIP_DETAIL_LOGGING
     switch (mState)
     {
     case State::Initialized:
         return "Initialized";

     case State::AddAttribute:
         return "AddAttribute";

     case State::AwaitingTimedStatus:
         return "AwaitingTimedStatus";

     case State::AwaitingResponse:
         return "AwaitingResponse";

     case State::ResponseReceived:
         return "ResponseReceived";

     case State::AwaitingDestruction:
         return "AwaitingDestruction";
     }
 #endif // CHIP_DETAIL_LOGGING
     return "N/A";
 }

 void WriteClient::MoveToState(const State aTargetState)
 {
     mState = aTargetState;
     ChipLogDetail(DataManagement, "WriteClient moving to [%10.10s]", GetStateStr());
 }

 CHIP_ERROR WriteClient::SendWriteRequest(const SessionHandle & session, System::Clock::Timeout timeout)
 {
     CHIP_ERROR err = CHIP_NO_ERROR;

     VerifyOrExit(mState == State::AddAttribute, err = CHIP_ERROR_INCORRECT_STATE);

     err = FinalizeMessage(false /* hasMoreChunks */);
     SuccessOrExit(err);

     // Create a new exchange context.
     mpExchangeCtx = mpExchangeMgr->NewContext(session, this);
     VerifyOrExit(mpExchangeCtx != nullptr, err = CHIP_ERROR_NO_MEMORY);
     VerifyOrReturnError(!(mpExchangeCtx->IsGroupExchangeContext() && mHasDataVersion), CHIP_ERROR_INVALID_MESSAGE_TYPE);
     mpExchangeCtx->SetResponseTimeout(timeout);

     if (mTimedWriteTimeoutMs.HasValue())
     {
         err = TimedRequest::Send(mpExchangeCtx, mTimedWriteTimeoutMs.Value());
         SuccessOrExit(err);
         MoveToState(State::AwaitingTimedStatus);
     }
     else
     {
         err = SendWriteRequest();
         SuccessOrExit(err);
     }

 exit:
     if (err != CHIP_NO_ERROR)
     {
         ChipLogError(DataManagement, "Write client failed to SendWriteRequest: %s", ErrorStr(err));
     }
     else
     {
         // TODO: Ideally this would happen async, but to make sure that we
         // handle this object dying (e.g. due to IM enging shutdown) while the
         // async bits are pending we'd need to malloc some state bit that we can
         // twiddle if we die.  For now just do the OnDone callback sync.
         if (session->IsGroupSession())
         {
             // Always shutdown on Group communication
             ChipLogDetail(DataManagement, "Closing on group Communication ");

             // Tell the application to release the object.
             // TODO: Consumers expect to hand off ownership of the WriteClient and wait for OnDone
             // after SendWriteRequest returns success.  Calling OnDone before returning is weird.
             // Need to refactor the code to avoid this.
             Close();
         }
     }

     return err;
 }

 CHIP_ERROR WriteClient::SendWriteRequest()
 {
     using namespace Protocols::InteractionModel;
     using namespace Messaging;

     System::PacketBufferHandle data = mChunks.PopHead();

     if (!mChunks.IsNull() && mpExchangeCtx->IsGroupExchangeContext())
     {
         // Reject this request if we have more than one chunk (mChunks is not null after PopHead()), and this is a group
         // exchange context.
         return CHIP_ERROR_INCORRECT_STATE;
     }

     // kExpectResponse is ignored by ExchangeContext in case of groupcast
     ReturnErrorOnFailure(mpExchangeCtx->SendMessage(MsgType::WriteRequest, std::move(data), SendMessageFlags::kExpectResponse));
     MoveToState(State::AwaitingResponse);
     return CHIP_NO_ERROR;
 }

 CHIP_ERROR WriteClient::OnMessageReceived(Messaging::ExchangeContext * apExchangeContext, const PayloadHeader & aPayloadHeader,
                                           System::PacketBufferHandle && aPayload)
 {
     if (mState == State::AwaitingResponse &&
         // We had sent the last chunk of data, and received all responses
         mChunks.IsNull())
     {
         MoveToState(State::ResponseReceived);
     }

     CHIP_ERROR err = CHIP_NO_ERROR;
     // Assert that the exchange context matches the client's current context.
     // This should never fail because even if SendWriteRequest is called
     // back-to-back, the second call will call Close() on the first exchange,
     // which clears the OnMessageReceived callback.
     VerifyOrExit(apExchangeContext == mpExchangeCtx, err = CHIP_ERROR_INCORRECT_STATE);

     if (mState == State::AwaitingTimedStatus)
     {
         err = HandleTimedStatus(aPayloadHeader, std::move(aPayload));
         // Skip all other processing here (which is for the response to the
         // write request), no matter whether err is success or not.
         goto exit;
     }

     if (aPayloadHeader.HasMessageType(Protocols::InteractionModel::MsgType::WriteResponse))
     {
         err = ProcessWriteResponseMessage(std::move(aPayload));
         SuccessOrExit(err);
         if (!mChunks.IsNull())
         {
             // Send the next chunk.
             SuccessOrExit(SendWriteRequest());
         }
     }
     else if (aPayloadHeader.HasMessageType(Protocols::InteractionModel::MsgType::StatusResponse))
     {
         err = StatusResponse::ProcessStatusResponse(std::move(aPayload));
         SuccessOrExit(err);
     }
     else
     {
         err = CHIP_ERROR_INVALID_MESSAGE_TYPE;
     }

 exit:
     if (mpCallback != nullptr)
     {
         if (err != CHIP_NO_ERROR)
         {
             mpCallback->OnError(this, err);
         }
     }

     if (mState != State::AwaitingResponse)
     {
         Close();
     }
     // Else we got a response to a Timed Request and just sent the write.

     return err;
 }

 void WriteClient::OnResponseTimeout(Messaging::ExchangeContext * apExchangeContext)
 {
     ChipLogError(DataManagement, "Time out! failed to receive write response from Exchange: " ChipLogFormatExchange,
                  ChipLogValueExchange(apExchangeContext));

     if (mpCallback != nullptr)
     {
         mpCallback->OnError(this, CHIP_ERROR_TIMEOUT);
     }
     Close();
 }

 CHIP_ERROR WriteClient::ProcessAttributeStatusIB(AttributeStatusIB::Parser & aAttributeStatusIB)
 {
     CHIP_ERROR err = CHIP_NO_ERROR;
     AttributePathIB::Parser attributePathParser;
     StatusIB statusIB;
     StatusIB::Parser StatusIBParser;
     ConcreteDataAttributePath attributePath;

     err = aAttributeStatusIB.GetPath(&attributePathParser);
     SuccessOrExit(err);

     err = attributePathParser.GetCluster(&(attributePath.mClusterId));
     SuccessOrExit(err);
     err = attributePathParser.GetEndpoint(&(attributePath.mEndpointId));
     SuccessOrExit(err);
     err = attributePathParser.GetAttribute(&(attributePath.mAttributeId));
     SuccessOrExit(err);
     err = attributePathParser.GetListIndex(attributePath);
     SuccessOrExit(err);

     err = aAttributeStatusIB.GetErrorStatus(&(StatusIBParser));
     if (CHIP_NO_ERROR == err)
     {
         err = StatusIBParser.DecodeStatusIB(statusIB);
         SuccessOrExit(err);
         if (mpCallback != nullptr)
         {
             mpCallback->OnResponse(this, attributePath, statusIB);
         }
     }

 exit:
     return err;
 }

 CHIP_ERROR WriteClient::HandleTimedStatus(const PayloadHeader & aPayloadHeader, System::PacketBufferHandle && aPayload)
 {
     ReturnErrorOnFailure(TimedRequest::HandleResponse(aPayloadHeader, std::move(aPayload)));

     return SendWriteRequest();
 }

 } // namespace app
 } // namespace chip
	/*
	*
	* Copyright (c) 2021 Project CHIP Authors
	* All rights reserved.
	*
	* 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.
	*/

	/**
	* @file
	* This file defines the initiator side of a CHIP Write Interaction.
	*
	*/

	#include "lib/core/CHIPError.h"
	#include <app/AppBuildConfig.h>
	#include <app/InteractionModelEngine.h>
	#include <app/TimedRequest.h>
	#include <app/WriteClient.h>

	namespace chip {
	namespace app {

	void WriteClient::Close()
	{
	MoveToState(State::AwaitingDestruction);

	// OnDone below can destroy us before we unwind all the way back into the
	// exchange code and it tries to close itself. Make sure that it doesn't
	// try to notify us that it's closing, since we will be dead.
	//
	// For more details, see #10344.
	if (mpExchangeCtx != nullptr)
	{
	mpExchangeCtx->SetDelegate(nullptr);
	}

	mpExchangeCtx = nullptr;

	if (mpCallback)
	{
	mpCallback->OnDone(this);
	}
	}

	void WriteClient::Abort()
	{
	//
	// If the exchange context hasn't already been gracefully closed
	// (signaled by setting it to null), then we need to forcibly
	// tear it down.
	//
	if (mpExchangeCtx != nullptr)
	{
	// We might be a delegate for this exchange, and we don't want the
	// OnExchangeClosing notification in that case. Null out the delegate
	// to avoid that.
	//
	// TODO: This makes all sorts of assumptions about what the delegate is
	// (notice the "might" above!) that might not hold in practice. We
	// really need a better solution here....
	mpExchangeCtx->SetDelegate(nullptr);
	mpExchangeCtx->Abort();
	mpExchangeCtx = nullptr;
	}
	}

	CHIP_ERROR WriteClient::ProcessWriteResponseMessage(System::PacketBufferHandle && payload)
	{
	CHIP_ERROR err = CHIP_NO_ERROR;
	System::PacketBufferTLVReader reader;
	TLV::TLVReader attributeStatusesReader;
	WriteResponseMessage::Parser writeResponse;
	AttributeStatusIBs::Parser attributeStatusesParser;

	reader.Init(std::move(payload));

	err = writeResponse.Init(reader);
	SuccessOrExit(err);

	#if CHIP_CONFIG_IM_ENABLE_SCHEMA_CHECK
	err = writeResponse.CheckSchemaValidity();
	SuccessOrExit(err);
	#endif
	err = writeResponse.GetWriteResponses(&attributeStatusesParser);
	SuccessOrExit(err);

	attributeStatusesParser.GetReader(&attributeStatusesReader);

	while (CHIP_NO_ERROR == (err = attributeStatusesReader.Next()))
	{
	VerifyOrExit(TLV::AnonymousTag() == attributeStatusesReader.GetTag(), err = CHIP_ERROR_INVALID_TLV_TAG);

	AttributeStatusIB::Parser element;

	err = element.Init(attributeStatusesReader);
	SuccessOrExit(err);

	err = ProcessAttributeStatusIB(element);
	SuccessOrExit(err);
	}

	// if we have exhausted this container
	if (CHIP_END_OF_TLV == err)
	{
	err = CHIP_NO_ERROR;
	}
	SuccessOrExit(err);
	ReturnErrorOnFailure(writeResponse.ExitContainer());

	exit:
	return err;
	}

	CHIP_ERROR WriteClient::PrepareAttributeIB(const ConcreteDataAttributePath & aPath)
	{
	AttributeDataIBs::Builder & writeRequests = mWriteRequestBuilder.GetWriteRequests();
	AttributeDataIB::Builder & attributeDataIB = writeRequests.CreateAttributeDataIBBuilder();
	ReturnErrorOnFailure(writeRequests.GetError());
	if (aPath.mDataVersion.HasValue())
	{
	attributeDataIB.DataVersion(aPath.mDataVersion.Value());
	mHasDataVersion = true;
	}
	ReturnErrorOnFailure(attributeDataIB.GetError());
	AttributePathIB::Builder & path = attributeDataIB.CreatePath();

	// We are using kInvalidEndpointId just for group write requests. This is not the correct use of ConcreteDataAttributePath.
	// TODO: update AttributePathParams or ConcreteDataAttributePath for a class supports both nullable list index and missing
	// endpoint id.
	if (aPath.mEndpointId != kInvalidEndpointId)
	{
	path.Endpoint(aPath.mEndpointId);
	}
	path.Cluster(aPath.mClusterId).Attribute(aPath.mAttributeId);
	if (aPath.IsListItemOperation())
	{
	if (aPath.mListOp == ConcreteDataAttributePath::ListOperation::AppendItem)
	{
	path.ListIndex(DataModel::NullNullable);
	}
	else
	{
	// We do not support other list operations (i.e. update, delete etc) for now.
	return CHIP_ERROR_UNSUPPORTED_CHIP_FEATURE;
	}
	}
	ReturnErrorOnFailure(path.EndOfAttributePathIB().GetError());

	return CHIP_NO_ERROR;
	}

	CHIP_ERROR WriteClient::FinishAttributeIB()
	{
	AttributeDataIB::Builder & attributeDataIB = mWriteRequestBuilder.GetWriteRequests().GetAttributeDataIBBuilder();
	attributeDataIB.EndOfAttributeDataIB();
	ReturnErrorOnFailure(attributeDataIB.GetError());
	MoveToState(State::AddAttribute);
	return CHIP_NO_ERROR;
	}

	TLV::TLVWriter * WriteClient::GetAttributeDataIBTLVWriter()
	{
	return mWriteRequestBuilder.GetWriteRequests().GetAttributeDataIBBuilder().GetWriter();
	}

	CHIP_ERROR WriteClient::FinalizeMessage(bool aHasMoreChunks)
	{
	System::PacketBufferHandle packet;
	VerifyOrReturnError(mState == State::AddAttribute, CHIP_ERROR_INCORRECT_STATE);

	TLV::TLVWriter * writer = mWriteRequestBuilder.GetWriter();
	ReturnErrorCodeIf(writer == nullptr, CHIP_ERROR_INCORRECT_STATE);
	ReturnErrorOnFailure(writer->UnreserveBuffer(kReservedSizeForTLVEncodingOverhead));

	AttributeDataIBs::Builder & attributeDataIBsBuilder = mWriteRequestBuilder.GetWriteRequests().EndOfAttributeDataIBs();
	ReturnErrorOnFailure(attributeDataIBsBuilder.GetError());

	mWriteRequestBuilder.MoreChunkedMessages(aHasMoreChunks).EndOfWriteRequestMessage();
	ReturnErrorOnFailure(mWriteRequestBuilder.GetError());
	ReturnErrorOnFailure(mMessageWriter.Finalize(&packet));
	mChunks.AddToEnd(std::move(packet));
	return CHIP_NO_ERROR;
	}

	CHIP_ERROR WriteClient::EnsureMessage()
	{
	if (mState != State::AddAttribute)
	{
	return StartNewMessage();
	}
	return CHIP_NO_ERROR;
	}

	CHIP_ERROR WriteClient::StartNewMessage()
	{
	uint16_t reservedSize = 0;

	if (mState == State::AddAttribute)
	{
	ReturnErrorOnFailure(FinalizeMessage(true));
	}

	// Do not allow timed request with chunks.
	VerifyOrReturnError(!(mTimedWriteTimeoutMs.HasValue() && !mChunks.IsNull()), CHIP_ERROR_NO_MEMORY);

	System::PacketBufferHandle packet = System::PacketBufferHandle::New(kMaxSecureSduLengthBytes);
	VerifyOrReturnError(!packet.IsNull(), CHIP_ERROR_NO_MEMORY);

	// Always limit the size of the packet to fit within kMaxSecureSduLengthBytes regardless of the available buffer capacity.
	if (packet->AvailableDataLength() > kMaxSecureSduLengthBytes)
	{
	reservedSize = static_cast<uint16_t>(packet->AvailableDataLength() - kMaxSecureSduLengthBytes);
	}

	// ... and we need to reserve some extra space for the MIC field.
	reservedSize = static_cast<uint16_t>(reservedSize + Crypto::CHIP_CRYPTO_AEAD_MIC_LENGTH_BYTES);

	// ... and the overhead for end of AttributeDataIBs (end of container), more chunks flag, end of WriteRequestMessage (another
	// end of container).
	reservedSize = static_cast<uint16_t>(reservedSize + kReservedSizeForTLVEncodingOverhead);

	#if CONFIG_IM_BUILD_FOR_UNIT_TEST
	// ... and for unit tests.
	reservedSize = static_cast<uint16_t>(reservedSize + mReservedSize);
	#endif

	mMessageWriter.Init(std::move(packet));

	ReturnErrorOnFailure(mMessageWriter.ReserveBuffer(reservedSize));

	ReturnErrorOnFailure(mWriteRequestBuilder.Init(&mMessageWriter));
	mWriteRequestBuilder.SuppressResponse(mSuppressResponse);
	mWriteRequestBuilder.TimedRequest(mTimedWriteTimeoutMs.HasValue());
	ReturnErrorOnFailure(mWriteRequestBuilder.GetError());
	mWriteRequestBuilder.CreateWriteRequests();
	ReturnErrorOnFailure(mWriteRequestBuilder.GetError());

	TLV::TLVWriter * writer = mWriteRequestBuilder.GetWriter();
	VerifyOrReturnError(writer != nullptr, CHIP_ERROR_INCORRECT_STATE);

	return CHIP_NO_ERROR;
	}

	CHIP_ERROR WriteClient::TryPutSinglePreencodedAttributeWritePayload(const ConcreteDataAttributePath & attributePath,
	const TLV::TLVReader & data)
	{
	TLV::TLVReader dataToWrite;
	dataToWrite.Init(data);

	TLV::TLVWriter * writer = nullptr;

	ReturnErrorOnFailure(PrepareAttributeIB(attributePath));
	VerifyOrReturnError((writer = GetAttributeDataIBTLVWriter()) != nullptr, CHIP_ERROR_INCORRECT_STATE);
	ReturnErrorOnFailure(writer->CopyElement(TLV::ContextTag(to_underlying(AttributeDataIB::Tag::kData)), dataToWrite));
	ReturnErrorOnFailure(FinishAttributeIB());
	return CHIP_NO_ERROR;
	}

	CHIP_ERROR WriteClient::PutSinglePreencodedAttributeWritePayload(const chip::app::ConcreteDataAttributePath & attributePath,
	const TLV::TLVReader & data)
	{
	TLV::TLVWriter backupWriter;

	mWriteRequestBuilder.GetWriteRequests().Checkpoint(backupWriter);

	// First attempt to write this attribute.
	CHIP_ERROR err = TryPutSinglePreencodedAttributeWritePayload(attributePath, data);
	if (err == CHIP_ERROR_NO_MEMORY \|\| err == CHIP_ERROR_BUFFER_TOO_SMALL)
	{
	// If it failed with no memory, then we create a new chunk for it.
	mWriteRequestBuilder.GetWriteRequests().Rollback(backupWriter);
	mWriteRequestBuilder.GetWriteRequests().ResetError();
	ReturnErrorOnFailure(StartNewMessage());
	err = TryPutSinglePreencodedAttributeWritePayload(attributePath, data);
	// Since we have created a new chunk for this element, the encode is expected to succeed.
	}
	return err;
	}

	CHIP_ERROR WriteClient::PutPreencodedAttribute(const ConcreteDataAttributePath & attributePath, const TLV::TLVReader & data)
	{
	ReturnErrorOnFailure(EnsureMessage());

	// ListIndex is missing and the data is an array -- we are writing a whole list.
	if (!attributePath.IsListOperation() && data.GetType() == TLV::TLVType::kTLVType_Array)
	{
	TLV::TLVReader dataReader;
	TLV::TLVReader valueReader;
	CHIP_ERROR err = CHIP_NO_ERROR;

	ConcreteDataAttributePath path = attributePath;

	dataReader.Init(data);
	dataReader.OpenContainer(valueReader);

	// Encode an empty list for the chunking protocol.
	ReturnErrorOnFailure(EncodeSingleAttributeDataIB(path, DataModel::List<uint8_t>()));

	if (err == CHIP_NO_ERROR)
	{
	path.mListOp = ConcreteDataAttributePath::ListOperation::AppendItem;
	while ((err = valueReader.Next()) == CHIP_NO_ERROR)
	{
	ReturnErrorOnFailure(PutSinglePreencodedAttributeWritePayload(path, valueReader));
	}
	}

	if (err == CHIP_END_OF_TLV)
	{
	err = CHIP_NO_ERROR;
	}
	return err;
	}
	// We are writing a non-list attribute, or we are writing a single element of a list.
	return PutSinglePreencodedAttributeWritePayload(attributePath, data);
	}

	const char * WriteClient::GetStateStr() const
	{
	#if CHIP_DETAIL_LOGGING
	switch (mState)
	{
	case State::Initialized:
	return "Initialized";

	case State::AddAttribute:
	return "AddAttribute";

	case State::AwaitingTimedStatus:
	return "AwaitingTimedStatus";

	case State::AwaitingResponse:
	return "AwaitingResponse";

	case State::ResponseReceived:
	return "ResponseReceived";

	case State::AwaitingDestruction:
	return "AwaitingDestruction";
	}
	#endif // CHIP_DETAIL_LOGGING
	return "N/A";
	}

	void WriteClient::MoveToState(const State aTargetState)
	{
	mState = aTargetState;
	ChipLogDetail(DataManagement, "WriteClient moving to [%10.10s]", GetStateStr());
	}

	CHIP_ERROR WriteClient::SendWriteRequest(const SessionHandle & session, System::Clock::Timeout timeout)
	{
	CHIP_ERROR err = CHIP_NO_ERROR;

	VerifyOrExit(mState == State::AddAttribute, err = CHIP_ERROR_INCORRECT_STATE);

	err = FinalizeMessage(false /* hasMoreChunks */);
	SuccessOrExit(err);

	// Create a new exchange context.
	mpExchangeCtx = mpExchangeMgr->NewContext(session, this);
	VerifyOrExit(mpExchangeCtx != nullptr, err = CHIP_ERROR_NO_MEMORY);
	VerifyOrReturnError(!(mpExchangeCtx->IsGroupExchangeContext() && mHasDataVersion), CHIP_ERROR_INVALID_MESSAGE_TYPE);
	mpExchangeCtx->SetResponseTimeout(timeout);

	if (mTimedWriteTimeoutMs.HasValue())
	{
	err = TimedRequest::Send(mpExchangeCtx, mTimedWriteTimeoutMs.Value());
	SuccessOrExit(err);
	MoveToState(State::AwaitingTimedStatus);
	}
	else
	{
	err = SendWriteRequest();
	SuccessOrExit(err);
	}

	exit:
	if (err != CHIP_NO_ERROR)
	{
	ChipLogError(DataManagement, "Write client failed to SendWriteRequest: %s", ErrorStr(err));
	}
	else
	{
	// TODO: Ideally this would happen async, but to make sure that we
	// handle this object dying (e.g. due to IM enging shutdown) while the
	// async bits are pending we'd need to malloc some state bit that we can
	// twiddle if we die. For now just do the OnDone callback sync.
	if (session->IsGroupSession())
	{
	// Always shutdown on Group communication
	ChipLogDetail(DataManagement, "Closing on group Communication ");

	// Tell the application to release the object.
	// TODO: Consumers expect to hand off ownership of the WriteClient and wait for OnDone
	// after SendWriteRequest returns success. Calling OnDone before returning is weird.
	// Need to refactor the code to avoid this.
	Close();
	}
	}

	return err;
	}

	CHIP_ERROR WriteClient::SendWriteRequest()
	{
	using namespace Protocols::InteractionModel;
	using namespace Messaging;

	System::PacketBufferHandle data = mChunks.PopHead();

	if (!mChunks.IsNull() && mpExchangeCtx->IsGroupExchangeContext())
	{
	// Reject this request if we have more than one chunk (mChunks is not null after PopHead()), and this is a group
	// exchange context.
	return CHIP_ERROR_INCORRECT_STATE;
	}

	// kExpectResponse is ignored by ExchangeContext in case of groupcast
	ReturnErrorOnFailure(mpExchangeCtx->SendMessage(MsgType::WriteRequest, std::move(data), SendMessageFlags::kExpectResponse));
	MoveToState(State::AwaitingResponse);
	return CHIP_NO_ERROR;
	}

	CHIP_ERROR WriteClient::OnMessageReceived(Messaging::ExchangeContext * apExchangeContext, const PayloadHeader & aPayloadHeader,
	System::PacketBufferHandle && aPayload)
	{
	if (mState == State::AwaitingResponse &&
	// We had sent the last chunk of data, and received all responses
	mChunks.IsNull())
	{
	MoveToState(State::ResponseReceived);
	}

	CHIP_ERROR err = CHIP_NO_ERROR;
	// Assert that the exchange context matches the client's current context.
	// This should never fail because even if SendWriteRequest is called
	// back-to-back, the second call will call Close() on the first exchange,
	// which clears the OnMessageReceived callback.
	VerifyOrExit(apExchangeContext == mpExchangeCtx, err = CHIP_ERROR_INCORRECT_STATE);

	if (mState == State::AwaitingTimedStatus)
	{
	err = HandleTimedStatus(aPayloadHeader, std::move(aPayload));
	// Skip all other processing here (which is for the response to the
	// write request), no matter whether err is success or not.
	goto exit;
	}

	if (aPayloadHeader.HasMessageType(Protocols::InteractionModel::MsgType::WriteResponse))
	{
	err = ProcessWriteResponseMessage(std::move(aPayload));
	SuccessOrExit(err);
	if (!mChunks.IsNull())
	{
	// Send the next chunk.
	SuccessOrExit(SendWriteRequest());
	}
	}
	else if (aPayloadHeader.HasMessageType(Protocols::InteractionModel::MsgType::StatusResponse))
	{
	err = StatusResponse::ProcessStatusResponse(std::move(aPayload));
	SuccessOrExit(err);
	}
	else
	{
	err = CHIP_ERROR_INVALID_MESSAGE_TYPE;
	}

	exit:
	if (mpCallback != nullptr)
	{
	if (err != CHIP_NO_ERROR)
	{
	mpCallback->OnError(this, err);
	}
	}

	if (mState != State::AwaitingResponse)
	{
	Close();
	}
	// Else we got a response to a Timed Request and just sent the write.

	return err;
	}

	void WriteClient::OnResponseTimeout(Messaging::ExchangeContext * apExchangeContext)
	{
	ChipLogError(DataManagement, "Time out! failed to receive write response from Exchange: " ChipLogFormatExchange,
	ChipLogValueExchange(apExchangeContext));

	if (mpCallback != nullptr)
	{
	mpCallback->OnError(this, CHIP_ERROR_TIMEOUT);
	}
	Close();
	}

	CHIP_ERROR WriteClient::ProcessAttributeStatusIB(AttributeStatusIB::Parser & aAttributeStatusIB)
	{
	CHIP_ERROR err = CHIP_NO_ERROR;
	AttributePathIB::Parser attributePathParser;
	StatusIB statusIB;
	StatusIB::Parser StatusIBParser;
	ConcreteDataAttributePath attributePath;

	err = aAttributeStatusIB.GetPath(&attributePathParser);
	SuccessOrExit(err);

	err = attributePathParser.GetCluster(&(attributePath.mClusterId));
	SuccessOrExit(err);
	err = attributePathParser.GetEndpoint(&(attributePath.mEndpointId));
	SuccessOrExit(err);
	err = attributePathParser.GetAttribute(&(attributePath.mAttributeId));
	SuccessOrExit(err);
	err = attributePathParser.GetListIndex(attributePath);
	SuccessOrExit(err);

	err = aAttributeStatusIB.GetErrorStatus(&(StatusIBParser));
	if (CHIP_NO_ERROR == err)
	{
	err = StatusIBParser.DecodeStatusIB(statusIB);
	SuccessOrExit(err);
	if (mpCallback != nullptr)
	{
	mpCallback->OnResponse(this, attributePath, statusIB);
	}
	}

	exit:
	return err;
	}

	CHIP_ERROR WriteClient::HandleTimedStatus(const PayloadHeader & aPayloadHeader, System::PacketBufferHandle && aPayload)
	{
	ReturnErrorOnFailure(TimedRequest::HandleResponse(aPayloadHeader, std::move(aPayload)));

	return SendWriteRequest();
	}

	} // namespace app
	} // namespace chip