DataFlex

A Simple Web Crawler

See more Spider Examples

This demonstrates a very simple web crawler using the Chilkat Spider component. DataFlex

Use ChilkatAx-win32.pkg

Procedure Test
    Boolean iSuccess
    Handle hoSpider
    Handle hoSeenDomains
    Handle hoSeedUrls
    String sUrl
    String sDomain
    Integer i
    String sDomain
    String sBaseDomain
    String sTemp1
    Integer iTemp1
    Boolean bTemp1

    Move False To iSuccess

    Get Create (RefClass(cComChilkatSpider)) To hoSpider
    If (Not(IsComObjectCreated(hoSpider))) Begin
        Send CreateComObject of hoSpider
    End

    Get Create (RefClass(cComCkStringArray)) To hoSeenDomains
    If (Not(IsComObjectCreated(hoSeenDomains))) Begin
        Send CreateComObject of hoSeenDomains
    End
    Get Create (RefClass(cComCkStringArray)) To hoSeedUrls
    If (Not(IsComObjectCreated(hoSeedUrls))) Begin
        Send CreateComObject of hoSeedUrls
    End

    Set ComUnique Of hoSeenDomains To True
    Set ComUnique Of hoSeedUrls To True

    // You will need to change the start URL to something else...
    Get ComAppend Of hoSeedUrls "http://something.whateverYouWant.com/" To iSuccess

    // Set outbound URL exclude patterns
    // URLs matching any of these patterns will not be added to the 
    // collection of outbound links.
    Send ComAddAvoidOutboundLinkPattern To hoSpider "*?id=*"
    Send ComAddAvoidOutboundLinkPattern To hoSpider "*.mypages.*"
    Send ComAddAvoidOutboundLinkPattern To hoSpider "*.personal.*"
    Send ComAddAvoidOutboundLinkPattern To hoSpider "*.comcast.*"
    Send ComAddAvoidOutboundLinkPattern To hoSpider "*.aol.*"
    Send ComAddAvoidOutboundLinkPattern To hoSpider "*~*"

    // Use a cache so we don't have to re-fetch URLs previously fetched.
    Set ComCacheDir Of hoSpider To "c:/spiderCache/"
    Set ComFetchFromCache Of hoSpider To True
    Set ComUpdateCache Of hoSpider To True

    While ((ComCount(hoSeedUrls)) > 0)

        Get ComPop Of hoSeedUrls To sUrl
        Send ComInitialize To hoSpider sUrl

        // Spider 5 URLs of this domain.
        // but first, save the base domain in seenDomains
        Get ComGetUrlDomain Of hoSpider sUrl To sDomain
        Get ComGetBaseDomain Of hoSpider sDomain To sTemp1
        Get ComAppend Of hoSeenDomains sTemp1 To iSuccess

        For i From 0 To 4
            Get ComCrawlNext Of hoSpider To iSuccess
            If (iSuccess = True) Begin

                // Display the URL we just crawled.
                Get ComLastUrl Of hoSpider To sTemp1
                Showln sTemp1

                // If the last URL was retrieved from cache,
                // we won't wait.  Otherwise we'll wait 1 second
                // before fetching the next URL.
                Get ComLastFromCache Of hoSpider To bTemp1
                If (bTemp1 <> True) Begin
                    Send ComSleepMs To hoSpider 1000
                End

            End
            Else Begin
                // cause the loop to exit..
                Move 999 To i
            End

        Loop

        // Add the outbound links to seedUrls, except
        // for the domains we've already seen.
        Get ComNumOutboundLinks Of hoSpider To iTemp1
        For i From 0 To (iTemp1 - 1)

            Get ComGetOutboundLink Of hoSpider i To sUrl
            Get ComGetUrlDomain Of hoSpider sUrl To sDomain
            Get ComGetBaseDomain Of hoSpider sDomain To sBaseDomain
            Get ComContains Of hoSeenDomains sBaseDomain To bTemp1
            If (bTemp1 = False) Begin
                // Don't let our list of seedUrls grow too large.
                Get ComCount Of hoSeedUrls To iTemp1
                If (iTemp1 < 1000) Begin
                    Get ComAppend Of hoSeedUrls sUrl To iSuccess
                End

            End

        Loop

    Loop



End_Procedure