利用bitmap对大规模数据排序_clickhouse bitmap 数据排序-优快云博客

本文链接：https://blog.youkuaiyun.com/HerosOfEarth/article/details/104832301

这篇博客介绍了如何使用Bitmap进行大规模数据的排序。通过详细解析`bitmap.h`和`bitmap.cpp`两个关键文件的实现，阐述了Bitmap数据结构在处理海量数据时的优势和具体操作步骤，展示了其在排序算法中的高效性能。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

bitmap.h

#ifndef _BITMAP_H_
#define _BITMAP_H_
#include <cstdio>
#include <cstdlib>
#include <iostream>
#include <limits.h>
#include <cstring>
using namespace std;

class BitMap {
public:
	BitMap(uint64_t size);
	~BitMap();
	uint64_t getSize() { return size; }
	bool getBit(uint64_t pos);
	void setBit(uint64_t pos);
	void resetBit(uint64_t pos);

private:
	uint64_t size;
	char* pointer;
};
#endif

bitmap.cpp

#include "bitmap.h"
BitMap::BitMap(uint64_t size_) {
	if (size < UINT64_MAX) {
		size = size_;
		uint64_t byteNum = size / 8 + 1;
		pointer = (char*)malloc(sizeof(char) * byteNum);
		memset(pointer, 0, byteNum);
	} else {
		cout << "Error! Over the max size.\n";
		exit(-1);
	}
}

BitMap::~BitMap(void) {
	if (pointer) {
		free(pointer);
	}
}

bool BitMap::getBit(uint64_t pos) {
	uint64_t index = pos / 8;
	int offset = pos % 8;
	return (*(pointer + index) & (1 << offset));
}

void BitMap::setBit(uint64_t pos) {
	uint64_t numberOfchar = pos / 8;
	int restOfchar = pos % 8;
	*(pointer + numberOfchar) |= 1 << restOfchar;
}

void BitMap::resetBit(uint64_t pos) {
	uint64_t index = pos / 8;
	int offset = pos % 8;
	char tmp = *(pointer + index);
	*(pointer + index) &= ~(1 << offset);
}

排序

void sortByBitMap(vector<uint64_t>& nums) {
	uint64_t size = nums.size();
	BitMap bm = BitMap(size);
	for (uint64_t x : nums) {
		bm.setBit(x);
	}
	uint64_t index = 0;
	for (uint64_t i = 1; i <= bm.getSize(); i++) {
		if (bm.getBit(i))
			nums[index++] = i;
	}
}