/*
    Copyright (c) 2016, Taiga Nomi
    All rights reserved.

    Redistribution and use in source and binary forms, with or without
    modification, are permitted provided that the following conditions are met:
    * Redistributions of source code must retain the above copyright
    notice, this list of conditions and the following disclaimer.
    * Redistributions in binary form must reproduce the above copyright
    notice, this list of conditions and the following disclaimer in the
    documentation and/or other materials provided with the distribution.
    * Neither the name of the <organization> nor the
    names of its contributors may be used to endorse or promote products
    derived from this software without specific prior written permission.

    THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY
    EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
    WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
    DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY
    DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
    (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
    LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
    ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
    (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
    SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#pragma once
#include <string>
#include <iostream>
#include <cstdio>
 #include "gtest/gtest.h"
#include "tiny_dnn/tiny_dnn.h"

namespace tiny_dnn {

template <typename Container, typename T>
inline bool is_near_container(const Container& expected, const Container& actual, T abs_error) {
    auto i1 = std::begin(expected);
    auto i2 = std::begin(actual);

    for (; i1 != std::end(expected); ++i1, ++i2) {
        if(std::abs(*i1 - *i2) > abs_error) return false;
    }
    return true;
}

template <typename Container>
inline bool is_different_container(const Container& expected, const Container& actual) {
    auto i1 = std::begin(expected);
    auto i2 = std::begin(actual);

    for (; i1 != std::end(expected); ++i1, ++i2) {
        if (*i1 != *i2) return true;
    }
    return false;
}

inline bool exists(const std::string& path) {
    if (FILE *file = std::fopen(path.c_str(), "r")) {
        fclose(file);
        return true;
    } else {
        return false;
    }
}

inline std::string unique_path() {
    std::string pattern = "%%%%-%%%%-%%%%-%%%%";

    for (auto p = pattern.begin(); p != pattern.end(); ++p) {
        if (*p == '%') *p = (rand()%10)+'0';
    }
    return exists(pattern) ? unique_path() : pattern;
}

vec_t forward_pass(layer& src, const vec_t& vec) {
    src.setup(false);
    (*src.inputs()[0]->get_data())[0] = vec;
    src.forward();
    return src.output()[0][0];
}

template <typename N>
vec_t forward_pass(network<N>& net, const vec_t& vec) {
    return net.predict(vec);
}

template <typename T>
void serialization_test(T& src, T& dst)
{
    //EXPECT_FALSE(src.has_same_weights(dst, 1E-5));

    std::string tmp_file_path = unique_path();

    // write
    {
        std::ofstream ofs(tmp_file_path.c_str());
        src.save(ofs);
    }

    // read
    {
        std::ifstream ifs(tmp_file_path.c_str());
        dst.load(ifs);
    }

    std::remove(tmp_file_path.c_str());

    vec_t v(src.in_data_size());
    uniform_rand(v.begin(), v.end(), -1.0f, 1.0f);

    EXPECT_TRUE(src.has_same_weights(dst, 1E-5f));

    vec_t r1 = forward_pass(src, v);
    vec_t r2 = forward_pass(dst, v);

    EXPECT_TRUE(is_near_container(r1, r2, 1E-4f));
}


template <typename T>
void quantized_serialization_test(T& src, T& dst)
{
    //EXPECT_FALSE(src.has_same_weights(dst, 1E-5));

    std::string tmp_file_path = unique_path();

    // write
    {
        std::ofstream ofs(tmp_file_path.c_str());
        src.save(ofs);
    }

    // read
    {
        std::ifstream ifs(tmp_file_path.c_str());
        dst.load(ifs);
    }

    std::remove(tmp_file_path.c_str());

    vec_t v(src.in_data_size());
    uniform_rand(v.begin(), v.end(), -1.0, 1.0);

    EXPECT_TRUE(src.has_same_weights(dst, 1E-5));

    vec_t r1 = forward_pass(src, v);
    vec_t r2 = forward_pass(dst, v);

    EXPECT_TRUE(is_near_container(r1, r2, 1E-2));
}

template <typename T>
inline T epsilon() {
    return 0;
}

template <>
inline float epsilon() {
    return 1e-2f;
}

template <>
inline double epsilon() {
    return 1e-4;
}

inline bool resolve_path(const std::string& filename, std::string& path) {
    static const char* path_list[] = {
       "",
       "./test/",
       "../test/",
       "../../test/",
       "../../../test/",
       "../../tiny-dnn/test/",
       "../../../tiny-dnn/test/" 
    };

    for (size_t i = 0; i < sizeof(path_list) / sizeof(path_list[0]); i++) {
        if (exists(path_list[i] + filename)) {
            path = path_list[i] + filename;
            return true;
        }
    }
    return false;
}

namespace {
    std::pair<std::vector<tensor_t>, std::vector<std::vector<label_t>>> generate_gradient_check_data(
        serial_size_t input_dimension, serial_size_t sample_count = 5, serial_size_t class_count = 2)
    {
        const serial_size_t input_channel_count = 1;
        const serial_size_t output_channel_count = 1;
        std::vector<tensor_t> a(sample_count, tensor_t(input_channel_count, vec_t(input_dimension, 0.0)));
        std::vector<std::vector<label_t>> t(sample_count, std::vector<label_t>(output_channel_count));

        for (serial_size_t sample = 0; sample < sample_count; ++sample) {
            for (serial_size_t input_channel = 0; input_channel < input_channel_count; ++input_channel) {
                vec_t& v = a[sample][input_channel];
                uniform_rand(v.begin(), v.end(), -1, 1);
            }
            for (serial_size_t output_channel = 0; output_channel < output_channel_count; ++output_channel) {
                t[sample][output_channel] = sample % class_count;
            }
        }

        return std::make_pair(a, t);
    }
}

#ifndef CNN_NO_SERIALIZATION
inline std::string layer_to_json(const layer& src) {
    std::ostringstream os;
    {
        cereal::JSONOutputArchive oa(os);
        layer::save_layer(oa, src);
    }
    return os.str();
}

inline std::shared_ptr<layer> json_to_layer(const std::string& src) {
    std::stringstream ss;
    ss << src;
    cereal::JSONInputArchive oa(ss);
    return layer::load_layer(oa);
}
#endif

} // namespace tiny_dnn