11.2 MapReduce 编程案例

      MapReduce 对大数据进行分割，分为一定大小的数据；然后将分割的数据交给多个 Mapper 函数处理。处理后产生多个规模较小的数据再提交给 Reducer 函数处理，得到更小规模的数据或最终结果。
（1）大文件切分。

import os
import os.path
import time

def FileSplit(sourceFile, targetFolder):
    if not os.path.isfile(sourceFile):
        print(sourceFile, ' does not exist.')
        return
    if not os.path.isdir(targetFolder): # 切分为多个小文件，放到指定文件夹中
        os.mkdir(targetFolder)
    tempData = []
    number = 10                       # 每个小文件的记录条数
    fileNum = 1                         # 小文件序号
    with open(sourceFile, 'r') as srcFile:
        dataLine = srcFile.readline()
        print('data:', dataLine)
        while dataLine:
            for i in range(number):
                tempData.append(dataLine)
                dataLine = srcFile.readline()
                if not dataLine:
                    break
            desFile = os.path.join(targetFolder, sourceFile[0:-4] + str(fileNum) + '.txt')
            with open(desFile, 'a+') as f:
                f.writelines(tempData)
            tempData = []
            fileNum = fileNum + 1
            print(fileNum)

if __name__ == '__main__':
    sourceFile = 'test.txt'
    targetFolder = 'test'
    FileSplit(sourceFile, targetFolder)

（2）Mapper 代码。

import os
import re
import threading
import time

def Map(sourceFile):
    if not os.path.exists(sourceFile):
        print(sourceFile, ' does not exist.')
        return
    pattern = re.compile(r'[0-9]{1,2}/[0-9]{1,2}/[0-9]{4}') # 用正则表达式匹配日期
    result = {}
    with open(sourceFile, 'r') as srcFile:
        for dataLine in srcFile:
            print('dataline:', dataLine)
            r = pattern.findall(dataLine)
            print(r)
            if r:
                t = result.get(r[0], 0)
                t += 1
                result[r[0]] = t
    desFile = sourceFile[0:-4] + '_map.txt'     # 结果文件
    with open(desFile, 'a+') as fp:
        for k, v in result.items():
            fp.write(k + ':' + str(v) +'\n')

if __name__ == '__main__':
    desFolder = 'test'
    files = os.listdir(desFolder)

    def Main(i):
        Map(desFolder + '\\' + files[i])

    fileNumber = len(files)
    for i in range(fileNumber):
        t = threading.Thread(target = Main, args = (i,))
        t.start()

（3）Reducer 代码。

import os

def Reduce(sourceFolder, targetFile):
    if not os.path.isdir(sourceFolder):
        print(sourceFolder, ' does not exist.')
        return
    result = {}
    # 用列表推导式获取文件夹的 Mapper 结果文件
    allFiles = [sourceFolder + '\\' + f for f in os.listdir(sourceFolder) if f.endswith('_map.txt')]
    for f in allFiles:
        with open(f, 'r') as fp:
            for line in fp:
                line = line.strip()
                if not line:
                    continue
                position = line.index(':')
                key = line[0:position]
                value = int(line[position + 1:])
                result[key] = result.get(key, 0) + value
        with open(targetFile, 'w') as fp:
            for k, v in result.items():
                fp.write(k + ':' + str(v) + '\n')

if __name__ == '__main__':
    Reduce('test', 'test\\result.txt')