【图】实例讲解ASP实现抓取网上房产信息

<%@LANGUAGE="VBSCRIPT" CODEPAGE="936"%
<!-- #include file="conn.asp" --

<!-- #include file="inc/function.asp" --
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd"
<html
<head
<titleUntitled Document</title
<meta http-equiv="Content-Type" content="text/html; charset=gb2312"
<meta http-equiv="refresh" content="300;URL=steal_house.asp"
</head

<body
<%
on error resume next
'
Server.ScriptTimeout = 999999
'========================================================
'字符编码函数
'====================================================
Function BytesToBstr(body,code)
        dim objstream
        set objstream = Server.CreateObject("adodb.stream")
        objstream.Type = 1
        objstream.Mode =3
        objstream.Open
        objstream.Write body
        objstream.Position = 0
        objstream.Type = 2
        objstream.Charset =code
        BytesToBstr = objstream.ReadText
        objstream.Close
        set objstream = nothing
End Function

'取行字符串在另一字符串中的出现位置
Function Newstring(wstr,strng)
Newstring=Instr(lcase(wstr),lcase(strng))
if Newstring<=0 then Newstring=Len(wstr)
End Function
'替换字符串函数
function ReplaceStr(ori,str1,str2)
ReplaceStr=replace(ori,str1,str2)
end function
'====================================================
function ReadXml(url,code,start,ends)
set oSend=createobject("Microsoft.XMLHTTP")
SourceCode = oSend.open ("GET",url,false)
oSend.send()
ReadXml=BytesToBstr(oSend.responseBody,code )
start=Instr(ReadXml,start)
ReadXml=mid(ReadXml,start)
ends=Instr(ReadXml,ends)
ReadXml=left(ReadXml,ends-1)
end function

function SubStr(body,start,ends)
start=Instr(body,start)
SubStr=mid(body,start+len(start)+1)
ends=Instr(SubStr,ends)
SubStr=left(SubStr,ends-1)
end function

dim getcont,NewsContent
dim url,title
url="http://www.***.com"'新闻网址knowsky.com
getcont=ReadXml(url,"gb2312","<table class=k2 border=""0""","</table")
getcont=RegexHtml(getcont)
dim KeyId,NewsClass,City,Position,HouseType,Level,Area,Price,Demostra

dim ContactMan,Contact
for i=2 to ubound(getcont)
response.Write(getcont(i)&"__<br")

tempLink=mid(getcont(i),instr(getcont(i),"href=""")+6,instr(getcont(i),""" onClick")-10)
tempLink=replace(tempLink,"../","")

response.Write(i&":"&tempLink&"<br")
NewsContent=ReadXml(tempLink,"gb2312","<td valign=""bottom"" width=""400""","<hr width=""760"" noshade size=""1"" color=""#808080"" ")
NewsContent=RemoveHtml(NewsContent)
NewsContent=replace(NewsContent,VbCrLf,"")
NewsContent=replace(NewsContent,vbNewLine,"")
NewsContent=replace(NewsContent," ","")
NewsContent=replace(NewsContent," ","")
NewsContent=replace(NewsContent," ","")
NewsContent=replace(NewsContent,"n","")
NewsContent=replace(NewsContent,chr(10),"")
NewsContent=replace(NewsContent,chr(13),"")
'===============get Content=======================
response.Write(NewsContent)
KeyId=SubStr(NewsContent,"列号：","信息类别：")
NewsClass=SubStr(NewsContent,"类别：","所在城市：")
City=SubStr(NewsContent,"