Store std::string in std::map w/ and w/o lowercase process, and find arbitrary string using various methods (Insensitive case)

insensitive_string_match.cpp

#include <algorithm>
#include <cctype>
#include <chrono>
#include <fstream>
#include <iostream>
#include <map>
#include <random>
#include <string>
#include <vector>

int main()
{
	const int NUM_ITERATIONS = 100'000'00;

	// Prepare data to find arbitrary string
	std::ifstream input("test_case.txt");
	std::vector<std::string> vTestData;

	while (!input.eof())
	{
		std::string str;
		input >> str;
		vTestData.emplace_back(str);
	}

	input.close();

	// Prepare random
	auto seed = std::chrono::high_resolution_clock::now().time_since_epoch().count();
	std::mt19937 mt_rand(seed);

	// Case 1: Store std::string in std::map with lowercase process
	input.open("sample.txt");
	std::map<std::string, int> map1;

	std::cout << "Store std::string in std::map with lowercase process\n";
	auto start = std::chrono::high_resolution_clock::now();

	while (!input.eof())
	{
		std::string str;
		input >> str;
		std::transform(str.begin(), str.end(), str.begin(), ::tolower);

		map1.insert(std::make_pair(str, 1));
	}

	auto end = std::chrono::high_resolution_clock::now();
	auto time = std::chrono::duration_cast<std::chrono::milliseconds>(end - start);
	std::cout << time.count() << "\n";

	input.close();

	// Case 1-1: Find arbitrary string using for-statement
	// NOTE: auto is much slower than auto& because of copy cost
	std::cout << "Find arbitrary string using for-statement\n";
	start = std::chrono::high_resolution_clock::now();

	for (int i = 0; i < NUM_ITERATIONS; ++i)
	{
		std::string str = vTestData[mt_rand() % vTestData.size()];

		for (auto& data : map1)
		{
			if (stricmp(str.c_str(), data.first.c_str()) == 0)
			{
				// Found it!
				break;
			}
		}
	}

	end = std::chrono::high_resolution_clock::now();
	time = std::chrono::duration_cast<std::chrono::milliseconds>(end - start);
	std::cout << time.count() << "\n";

	// Case 1-2: Find arbitrary string using std::map::find
	std::cout << "Find arbitrary string using std::map::find\n";
	start = std::chrono::high_resolution_clock::now();

	for (int i = 0; i < NUM_ITERATIONS; ++i)
	{
		std::string str = vTestData[mt_rand() % vTestData.size()];
		std::transform(str.begin(), str.end(), str.begin(), ::tolower);

		auto search = map1.find(str);

		if (search != map1.end())
		{
			// Found it!
		}
		else
		{
			// Not found
		}
	}

	end = std::chrono::high_resolution_clock::now();
	time = std::chrono::duration_cast<std::chrono::milliseconds>(end - start);
	std::cout << time.count() << "\n";

	// Case 2: Store std::string in std::map without lowercase process
	input.open("samples.txt");
	std::map<std::string, int> map2;

	std::cout << "Store std::string in std::map without lowercase process\n";
	start = std::chrono::high_resolution_clock::now();

	while (!input.eof())
	{
		std::string str;
		input >> str;

		map2.insert(std::make_pair(str, 1));
	}

	end = std::chrono::high_resolution_clock::now();
	time = std::chrono::duration_cast<std::chrono::milliseconds>(end - start);
	std::cout << time.count() << "\n";

	input.close();

	// Case 2-1: Find arbitrary string using for-statement
	// NOTE: auto is much slower than auto& because of copy cost
	std::cout << "Find arbitrary string using for-statement\n";
	start = std::chrono::high_resolution_clock::now();

	for (int i = 0; i < NUM_ITERATIONS; ++i)
	{
		std::string str = vTestData[mt_rand() % vTestData.size()];

		for (auto& data : map2)
		{
			if (stricmp(str.c_str(), data.first.c_str()) == 0)
			{
				// Found it!
				break;
			}
		}
	}

	end = std::chrono::high_resolution_clock::now();
	time = std::chrono::duration_cast<std::chrono::milliseconds>(end - start);
	std::cout << time.count() << "\n";

	// Case 2-2: Find arbitrary string using std::find_if
	std::cout << "Find arbitrary string using std::find_if\n";
	start = std::chrono::high_resolution_clock::now();

	for (int i = 0; i < NUM_ITERATIONS; ++i)
	{
		std::string str = vTestData[mt_rand() % vTestData.size()];

		std::find_if(map2.begin(), map2.end(), [&str](const std::pair<std::string, int>& v)
		{
			return stricmp(v.first.c_str(), str.c_str()) == 0;
		});
	}

	end = std::chrono::high_resolution_clock::now();
	time = std::chrono::duration_cast<std::chrono::milliseconds>(end - start);
	std::cout << time.count() << "\n";

	return 0;
}