/**
 * Copyright (c) 2026 Huawei Technologies Co., Ltd.
 * This program is free software, you can redistribute it and/or modify it under the terms and conditions of
 * CANN Open Software License Agreement Version 2.0 (the "License").
 * Please refer to the License for details. You may not use this file except in compliance with the License.
 * THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
 * INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
 * See LICENSE in the root of the software repository for the full text of the License.
 */

#include "aclnn_neg.h"
#include "neg.h"
#include "aclnn_kernels/cast.h"
#include "aclnn_kernels/contiguous.h"
#include "aclnn_kernels/common/op_error_check.h"
#include "opdev/common_types.h"
#include "opdev/data_type_utils.h"
#include "opdev/format_utils.h"
#include "opdev/op_dfx.h"
#include "opdev/op_executor.h"
#include "opdev/op_log.h"
#include "opdev/shape_utils.h"
#include "opdev/tensor_view_utils.h"
#include "opdev/platform.h"
#include "platform/soc_spec.h"

using namespace op;
#ifdef __cplusplus
extern "C" {
#endif

static const std::initializer_list<op::DataType> dtype_support_list = {
    op::DataType::DT_INT8,  op::DataType::DT_INT32,  op::DataType::DT_INT64,     op::DataType::DT_FLOAT16,
    op::DataType::DT_FLOAT, op::DataType::DT_DOUBLE, op::DataType::DT_COMPLEX64, op::DataType::DT_COMPLEX128};

static const std::initializer_list<op::DataType> dtype_support_list_afterV200 = {
    op::DataType::DT_INT8,      op::DataType::DT_INT32,      op::DataType::DT_INT64,
    op::DataType::DT_FLOAT16,   op::DataType::DT_FLOAT,      op::DataType::DT_DOUBLE,
    op::DataType::DT_COMPLEX64, op::DataType::DT_COMPLEX128, op::DataType::DT_BF16};

static bool CheckNotNull(const aclTensor* self, const aclTensor* out)
{
    OP_CHECK_NULL(self, return false);
    OP_CHECK_NULL(out, return false);

    return true;
}

static bool CheckDtypeValid(const aclTensor* self)
{
    bool isAfterV200 =
        (GetCurrentPlatformInfo().GetCurNpuArch() == NpuArch::DAV_2201 ||
         GetCurrentPlatformInfo().GetCurNpuArch() == NpuArch::DAV_3510);
    bool isSupport = isAfterV200 ? CheckType(self->GetDataType(), dtype_support_list_afterV200) :
                                   CheckType(self->GetDataType(), dtype_support_list);
    if (!isSupport) {
        OP_LOGE(
            ACLNN_ERR_PARAM_INVALID, "Self dtype %s should be in dtype support list %s.",
            ToString(self->GetDataType()).GetString(),
            isAfterV200 ? ToString(dtype_support_list_afterV200).GetString() :
                          ToString(dtype_support_list).GetString());
        return false;
    }
    return true;
}

static bool CheckPromoteType(const aclTensor* self, const aclTensor* out)
{
    OP_CHECK_RESULT_DTYPE_CAST_FAILED(self->GetDataType(), out->GetDataType(), return false);

    return true;
}

static bool CheckShape(const aclTensor* self, const aclTensor* out)
{
    const size_t MAX_DIM = 8;
    OP_CHECK_MAX_DIM(self, MAX_DIM, return false);
    OP_CHECK_SHAPE_NOT_EQUAL(self, out, return false);
    return true;
}

static aclnnStatus CheckParams(const aclTensor* self, const aclTensor* out)
{
    // 1. 检查参数是否为空指针
    CHECK_RET(CheckNotNull(self, out), ACLNN_ERR_PARAM_NULLPTR);

    // 2. 检查输入的数据类型是否在API支持的数据类型范围之内，需要根据api定义校验
    CHECK_RET(CheckDtypeValid(self), ACLNN_ERR_PARAM_INVALID);

    // 3. 检查self的数据类型能否转换为输出数据类型
    CHECK_RET(CheckPromoteType(self, out), ACLNN_ERR_PARAM_INVALID);

    // 4. 检查shape
    CHECK_RET(CheckShape(self, out), ACLNN_ERR_PARAM_INVALID);

    return ACLNN_SUCCESS;
}

aclnnStatus aclnnNegGetWorkspaceSize(
    const aclTensor* self, aclTensor* out, uint64_t* workspaceSize, aclOpExecutor** executor)
{
    L2_DFX_PHASE_1(aclnnNeg, DFX_IN(self), DFX_OUT(out));
    // 固定写法，参数检查
    auto ret = CheckParams(self, out);
    CHECK_RET(ret == ACLNN_SUCCESS, ret);
    // 固定写法，创建OpExecutor
    auto uniqueExecutor = CREATE_EXECUTOR();
    CHECK_RET(uniqueExecutor.get() != nullptr, ACLNN_ERR_INNER_CREATE_EXECUTOR);
    // neg算子的空tensor在kernel中支持
    if (self->IsEmpty()) {
        *workspaceSize = 0;
        uniqueExecutor.ReleaseTo(executor);
        return ACLNN_SUCCESS;
    }
    // 固定写法，将输入self转换成连续的tensor
    auto selfContiguous = l0op::Contiguous(self, uniqueExecutor.get());
    CHECK_RET(selfContiguous != nullptr, ACLNN_ERR_INNER_NULLPTR);
    // 调用Neg算子kernel
    auto negOpOut = l0op::Neg(selfContiguous, uniqueExecutor.get());
    CHECK_RET(negOpOut != nullptr, ACLNN_ERR_INNER_NULLPTR);
    // 固定写法，将计算结果转换成输出out的数据类型
    auto castOut = l0op::Cast(negOpOut, out->GetDataType(), uniqueExecutor.get());
    CHECK_RET(castOut != nullptr, ACLNN_ERR_INNER_NULLPTR);
    // 固定写法，将计算结果拷贝到输出out上，out可能是非连续的tensor
    auto viewCopyResult = l0op::ViewCopy(castOut, out, uniqueExecutor.get());
    CHECK_RET(viewCopyResult != nullptr, ACLNN_ERR_INNER_NULLPTR);
    // 固定写法，获取计算过程中需要使用的workspace大小
    *workspaceSize = uniqueExecutor->GetWorkspaceSize();
    uniqueExecutor.ReleaseTo(executor);
    return ACLNN_SUCCESS;
}

aclnnStatus aclnnInplaceNegGetWorkspaceSize(aclTensor* selfRef, uint64_t* workspaceSize, aclOpExecutor** executor)
{
    return aclnnNegGetWorkspaceSize(selfRef, selfRef, workspaceSize, executor);
}

aclnnStatus aclnnNeg(void* workspace, uint64_t workspaceSize, aclOpExecutor* executor, const aclrtStream stream)
{
    L2_DFX_PHASE_2(aclnnNeg);
    // 固定写法，调用框架能力，完成计算
    return CommonOpExecutorRun(workspace, workspaceSize, executor, stream);
}

aclnnStatus aclnnInplaceNeg(void* workspace, uint64_t workspaceSize, aclOpExecutor* executor, aclrtStream stream)
{
    L2_DFX_PHASE_2(aclnnInplaceNeg);
    // 固定写法，调用框架能力，完成计算
    return CommonOpExecutorRun(workspace, workspaceSize, executor, stream);
}

#ifdef __cplusplus
}
#endif