python之扫描本地信息到Solr服务

##scan_to_solr.py


#-*-coding:utf-8-*-
import urllib
import os,sys,re
import urllib2,httplib
import threading
import time
urlMap = {}
gEnd = False
gExit = False
gOK = 0
gErr = 0
reload(sys)
#sys.setdefaultencoding('utf-8')
sys.setdefaultencoding('gbk')
gCount = 0
gErrorList = []
gErrNumber = 0
def getUrl(url):
    try:
        return urllib2.urlopen(url,timeout=240).read()
    except:
        print "XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXload error~"
        return None

def addIndex(key,value):
    url = u'http://192.168.4.60/baidu/src/addServer.php?ch=%s&jp=%s' % (key.decode("gbk"),value)
    ret = getUrl(url.encode("utf-8"))
    if ret != None:
    	return True
    return False

def getList(dirname,pFunc):
    try:
        ls=os.listdir(dirname)
    except:
        print dirname,'is access deny'
    else:
        for file in ls:
            temp = os.path.join(dirname,file)
            if(os.path.isdir(temp)):
                getList(temp,pFunc)
            else:
                pFunc(dirname,file)

"""
    callBack Function 
"""
def doAddFile(dir,file):
    global gCount
    global urlMap,gOK,gErr,gErrNumber,gErrorList
    if file == "entries" or file == "toIndex.py" or file == "index.txt":
    	return
    if re.search("(.ppt|.pptx|.xls|.txt|.doc|.vsd|.pdf|.rar|.xlsx|.zip|.docx|.tar|.gz|.iso|.rpm|.sql)$",file):
		gCount = gCount + 1
		dir = dir.replace("\\","/")
		svnUrl = 'svn://192.168.0.248:9997/DOC/' + u'软件部/'
		anchor = "%s%s" % (svnUrl,dir[11:].decode("gbk"))
		print file
		file = file.replace(" ","")
		print "###############"
		print file
		print "###############(%d)" % (gCount)
		ret = addIndex(file,anchor)
		if ret == True:
			gOK = gOK + 1
			print "OOO KKK(%d)\n" % (gOK)
			time.sleep(1)
		else:
			gErr = gErr + 1
			print "NNN GGG(%d)\n" % (gErr)
			addIndex(key,value)

"""
    Main Inport
"""
print "\n============================== Begin ScanDir ========================================"
getList(os.getcwd(),doAddFile)
print "ScanFile(%d)" % (gCount)
print "\n========================================================================================"
print "OK(%d) ,Error(%d)" % (gOK, gErr)


@@用法说明

scan_to_solr.py

扫描本地文件提交到Solr的模糊检索服务中

  • 0
    点赞
  • 0
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值