ASP获取网页全部图片地址并保存为数组的正则
目前还是有BUG的,最新的测试页面在: http://www.reallydo.com/getimg.asp
正则分析页面在: http://jorkin.reallydo.com/article.asp?id=380
发现BUG请在后面留言,谢谢.
1.31修正
src=后面有空格不能正确匹配.已修正.
src=''为空时出错.已修正.
发现BUG: 图片路径有多个空格时只能保留一个.未修正.
2.18修正
图片路径有多个空格时只能保留一个的BUG.已修正.
代码如下:
<%
'功能:获取全部图片地址,保存到一个数组.
'来源:http://jorkin.reallydo.com/article.asp?id=448
'需要ReplaceAll函数:http://jorkin.reallydo.com/article.asp?id=406
Function getIMG(sString)
Dim sReallyDo, regEx, iReallyDo
Dim oMatches, cMatch
'//定义一个空数组
iReallyDo = -1
ReDim aReallyDo(iReallyDo)
If IsNull(sString) Then
getIMG = ""
Exit Function
End If
'//格式化HTML代码
'//将每个 <img 换行 方便正则替换
sReallyDo = sString
On Error Resume Next
sReallyDo = Replace(sReallyDo, vbCr, " ")
sReallyDo = Replace(sReallyDo, vbLf, " ")
sReallyDo = Replace(sReallyDo, vbTab, " ")
sReallyDo = Replace(sReallyDo, "<img ", vbCrLf & "<img ", 1, -1, 1)
sReallyDo = Replace(sReallyDo, "/>", " />", 1, -1, 1)
sReallyDo = ReplaceAll(sReallyDo, "= ", "=", True)
sReallyDo = ReplaceAll(sReallyDo, "> ", ">", True)
sReallyDo = Replace(sReallyDo, "><", ">" & vbCrLf & "<")
sReallyDo = Trim(sReallyDo)
On Error GoTo 0
Set regEx = New RegExp
regEx.IgnoreCase = True
regEx.Global = True
'//去除onclick,onload等脚本
regEx.Pattern = "\s[on].+?=([\""|\'])(.*?)\1"
sReallyDo = regEx.Replace(sReallyDo, "")
'//将SRC不带引号的图片地址加上引号
regEx.Pattern = "<img.*?\ssrc=([^\""\'\s][^\""\'\s>]*).*?>"
sReallyDo = regEx.Replace(sReallyDo, "<img src=""$1"" />")
'//正则匹配图片SRC地址
regEx.Pattern = "<img.*?\ssrc=([\""\'])([^\""\']+?)\1.*?>"
Set oMatches = regEx.Execute(sReallyDo)
'//将图片地址存入数组
For Each cMatch in oMatches
iReallyDo = iReallyDo + 1
ReDim Preserve aReallyDo(iReallyDo)
aReallyDo(iReallyDo) = regEx.Replace(cMatch.Value, "$2")
Next
getIMG = aReallyDo
End Function
%>