src/app/AttributeAccessInterface.h - third_party/github/project-chip/connectedhomeip - Git at Google

 /*
  *
  *    Copyright (c) 2021-2022 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.
  */

 #pragma once

 #include <access/SubjectDescriptor.h>
 #include <app/ConcreteAttributePath.h>
 #include <app/MessageDef/AttributeReportIBs.h>
 #include <app/data-model/DecodableList.h>
 #include <app/data-model/Decode.h>
 #include <app/data-model/Encode.h>
 #include <app/data-model/FabricScoped.h>
 #include <app/data-model/List.h> // So we can encode lists
 #include <app/data-model/TagBoundEncoder.h>
 #include <app/util/basic-types.h>
 #include <lib/core/Optional.h>
 #include <lib/core/TLV.h>
 #include <lib/support/logging/CHIPLogging.h>

 /**
  * Callback class that clusters can implement in order to interpose custom
  * attribute-handling logic.  An AttributeAccessInterface instance is associated
  * with some specific cluster.  A single instance may be used for a specific
  * endpoint or for all endpoints.
  *
  * Instances of AttributeAccessInterface that are registered via
  * registerAttributeAccessOverride will be consulted before taking the normal
  * attribute access codepath and can use that codepath as a fallback if desired.
  */
 namespace chip {
 namespace app {

 /**
  * The AttributeReportBuilder is a helper class for filling a single report in AttributeReportIBs.
  *
  * Possible usage of AttributeReportBuilder might be:
  *
  * AttributeReportBuilder builder;
  * ReturnErrorOnFailure(builder.PrepareAttribute(...));
  * ReturnErrorOnFailure(builder.Encode(...));
  * ReturnErrorOnFailure(builder.FinishAttribute());
  */
 class AttributeReportBuilder
 {
 public:
     /**
      * PrepareAttribute encodes the "header" part of an attribute report including the path and data version.
      * Path will be encoded according to section 10.5.4.3.1 in the spec.
      * Note: Only append is supported currently (encode a null list index), other operations won't encode a list index in the
      * attribute path field.
      * TODO: Add support for encoding a single element in the list (path with a valid list index).
      */
     CHIP_ERROR PrepareAttribute(AttributeReportIBs::Builder & aAttributeReportIBs, const ConcreteDataAttributePath & aPath,
                                 DataVersion aDataVersion);

     /**
      * FinishAttribute encodes the "footer" part of an attribute report (it closes the containers opened in PrepareAttribute)
      */
     CHIP_ERROR FinishAttribute(AttributeReportIBs::Builder & aAttributeReportIBs);

     /**
      * EncodeValue encodes the value field of the report, it should be called exactly once.
      */
     template <typename T, std::enable_if_t<!DataModel::IsFabricScoped<T>::value, bool> = true, typename... Ts>
     CHIP_ERROR EncodeValue(AttributeReportIBs::Builder & aAttributeReportIBs, TLV::Tag tag, T && item, Ts &&... aArgs)
     {
         return DataModel::Encode(*(aAttributeReportIBs.GetAttributeReport().GetAttributeData().GetWriter()), tag, item,
                                  std::forward<Ts>(aArgs)...);
     }

     template <typename T, std::enable_if_t<DataModel::IsFabricScoped<T>::value, bool> = true, typename... Ts>
     CHIP_ERROR EncodeValue(AttributeReportIBs::Builder & aAttributeReportIBs, TLV::Tag tag, FabricIndex accessingFabricIndex,
                            T && item, Ts &&... aArgs)
     {
         return DataModel::EncodeForRead(*(aAttributeReportIBs.GetAttributeReport().GetAttributeData().GetWriter()), tag,
                                         accessingFabricIndex, item, std::forward<Ts>(aArgs)...);
     }
 };

 /**
  * The AttributeValueEncoder is a helper class for filling report payloads into AttributeReportIBs.
  * The attribute value encoder can be initialized with a AttributeEncodeState for saving and recovering its state between encode
  * sessions (chunkings).
  *
  * When Encode returns recoverable errors (e.g. CHIP_ERROR_NO_MEMORY) the state can be used to initialize the AttributeValueEncoder
  * for future use on the same attribute path.
  */
 class AttributeValueEncoder
 {
 public:
     class ListEncodeHelper
     {
     public:
         ListEncodeHelper(AttributeValueEncoder & encoder) : mAttributeValueEncoder(encoder) {}

         template <typename T, std::enable_if_t<DataModel::IsFabricScoped<T>::value, bool> = true>
         CHIP_ERROR Encode(T && aArg) const
         {
             VerifyOrReturnError(aArg.GetFabricIndex() != kUndefinedFabricIndex, CHIP_ERROR_INVALID_FABRIC_INDEX);

             // If we are encoding for a fabric filtered attribute read and the fabric index does not match that present in the
             // request, skip encoding this list item.
             VerifyOrReturnError(!mAttributeValueEncoder.mIsFabricFiltered ||
                                     aArg.GetFabricIndex() == mAttributeValueEncoder.mAccessingFabricIndex,
                                 CHIP_NO_ERROR);
             return mAttributeValueEncoder.EncodeListItem(mAttributeValueEncoder.mAccessingFabricIndex, std::forward<T>(aArg));
         }

         template <typename T, std::enable_if_t<!DataModel::IsFabricScoped<T>::value, bool> = true>
         CHIP_ERROR Encode(T && aArg) const
         {
             return mAttributeValueEncoder.EncodeListItem(std::forward<T>(aArg));
         }

     private:
         AttributeValueEncoder & mAttributeValueEncoder;
     };

     class AttributeEncodeState
     {
     public:
         AttributeEncodeState() : mAllowPartialData(false), mCurrentEncodingListIndex(kInvalidListIndex) {}
         bool AllowPartialData() const { return mAllowPartialData; }

     private:
         friend class AttributeValueEncoder;
         /**
          * When an attempt to encode an attribute returns an error, the buffer may contain tailing dirty data
          * (since the put was aborted).  The report engine normally rolls back the buffer to right before encoding
          * of the attribute started on errors.
          *
          * When chunking a list, EncodeListItem will atomically encode list items, ensuring that the
          * state of the buffer is valid to send (i.e. contains no trailing garbage), and return an error
          * if the list doesn't entirely fit.  In this situation, mAllowPartialData is set to communicate to the
          * report engine that it should not roll back the list items.
          *
          * TODO: There might be a better name for this variable.
          */
         bool mAllowPartialData = false;
         /**
          * If set to kInvalidListIndex, indicates that we have not encoded any data for the list yet and
          * need to start by encoding an empty list before we start encoding any list items.
          *
          * When set to a valid ListIndex value, indicates the index of the next list item that needs to be
          * encoded (i.e. the count of items encoded so far).
          */
         ListIndex mCurrentEncodingListIndex = kInvalidListIndex;
     };

     AttributeValueEncoder(AttributeReportIBs::Builder & aAttributeReportIBsBuilder, FabricIndex aAccessingFabricIndex,
                           const ConcreteAttributePath & aPath, DataVersion aDataVersion, bool aIsFabricFiltered = false,
                           const AttributeEncodeState & aState = AttributeEncodeState()) :
         mAttributeReportIBsBuilder(aAttributeReportIBsBuilder),
         mAccessingFabricIndex(aAccessingFabricIndex), mPath(aPath.mEndpointId, aPath.mClusterId, aPath.mAttributeId),
         mDataVersion(aDataVersion), mIsFabricFiltered(aIsFabricFiltered), mEncodeState(aState)
     {}

     /**
      * Encode a single value.  This value will not be chunked; it will either be
      * entirely encoded or fail to be encoded.  Consumers are allowed to make
      * either one call to Encode or one call to EncodeList to handle a read.
      */
     template <typename... Ts>
     CHIP_ERROR Encode(Ts &&... aArgs)
     {
         mTriedEncode = true;
         return EncodeAttributeReportIB(std::forward<Ts>(aArgs)...);
     }

     /**
      * Encode an explicit null value.
      */
     CHIP_ERROR EncodeNull()
     {
         // Doesn't matter what type Nullable we use here.
         return Encode(DataModel::Nullable<uint8_t>());
     }

     /**
      * Encode an explicit empty list.
      */
     CHIP_ERROR EncodeEmptyList()
     {
         // Doesn't matter what type List we use here.
         return Encode(DataModel::List<uint8_t>());
     }

     /**
      * aCallback is expected to take a const auto & argument and Encode() on it as many times as needed to encode all the list
      * elements one by one.  If any of those Encode() calls returns failure, aCallback must stop encoding and return failure.  When
      * all items are encoded aCallback is expected to return success.
      *
      * aCallback may not be called.  Consumers must not assume it will be called.
      *
      * When EncodeList returns an error, the consumers must abort the encoding, and return the exact error to the caller.
      *
      * TODO: Can we hold a error state in the AttributeValueEncoder itself so functions in ember-compatibility-functions don't have
      * to rely on the above assumption?
      *
      * Consumers are allowed to make either one call to EncodeList or one call to Encode to handle a read.
      *
      */
     template <typename ListGenerator>
     CHIP_ERROR EncodeList(ListGenerator aCallback)
     {
         mTriedEncode = true;
         // Spec 10.5.4.3.1, 10.5.4.6 (Replace a list w/ Multiple IBs)
         // EmptyList acts as the beginning of the whole array type attribute report.
         // An empty list is encoded iff both mCurrentEncodingListIndex and mEncodeState.mCurrentEncodingListIndex are invalid
         // values. After encoding the empty list, mEncodeState.mCurrentEncodingListIndex and mCurrentEncodingListIndex are set to 0.
         ReturnErrorOnFailure(EnsureListStarted());
         CHIP_ERROR err = aCallback(ListEncodeHelper(*this));

         // Even if encoding list items failed, make sure we EnsureListEnded().
         // Since we encode list items atomically, in the case when we just
         // didn't fit the next item we want to make sure our list is properly
         // ended before the reporting engine starts chunking.
         EnsureListEnded();
         if (err == CHIP_NO_ERROR)
         {
             // The Encode procedure finished without any error, clear the state.
             mEncodeState = AttributeEncodeState();
         }
         return err;
     }

     bool TriedEncode() const { return mTriedEncode; }

     /**
      * The accessing fabric index for this read or subscribe interaction.
      */
     FabricIndex AccessingFabricIndex() const { return mAccessingFabricIndex; }

     /**
      * AttributeValueEncoder is a short lived object, and the state is persisted by mEncodeState and restored by constructor.
      */
     const AttributeEncodeState & GetState() const { return mEncodeState; }

 private:
     // We made EncodeListItem() private, and ListEncoderHelper will expose it by Encode()
     friend class ListEncodeHelper;

     template <typename... Ts>
     CHIP_ERROR EncodeListItem(Ts &&... aArgs)
     {
         // EncodeListItem must be called after EnsureListStarted(), thus mCurrentEncodingListIndex and
         // mEncodeState.mCurrentEncodingListIndex are not invalid values.
         if (mCurrentEncodingListIndex < mEncodeState.mCurrentEncodingListIndex)
         {
             // We have encoded this element in previous chunks, skip it.
             mCurrentEncodingListIndex++;
             return CHIP_NO_ERROR;
         }

         TLV::TLVWriter backup;
         mAttributeReportIBsBuilder.Checkpoint(backup);

         CHIP_ERROR err;
         if (mEncodingInitialList)
         {
             // Just encode a single item, with an anonymous tag.
             AttributeReportBuilder builder;
             err = builder.EncodeValue(mAttributeReportIBsBuilder, TLV::AnonymousTag(), std::forward<Ts>(aArgs)...);
         }
         else
         {
             err = EncodeAttributeReportIB(std::forward<Ts>(aArgs)...);
         }
         if (err != CHIP_NO_ERROR)
         {
             // For list chunking, ReportEngine should not rollback the buffer when CHIP_ERROR_NO_MEMORY or similar error occurred.
             // However, the error might be raised in the middle of encoding procedure, then the buffer may contain partial data,
             // unclosed containers etc. This line clears all possible partial data and makes EncodeListItem is atomic.
             mAttributeReportIBsBuilder.Rollback(backup);
             return err;
         }

         mCurrentEncodingListIndex++;
         mEncodeState.mCurrentEncodingListIndex++;
         return CHIP_NO_ERROR;
     }

     /**
      * Builds a single AttributeReportIB in AttributeReportIBs.  The caller is
      * responsible for setting up mPath correctly.
      *
      * In particular, when we are encoding a single element in the list, mPath
      * must indicate a null list index to represent an "append" operation.
      * operation.
      */
     template <typename... Ts>
     CHIP_ERROR EncodeAttributeReportIB(Ts &&... aArgs)
     {
         AttributeReportBuilder builder;
         ReturnErrorOnFailure(builder.PrepareAttribute(mAttributeReportIBsBuilder, mPath, mDataVersion));
         ReturnErrorOnFailure(builder.EncodeValue(mAttributeReportIBsBuilder, TLV::ContextTag(AttributeDataIB::Tag::kData),
                                                  std::forward<Ts>(aArgs)...));

         return builder.FinishAttribute(mAttributeReportIBsBuilder);
     }

     /**
      * EnsureListStarted sets our mCurrentEncodingListIndex to 0, and:
      *
      * * If we are just starting the list, gets us ready to encode list items.
      *
      * * If we are continuing a chunked list, guarantees that mPath.mListOp is
      *   AppendItem after it returns.
      */
     CHIP_ERROR EnsureListStarted();

     /**
      * EnsureListEnded writes out the end of the list and our attribute data IB,
      * if we were encoding our initial list
      */
     void EnsureListEnded();

     bool mTriedEncode = false;
     AttributeReportIBs::Builder & mAttributeReportIBsBuilder;
     const FabricIndex mAccessingFabricIndex;
     ConcreteDataAttributePath mPath;
     DataVersion mDataVersion;
     bool mIsFabricFiltered = false;
     // mEncodingInitialList is true if we're encoding a list and we have not
     // started chunking it yet, so we're encoding a single attribute report IB
     // for the whole list, not one per item.
     bool mEncodingInitialList = false;
     AttributeEncodeState mEncodeState;
     ListIndex mCurrentEncodingListIndex = kInvalidListIndex;
 };

 class AttributeValueDecoder
 {
 public:
     AttributeValueDecoder(TLV::TLVReader & aReader, const Access::SubjectDescriptor & aSubjectDescriptor) :
         mReader(aReader), mSubjectDescriptor(aSubjectDescriptor)
     {}

     template <typename T, typename std::enable_if_t<!DataModel::IsFabricScoped<T>::value, bool> = true>
     CHIP_ERROR Decode(T & aArg)
     {
         mTriedDecode = true;
         return DataModel::Decode(mReader, aArg);
     }

     template <typename T, typename std::enable_if_t<DataModel::IsFabricScoped<T>::value, bool> = true>
     CHIP_ERROR Decode(T & aArg)
     {
         mTriedDecode = true;
         // The WriteRequest comes with no fabric index, this will happen when receiving a write request on a PASE session before
         // AddNOC.
         VerifyOrReturnError(AccessingFabricIndex() != kUndefinedFabricIndex, CHIP_IM_GLOBAL_STATUS(UnsupportedAccess));
         ReturnErrorOnFailure(DataModel::Decode(mReader, aArg));
         aArg.SetFabricIndex(AccessingFabricIndex());
         return CHIP_NO_ERROR;
     }

     bool TriedDecode() const { return mTriedDecode; }

     /**
      * The accessing fabric index for this write interaction.
      */
     FabricIndex AccessingFabricIndex() const { return mSubjectDescriptor.fabricIndex; }

     /**
      * The accessing subject descriptor for this write interaction.
      */
     const Access::SubjectDescriptor & GetSubjectDescriptor() const { return mSubjectDescriptor; }

 private:
     TLV::TLVReader & mReader;
     bool mTriedDecode = false;
     const Access::SubjectDescriptor mSubjectDescriptor;
 };

 class AttributeAccessInterface
 {
 public:
     /**
      * aEndpointId can be Missing to indicate that this object is meant to be
      * used with all endpoints.
      */
     AttributeAccessInterface(Optional<EndpointId> aEndpointId, ClusterId aClusterId) :
         mEndpointId(aEndpointId), mClusterId(aClusterId)
     {}
     virtual ~AttributeAccessInterface() {}

     /**
      * Callback for reading attributes.
      *
      * @param [in] aPath indicates which exact data is being read.
      * @param [in] aEncoder the AttributeValueEncoder to use for encoding the
      *             data.
      *
      * The implementation can do one of three things:
      *
      * 1) Return a failure.  This is treated as a failed read and the error is
      *    returned to the client, by converting it to a StatusIB.
      * 2) Return success and attempt to encode data using aEncoder.  The data is
      *    returned to the client.
      * 3) Return success and not attempt to encode any data using aEncoder.  In
      *    this case, Ember attribute access will happen for the read. This may
      *    involve reading from the attribute store or external attribute
      *    callbacks.
      */
     virtual CHIP_ERROR Read(const ConcreteReadAttributePath & aPath, AttributeValueEncoder & aEncoder) = 0;

     /**
      * Callback for writing attributes.
      *
      * @param [in] aPath indicates which exact data is being written.
      * @param [in] aDecoder the AttributeValueDecoder to use for decoding the
      *             data.
      *
      * The implementation can do one of three things:
      *
      * 1) Return a failure.  This is treated as a failed write and the error is
      *    sent to the client, by converting it to a StatusIB.
      * 2) Return success and attempt to decode from aDecoder.  This is
      *    treated as a successful write.
      * 3) Return success and not attempt to decode from aDecoder.  In
      *    this case, Ember attribute access will happen for the write. This may
      *    involve writing to the attribute store or external attribute
      *    callbacks.
      */
     virtual CHIP_ERROR Write(const ConcreteDataAttributePath & aPath, AttributeValueDecoder & aDecoder) { return CHIP_NO_ERROR; }

     /**
      * Indicates the start of a series of list operations. This function will be called before the first Write operation of a series
      * of consequence attribute data of the same attribute.
      *
      * 1) This function will be called if the client tries to set a nullable list attribute to null.
      * 2) This function will only be called once for a series of consequent attribute data (regardless the kind of list operation)
      * of the same attribute.
      *
      * @param [in] aPath indicates the path of the modified list.
      */
     virtual void OnListWriteBegin(const ConcreteAttributePath & aPath) {}

     /**
      * Indicates the end of a series of list operations. This function will be called after the last Write operation of a series
      * of consequence attribute data of the same attribute.
      *
      * 1) This function will be called if the client tries to set a nullable list attribute to null.
      * 2) This function will only be called once for a series of consequent attribute data (regardless the kind of list operation)
      * of the same attribute.
      * 3) When aWriteWasSuccessful is true, the data written must be consistent or the list is untouched.
      *
      * @param [in] aPath indicates the path of the modified list
      * @param [in] aWriteWasSuccessful indicates whether the delivered list is complete.
      *
      */
     virtual void OnListWriteEnd(const ConcreteAttributePath & aPath, bool aWriteWasSuccessful) {}

     /**
      * Mechanism for keeping track of a chain of AttributeAccessInterfaces.
      */
     void SetNext(AttributeAccessInterface * aNext) { mNext = aNext; }
     AttributeAccessInterface * GetNext() const { return mNext; }

     /**
      * Check whether a this AttributeAccessInterface is relevant for a
      * particular endpoint+cluster.  An AttributeAccessInterface will be used
      * for a read from a particular cluster only when this function returns
      * true.
      */
     bool Matches(EndpointId aEndpointId, ClusterId aClusterId) const
     {
         return (!mEndpointId.HasValue() || mEndpointId.Value() == aEndpointId) && mClusterId == aClusterId;
     }

     /**
      * Check whether an AttributeAccessInterface is relevant for a particular
      * specific endpoint.  This is used to clean up overrides registered for an
      * endpoint that becomes disabled.
      */
     bool MatchesEndpoint(EndpointId aEndpointId) const { return mEndpointId.HasValue() && mEndpointId.Value() == aEndpointId; }

     /**
      * Check whether another AttributeAccessInterface wants to handle the same set of
      * attributes as we do.
      */
     bool Matches(const AttributeAccessInterface & aOther) const
     {
         return mClusterId == aOther.mClusterId &&
             (!mEndpointId.HasValue() || !aOther.mEndpointId.HasValue() || mEndpointId.Value() == aOther.mEndpointId.Value());
     }

 private:
     Optional<EndpointId> mEndpointId;
     ClusterId mClusterId;
     AttributeAccessInterface * mNext = nullptr;
 };

 } // namespace app
 } // namespace chip
	/*
	*
	* Copyright (c) 2021-2022 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.
	*/

	#pragma once

	#include <access/SubjectDescriptor.h>
	#include <app/ConcreteAttributePath.h>
	#include <app/MessageDef/AttributeReportIBs.h>
	#include <app/data-model/DecodableList.h>
	#include <app/data-model/Decode.h>
	#include <app/data-model/Encode.h>
	#include <app/data-model/FabricScoped.h>
	#include <app/data-model/List.h> // So we can encode lists
	#include <app/data-model/TagBoundEncoder.h>
	#include <app/util/basic-types.h>
	#include <lib/core/Optional.h>
	#include <lib/core/TLV.h>
	#include <lib/support/logging/CHIPLogging.h>

	/**
	* Callback class that clusters can implement in order to interpose custom
	* attribute-handling logic. An AttributeAccessInterface instance is associated
	* with some specific cluster. A single instance may be used for a specific
	* endpoint or for all endpoints.
	*
	* Instances of AttributeAccessInterface that are registered via
	* registerAttributeAccessOverride will be consulted before taking the normal
	* attribute access codepath and can use that codepath as a fallback if desired.
	*/
	namespace chip {
	namespace app {

	/**
	* The AttributeReportBuilder is a helper class for filling a single report in AttributeReportIBs.
	*
	* Possible usage of AttributeReportBuilder might be:
	*
	* AttributeReportBuilder builder;
	* ReturnErrorOnFailure(builder.PrepareAttribute(...));
	* ReturnErrorOnFailure(builder.Encode(...));
	* ReturnErrorOnFailure(builder.FinishAttribute());
	*/
	class AttributeReportBuilder
	{
	public:
	/**
	* PrepareAttribute encodes the "header" part of an attribute report including the path and data version.
	* Path will be encoded according to section 10.5.4.3.1 in the spec.
	* Note: Only append is supported currently (encode a null list index), other operations won't encode a list index in the
	* attribute path field.
	* TODO: Add support for encoding a single element in the list (path with a valid list index).
	*/
	CHIP_ERROR PrepareAttribute(AttributeReportIBs::Builder & aAttributeReportIBs, const ConcreteDataAttributePath & aPath,
	DataVersion aDataVersion);

	/**
	* FinishAttribute encodes the "footer" part of an attribute report (it closes the containers opened in PrepareAttribute)
	*/
	CHIP_ERROR FinishAttribute(AttributeReportIBs::Builder & aAttributeReportIBs);

	/**
	* EncodeValue encodes the value field of the report, it should be called exactly once.
	*/
	template <typename T, std::enable_if_t<!DataModel::IsFabricScoped<T>::value, bool> = true, typename... Ts>
	CHIP_ERROR EncodeValue(AttributeReportIBs::Builder & aAttributeReportIBs, TLV::Tag tag, T && item, Ts &&... aArgs)
	{
	return DataModel::Encode(*(aAttributeReportIBs.GetAttributeReport().GetAttributeData().GetWriter()), tag, item,
	std::forward<Ts>(aArgs)...);
	}

	template <typename T, std::enable_if_t<DataModel::IsFabricScoped<T>::value, bool> = true, typename... Ts>
	CHIP_ERROR EncodeValue(AttributeReportIBs::Builder & aAttributeReportIBs, TLV::Tag tag, FabricIndex accessingFabricIndex,
	T && item, Ts &&... aArgs)
	{
	return DataModel::EncodeForRead(*(aAttributeReportIBs.GetAttributeReport().GetAttributeData().GetWriter()), tag,
	accessingFabricIndex, item, std::forward<Ts>(aArgs)...);
	}
	};

	/**
	* The AttributeValueEncoder is a helper class for filling report payloads into AttributeReportIBs.
	* The attribute value encoder can be initialized with a AttributeEncodeState for saving and recovering its state between encode
	* sessions (chunkings).
	*
	* When Encode returns recoverable errors (e.g. CHIP_ERROR_NO_MEMORY) the state can be used to initialize the AttributeValueEncoder
	* for future use on the same attribute path.
	*/
	class AttributeValueEncoder
	{
	public:
	class ListEncodeHelper
	{
	public:
	ListEncodeHelper(AttributeValueEncoder & encoder) : mAttributeValueEncoder(encoder) {}

	template <typename T, std::enable_if_t<DataModel::IsFabricScoped<T>::value, bool> = true>
	CHIP_ERROR Encode(T && aArg) const
	{
	VerifyOrReturnError(aArg.GetFabricIndex() != kUndefinedFabricIndex, CHIP_ERROR_INVALID_FABRIC_INDEX);

	// If we are encoding for a fabric filtered attribute read and the fabric index does not match that present in the
	// request, skip encoding this list item.
	VerifyOrReturnError(!mAttributeValueEncoder.mIsFabricFiltered \|\|
	aArg.GetFabricIndex() == mAttributeValueEncoder.mAccessingFabricIndex,
	CHIP_NO_ERROR);
	return mAttributeValueEncoder.EncodeListItem(mAttributeValueEncoder.mAccessingFabricIndex, std::forward<T>(aArg));
	}

	template <typename T, std::enable_if_t<!DataModel::IsFabricScoped<T>::value, bool> = true>
	CHIP_ERROR Encode(T && aArg) const
	{
	return mAttributeValueEncoder.EncodeListItem(std::forward<T>(aArg));
	}

	private:
	AttributeValueEncoder & mAttributeValueEncoder;
	};

	class AttributeEncodeState
	{
	public:
	AttributeEncodeState() : mAllowPartialData(false), mCurrentEncodingListIndex(kInvalidListIndex) {}
	bool AllowPartialData() const { return mAllowPartialData; }

	private:
	friend class AttributeValueEncoder;
	/**
	* When an attempt to encode an attribute returns an error, the buffer may contain tailing dirty data
	* (since the put was aborted). The report engine normally rolls back the buffer to right before encoding
	* of the attribute started on errors.
	*
	* When chunking a list, EncodeListItem will atomically encode list items, ensuring that the
	* state of the buffer is valid to send (i.e. contains no trailing garbage), and return an error
	* if the list doesn't entirely fit. In this situation, mAllowPartialData is set to communicate to the
	* report engine that it should not roll back the list items.
	*
	* TODO: There might be a better name for this variable.
	*/
	bool mAllowPartialData = false;
	/**
	* If set to kInvalidListIndex, indicates that we have not encoded any data for the list yet and
	* need to start by encoding an empty list before we start encoding any list items.
	*
	* When set to a valid ListIndex value, indicates the index of the next list item that needs to be
	* encoded (i.e. the count of items encoded so far).
	*/
	ListIndex mCurrentEncodingListIndex = kInvalidListIndex;
	};

	AttributeValueEncoder(AttributeReportIBs::Builder & aAttributeReportIBsBuilder, FabricIndex aAccessingFabricIndex,
	const ConcreteAttributePath & aPath, DataVersion aDataVersion, bool aIsFabricFiltered = false,
	const AttributeEncodeState & aState = AttributeEncodeState()) :
	mAttributeReportIBsBuilder(aAttributeReportIBsBuilder),
	mAccessingFabricIndex(aAccessingFabricIndex), mPath(aPath.mEndpointId, aPath.mClusterId, aPath.mAttributeId),
	mDataVersion(aDataVersion), mIsFabricFiltered(aIsFabricFiltered), mEncodeState(aState)
	{}

	/**
	* Encode a single value. This value will not be chunked; it will either be
	* entirely encoded or fail to be encoded. Consumers are allowed to make
	* either one call to Encode or one call to EncodeList to handle a read.
	*/
	template <typename... Ts>
	CHIP_ERROR Encode(Ts &&... aArgs)
	{
	mTriedEncode = true;
	return EncodeAttributeReportIB(std::forward<Ts>(aArgs)...);
	}

	/**
	* Encode an explicit null value.
	*/
	CHIP_ERROR EncodeNull()
	{
	// Doesn't matter what type Nullable we use here.
	return Encode(DataModel::Nullable<uint8_t>());
	}

	/**
	* Encode an explicit empty list.
	*/
	CHIP_ERROR EncodeEmptyList()
	{
	// Doesn't matter what type List we use here.
	return Encode(DataModel::List<uint8_t>());
	}

	/**
	* aCallback is expected to take a const auto & argument and Encode() on it as many times as needed to encode all the list
	* elements one by one. If any of those Encode() calls returns failure, aCallback must stop encoding and return failure. When
	* all items are encoded aCallback is expected to return success.
	*
	* aCallback may not be called. Consumers must not assume it will be called.
	*
	* When EncodeList returns an error, the consumers must abort the encoding, and return the exact error to the caller.
	*
	* TODO: Can we hold a error state in the AttributeValueEncoder itself so functions in ember-compatibility-functions don't have
	* to rely on the above assumption?
	*
	* Consumers are allowed to make either one call to EncodeList or one call to Encode to handle a read.
	*
	*/
	template <typename ListGenerator>
	CHIP_ERROR EncodeList(ListGenerator aCallback)
	{
	mTriedEncode = true;
	// Spec 10.5.4.3.1, 10.5.4.6 (Replace a list w/ Multiple IBs)
	// EmptyList acts as the beginning of the whole array type attribute report.
	// An empty list is encoded iff both mCurrentEncodingListIndex and mEncodeState.mCurrentEncodingListIndex are invalid
	// values. After encoding the empty list, mEncodeState.mCurrentEncodingListIndex and mCurrentEncodingListIndex are set to 0.
	ReturnErrorOnFailure(EnsureListStarted());
	CHIP_ERROR err = aCallback(ListEncodeHelper(*this));

	// Even if encoding list items failed, make sure we EnsureListEnded().
	// Since we encode list items atomically, in the case when we just
	// didn't fit the next item we want to make sure our list is properly
	// ended before the reporting engine starts chunking.
	EnsureListEnded();
	if (err == CHIP_NO_ERROR)
	{
	// The Encode procedure finished without any error, clear the state.
	mEncodeState = AttributeEncodeState();
	}
	return err;
	}

	bool TriedEncode() const { return mTriedEncode; }

	/**
	* The accessing fabric index for this read or subscribe interaction.
	*/
	FabricIndex AccessingFabricIndex() const { return mAccessingFabricIndex; }

	/**
	* AttributeValueEncoder is a short lived object, and the state is persisted by mEncodeState and restored by constructor.
	*/
	const AttributeEncodeState & GetState() const { return mEncodeState; }

	private:
	// We made EncodeListItem() private, and ListEncoderHelper will expose it by Encode()
	friend class ListEncodeHelper;

	template <typename... Ts>
	CHIP_ERROR EncodeListItem(Ts &&... aArgs)
	{
	// EncodeListItem must be called after EnsureListStarted(), thus mCurrentEncodingListIndex and
	// mEncodeState.mCurrentEncodingListIndex are not invalid values.
	if (mCurrentEncodingListIndex < mEncodeState.mCurrentEncodingListIndex)
	{
	// We have encoded this element in previous chunks, skip it.
	mCurrentEncodingListIndex++;
	return CHIP_NO_ERROR;
	}

	TLV::TLVWriter backup;
	mAttributeReportIBsBuilder.Checkpoint(backup);

	CHIP_ERROR err;
	if (mEncodingInitialList)
	{
	// Just encode a single item, with an anonymous tag.
	AttributeReportBuilder builder;
	err = builder.EncodeValue(mAttributeReportIBsBuilder, TLV::AnonymousTag(), std::forward<Ts>(aArgs)...);
	}
	else
	{
	err = EncodeAttributeReportIB(std::forward<Ts>(aArgs)...);
	}
	if (err != CHIP_NO_ERROR)
	{
	// For list chunking, ReportEngine should not rollback the buffer when CHIP_ERROR_NO_MEMORY or similar error occurred.
	// However, the error might be raised in the middle of encoding procedure, then the buffer may contain partial data,
	// unclosed containers etc. This line clears all possible partial data and makes EncodeListItem is atomic.
	mAttributeReportIBsBuilder.Rollback(backup);
	return err;
	}

	mCurrentEncodingListIndex++;
	mEncodeState.mCurrentEncodingListIndex++;
	return CHIP_NO_ERROR;
	}

	/**
	* Builds a single AttributeReportIB in AttributeReportIBs. The caller is
	* responsible for setting up mPath correctly.
	*
	* In particular, when we are encoding a single element in the list, mPath
	* must indicate a null list index to represent an "append" operation.
	* operation.
	*/
	template <typename... Ts>
	CHIP_ERROR EncodeAttributeReportIB(Ts &&... aArgs)
	{
	AttributeReportBuilder builder;
	ReturnErrorOnFailure(builder.PrepareAttribute(mAttributeReportIBsBuilder, mPath, mDataVersion));
	ReturnErrorOnFailure(builder.EncodeValue(mAttributeReportIBsBuilder, TLV::ContextTag(AttributeDataIB::Tag::kData),
	std::forward<Ts>(aArgs)...));

	return builder.FinishAttribute(mAttributeReportIBsBuilder);
	}

	/**
	* EnsureListStarted sets our mCurrentEncodingListIndex to 0, and:
	*
	* * If we are just starting the list, gets us ready to encode list items.
	*
	* * If we are continuing a chunked list, guarantees that mPath.mListOp is
	* AppendItem after it returns.
	*/
	CHIP_ERROR EnsureListStarted();

	/**
	* EnsureListEnded writes out the end of the list and our attribute data IB,
	* if we were encoding our initial list
	*/
	void EnsureListEnded();

	bool mTriedEncode = false;
	AttributeReportIBs::Builder & mAttributeReportIBsBuilder;
	const FabricIndex mAccessingFabricIndex;
	ConcreteDataAttributePath mPath;
	DataVersion mDataVersion;
	bool mIsFabricFiltered = false;
	// mEncodingInitialList is true if we're encoding a list and we have not
	// started chunking it yet, so we're encoding a single attribute report IB
	// for the whole list, not one per item.
	bool mEncodingInitialList = false;
	AttributeEncodeState mEncodeState;
	ListIndex mCurrentEncodingListIndex = kInvalidListIndex;
	};

	class AttributeValueDecoder
	{
	public:
	AttributeValueDecoder(TLV::TLVReader & aReader, const Access::SubjectDescriptor & aSubjectDescriptor) :
	mReader(aReader), mSubjectDescriptor(aSubjectDescriptor)
	{}

	template <typename T, typename std::enable_if_t<!DataModel::IsFabricScoped<T>::value, bool> = true>
	CHIP_ERROR Decode(T & aArg)
	{
	mTriedDecode = true;
	return DataModel::Decode(mReader, aArg);
	}

	template <typename T, typename std::enable_if_t<DataModel::IsFabricScoped<T>::value, bool> = true>
	CHIP_ERROR Decode(T & aArg)
	{
	mTriedDecode = true;
	// The WriteRequest comes with no fabric index, this will happen when receiving a write request on a PASE session before
	// AddNOC.
	VerifyOrReturnError(AccessingFabricIndex() != kUndefinedFabricIndex, CHIP_IM_GLOBAL_STATUS(UnsupportedAccess));
	ReturnErrorOnFailure(DataModel::Decode(mReader, aArg));
	aArg.SetFabricIndex(AccessingFabricIndex());
	return CHIP_NO_ERROR;
	}

	bool TriedDecode() const { return mTriedDecode; }

	/**
	* The accessing fabric index for this write interaction.
	*/
	FabricIndex AccessingFabricIndex() const { return mSubjectDescriptor.fabricIndex; }

	/**
	* The accessing subject descriptor for this write interaction.
	*/
	const Access::SubjectDescriptor & GetSubjectDescriptor() const { return mSubjectDescriptor; }

	private:
	TLV::TLVReader & mReader;
	bool mTriedDecode = false;
	const Access::SubjectDescriptor mSubjectDescriptor;
	};

	class AttributeAccessInterface
	{
	public:
	/**
	* aEndpointId can be Missing to indicate that this object is meant to be
	* used with all endpoints.
	*/
	AttributeAccessInterface(Optional<EndpointId> aEndpointId, ClusterId aClusterId) :
	mEndpointId(aEndpointId), mClusterId(aClusterId)
	{}
	virtual ~AttributeAccessInterface() {}

	/**
	* Callback for reading attributes.
	*
	* @param [in] aPath indicates which exact data is being read.
	* @param [in] aEncoder the AttributeValueEncoder to use for encoding the
	* data.
	*
	* The implementation can do one of three things:
	*
	* 1) Return a failure. This is treated as a failed read and the error is
	* returned to the client, by converting it to a StatusIB.
	* 2) Return success and attempt to encode data using aEncoder. The data is
	* returned to the client.
	* 3) Return success and not attempt to encode any data using aEncoder. In
	* this case, Ember attribute access will happen for the read. This may
	* involve reading from the attribute store or external attribute
	* callbacks.
	*/
	virtual CHIP_ERROR Read(const ConcreteReadAttributePath & aPath, AttributeValueEncoder & aEncoder) = 0;

	/**
	* Callback for writing attributes.
	*
	* @param [in] aPath indicates which exact data is being written.
	* @param [in] aDecoder the AttributeValueDecoder to use for decoding the
	* data.
	*
	* The implementation can do one of three things:
	*
	* 1) Return a failure. This is treated as a failed write and the error is
	* sent to the client, by converting it to a StatusIB.
	* 2) Return success and attempt to decode from aDecoder. This is
	* treated as a successful write.
	* 3) Return success and not attempt to decode from aDecoder. In
	* this case, Ember attribute access will happen for the write. This may
	* involve writing to the attribute store or external attribute
	* callbacks.
	*/
	virtual CHIP_ERROR Write(const ConcreteDataAttributePath & aPath, AttributeValueDecoder & aDecoder) { return CHIP_NO_ERROR; }

	/**
	* Indicates the start of a series of list operations. This function will be called before the first Write operation of a series
	* of consequence attribute data of the same attribute.
	*
	* 1) This function will be called if the client tries to set a nullable list attribute to null.
	* 2) This function will only be called once for a series of consequent attribute data (regardless the kind of list operation)
	* of the same attribute.
	*
	* @param [in] aPath indicates the path of the modified list.
	*/
	virtual void OnListWriteBegin(const ConcreteAttributePath & aPath) {}

	/**
	* Indicates the end of a series of list operations. This function will be called after the last Write operation of a series
	* of consequence attribute data of the same attribute.
	*
	* 1) This function will be called if the client tries to set a nullable list attribute to null.
	* 2) This function will only be called once for a series of consequent attribute data (regardless the kind of list operation)
	* of the same attribute.
	* 3) When aWriteWasSuccessful is true, the data written must be consistent or the list is untouched.
	*
	* @param [in] aPath indicates the path of the modified list
	* @param [in] aWriteWasSuccessful indicates whether the delivered list is complete.
	*
	*/
	virtual void OnListWriteEnd(const ConcreteAttributePath & aPath, bool aWriteWasSuccessful) {}

	/**
	* Mechanism for keeping track of a chain of AttributeAccessInterfaces.
	*/
	void SetNext(AttributeAccessInterface * aNext) { mNext = aNext; }
	AttributeAccessInterface * GetNext() const { return mNext; }

	/**
	* Check whether a this AttributeAccessInterface is relevant for a
	* particular endpoint+cluster. An AttributeAccessInterface will be used
	* for a read from a particular cluster only when this function returns
	* true.
	*/
	bool Matches(EndpointId aEndpointId, ClusterId aClusterId) const
	{
	return (!mEndpointId.HasValue() \|\| mEndpointId.Value() == aEndpointId) && mClusterId == aClusterId;
	}

	/**
	* Check whether an AttributeAccessInterface is relevant for a particular
	* specific endpoint. This is used to clean up overrides registered for an
	* endpoint that becomes disabled.
	*/
	bool MatchesEndpoint(EndpointId aEndpointId) const { return mEndpointId.HasValue() && mEndpointId.Value() == aEndpointId; }

	/**
	* Check whether another AttributeAccessInterface wants to handle the same set of
	* attributes as we do.
	*/
	bool Matches(const AttributeAccessInterface & aOther) const
	{
	return mClusterId == aOther.mClusterId &&
	(!mEndpointId.HasValue() \|\| !aOther.mEndpointId.HasValue() \|\| mEndpointId.Value() == aOther.mEndpointId.Value());
	}

	private:
	Optional<EndpointId> mEndpointId;
	ClusterId mClusterId;
	AttributeAccessInterface * mNext = nullptr;
	};

	} // namespace app
	} // namespace chip