ETL应用：一种一次获取一个平台接口文件的方法

ETL应用场景中，若对端接口文件未能提供，任务会处于循环等待，直到对端提供为止，该方法极大的消耗了系统资源。为此想到了一种方法，一次获取一个平台的文件，实现思路如下：

1、第一次获取对端平台提供目录下给定日期的所有接口文件，并保存文件列表；

2、后续每隔n分钟重启获取任务，每次先获取文件列表，和上次列表进行对比，当发生如下情况时，会重新获取：

A、有新文件产生；

B、有文件大小变化

实现方法如下：

[ftp.properties]

ipaddress = 10.25.xxx.xxx
username = xxxxx
password = xxxxx

#\u5F53 encryption \u6709\u503C\u65F6\uFF0C\u5C06\u8FDB\u884C\u5BC6\u7801\u89E3\u6790
encryption = 

#\u5F53resolve \u4E3A False\u65F6\uFF0C\u9700\u8981\u66FF\u6362\u8FDC\u7A0B\u76EE\u5F55\u548C\u5F53\u524D\u76EE\u5F55\u7684\u53C2\u6570
resolve = 1
remoteDir = /bosscdr/tobak/jf_bass
localDir = /interface/cyg/[SDT_YYYYMMDD]

#\u4E0A\u6B21\u4FDD\u5B58\u7684\u6587\u4EF6\u83B7\u53D6\u5217\u8868
lastFileList = /interface/cyg/lastfilelist.txt

# -*-  coding:utf-8 -*-
'''
函数说明 ：获取远程文件
编写时间： 2015-5-5
编 写 人 ： chenyangang
----------------------------------------
实现方法：
   1、获取配置文件中制定的ftp服务器信息，用户名采用加密方式
   2、获取远程目录下的文件列表，若存在保存的文件列表，进行比对，提取差异文件
   3、根据差异文件进行文件获取
'''
import datetime
import ConfigParser
import os
import ftplib
import cPickle

class GetDataBaseDiff(object):

    def __init__(self, config, interfaceID = None, interfaceDate = None, delay = 0):
        self.config = config
        self.interfaceID = interfaceID

        #默认为当天日期
        if interfaceDate == None:
            self.interfaceDate = datetime.date.strftime(datetime.date.today() - \
                                 datetime.timedelta(delay),"%Y%m%d")

    def getConfig(self, interfaceDate):

        readConfig = ConfigParser.ConfigParser()
        with open(self.config,'r') as configFile:

            readConfig.readfp(configFile)
            hostaddr = readConfig.get('ftp.properties','ipaddress')
            username = readConfig.get('ftp.properties','username')          

            #是否解析参数和加密
            resolve = readConfig.get('ftp.properties','resolve')
            encryption = readConfig.get('ftp.properties','encryption')

            #目录信息
            remoteDir = readConfig.get('ftp.properties','remoteDir')
            localDir = readConfig.get('ftp.properties','localDir') 

            #存储上次获取文件列表
            lastFileList = readConfig.get('ftp.properties','lastFileList') 

            if encryption == '' :
                password = readConfig.get('ftp.properties','password')
            else:
                command = encryption + ' ' + readConfig.get('ftp.properties','password')
                password = os.popen(command)

            if resolve == '' :
                month = interfaceDate[0:6]
                remoteDir = remoteDir.replace(r"[SDT_YYYYMMDD]", interfaceDate)
                remoteDir = remoteDir.replace(r"[SDT_YYYYMM]",month)

                localDir = localDir.replace(r"[SDT_YYYYMMDD]", interfaceDate)
                localDir = localDir.replace(r"[SDT_YYYYMM]",month)

        return hostaddr, username, password, remoteDir, localDir, lastFileList

    def connect(self, hostaddr, username, password):

        try:
            connftp = ftplib.FTP(hostaddr)
        except ftplib.error_perm:
            print "The ipaddress (ipaddress) refused!" %{'ipaddress':hostaddr}

        try:
            connftp.login(username, password)
        except ftplib.error_perm:
            print "This username (username) refuse connect, please check your\
                         username or password!" %{'username':username}

        return connftp

    def getFileList(self, connftp, remoteDir):

        #获取文件详细信息，包括权限、文件大小、属主等信息，其中第5项为文件大小
        connftp.cwd(remoteDir)
        filesDetail = connftp.nlst('-l')

        #保存文件名称和大小
        fileList = {}

        for fileDetail in filesDetail:
            filelistFromDetail = fileDetail.strip().split()
            fileList[filelistFromDetail[-1]] = filelistFromDetail[4]

        return fileList

    def comparisonFileList(self, lastFileList, newFileList):

        #装载上一次文件获取信息

        if len(open(lastFileList, "rb").readlines()) > 0 :
            with open(lastFileList, "rb") as fp:
                try:
                    lastfileList = cPickle.load(fp)
                except EOFError:
                    print "Load (filename) was failed"%{'filename':lastFileList}
        else:
            lastfileList={}

        lastfileset = set(lastfileList.keys())
        newfileSet = set(newFileList.keys())

        #提取新增文件列表
        diffFileList = list(newfileSet - lastfileset)
        sameFileName = list(newfileSet & lastfileset)

        #提取前后文件大小不一致的文件列表
        for samefilename in sameFileName:
            if newFileList[samefilename] != lastfileList[samefilename]:
                diffFileList.append(samefilename)

        del lastfileList
        #保存最新文件获取列表
        fp = open(lastFileList, "wb")

        lastfileList = cPickle.dump(newFileList, fp)
        fp.close()

        return diffFileList

    def machedFileList(self, diffFileList, interfaceID, interfaceDate):
        return [flist for flist in diffFileList if interfaceID in flist \
                                              and interfaceDate in flist] 

    def download(self, connftp, localDir, getFileList):

        #进入本地目录
        if not os.path.isdir(localDir) :
            os.makedirs(localDir)

        try:
            os.chdir(localDir)
        except :
            print 'Dose\'t enter the directory , mybe you have not authority!'

        #获取最新文件
        for remotefile in getFileList:
            try:
                connftp.retrbinary("RETR %s"%remotefile, open(remotefile,"wb").write)
            except ftplib.error_perm:
                print 'ERROR: cannot read file "%s"' % remotefile

        connftp.quit()

if __name__ == '__main__' :
    interfaceDate = ''
    interfaceID = None
    getDataBaseDiff = GetDataBaseDiff('./config.properties', interfaceDate, 0)
    hostaddr, username, password, remoteDir, localDir, lastFileList = getDataBaseDiff.getConfig(interfaceDate)

    connectionFtp = getDataBaseDiff.connect(hostaddr, username, password)

    fileList = getDataBaseDiff.getFileList(connectionFtp, remoteDir)
    diffFileList = getDataBaseDiff.comparisonFileList(lastFileList, fileList)

    if interfaceID is not None and len(diffFileList) >0:
        getFileList = getDataBaseDiff.machedFileList(diffFileList, interfaceID, interfaceDate)
        getDataBaseDiff.download(connectionFtp, localDir, getFileList)
    else:
        getDataBaseDiff.download(connectionFtp, localDir, diffFileList)

如上，是学习python后，尝试编写的代码。可修改为配置文件中配置多个平台，获取多平台接口数据。