From 82a499bbb4cd2e405b3510eda741d27baf4c2836 Mon Sep 17 00:00:00 2001 From: edoardottt Date: Fri, 19 Nov 2021 17:26:02 +0100 Subject: [PATCH] Fix Ignore option --- crawler/colly.go | 48 ++++++++++++++++++++++++++++++------------------ 1 file changed, 30 insertions(+), 18 deletions(-) diff --git a/crawler/colly.go b/crawler/colly.go index b83cf88..47741e8 100644 --- a/crawler/colly.go +++ b/crawler/colly.go @@ -101,9 +101,11 @@ func Crawler(target string, txt string, html string, delayTime int, concurrency // Only those links are visited which are in AllowedDomains if (!intensive && utils.SameDomain(protocolTemp+"://"+target, absoluteUrl)) || (intensive && intensiveOk(targetTemp, absoluteUrl)) { - if ignoreBool && !IgnoreMatch(link, ignoreSlice) { - FinalResults = append(FinalResults, absoluteUrl) - c.Visit(absoluteUrl) + if ignoreBool { + if !IgnoreMatch(link, ignoreSlice) { + FinalResults = append(FinalResults, absoluteUrl) + c.Visit(absoluteUrl) + } } else { FinalResults = append(FinalResults, absoluteUrl) c.Visit(absoluteUrl) @@ -121,9 +123,11 @@ func Crawler(target string, txt string, html string, delayTime int, concurrency // Only those links are visited which are in AllowedDomains if (!intensive && utils.SameDomain(protocolTemp+"://"+target, absoluteUrl)) || (intensive && intensiveOk(targetTemp, absoluteUrl)) { - if ignoreBool && !IgnoreMatch(link, ignoreSlice) { - FinalResults = append(FinalResults, absoluteUrl) - c.Visit(absoluteUrl) + if ignoreBool { + if !IgnoreMatch(link, ignoreSlice) { + FinalResults = append(FinalResults, absoluteUrl) + c.Visit(absoluteUrl) + } } else { FinalResults = append(FinalResults, absoluteUrl) c.Visit(absoluteUrl) @@ -141,9 +145,11 @@ func Crawler(target string, txt string, html string, delayTime int, concurrency // Only those links are visited which are in AllowedDomains if (!intensive && utils.SameDomain(protocolTemp+"://"+target, absoluteUrl)) || (intensive && intensiveOk(targetTemp, absoluteUrl)) { - if ignoreBool && !IgnoreMatch(link, ignoreSlice) { - FinalResults = append(FinalResults, absoluteUrl) - c.Visit(absoluteUrl) + if ignoreBool { + if !IgnoreMatch(link, ignoreSlice) { + FinalResults = append(FinalResults, absoluteUrl) + c.Visit(absoluteUrl) + } } else { FinalResults = append(FinalResults, absoluteUrl) c.Visit(absoluteUrl) @@ -161,9 +167,11 @@ func Crawler(target string, txt string, html string, delayTime int, concurrency // Only those links are visited which are in AllowedDomains if (!intensive && utils.SameDomain(protocolTemp+"://"+target, absoluteUrl)) || (intensive && intensiveOk(targetTemp, absoluteUrl)) { - if ignoreBool && !IgnoreMatch(link, ignoreSlice) { - FinalResults = append(FinalResults, absoluteUrl) - c.Visit(absoluteUrl) + if ignoreBool { + if !IgnoreMatch(link, ignoreSlice) { + FinalResults = append(FinalResults, absoluteUrl) + c.Visit(absoluteUrl) + } } else { FinalResults = append(FinalResults, absoluteUrl) c.Visit(absoluteUrl) @@ -181,9 +189,11 @@ func Crawler(target string, txt string, html string, delayTime int, concurrency // Only those links are visited which are in AllowedDomains if (!intensive && utils.SameDomain(protocolTemp+"://"+target, absoluteUrl)) || (intensive && intensiveOk(targetTemp, absoluteUrl)) { - if ignoreBool && !IgnoreMatch(link, ignoreSlice) { - FinalResults = append(FinalResults, absoluteUrl) - c.Visit(absoluteUrl) + if ignoreBool { + if !IgnoreMatch(link, ignoreSlice) { + FinalResults = append(FinalResults, absoluteUrl) + c.Visit(absoluteUrl) + } } else { FinalResults = append(FinalResults, absoluteUrl) c.Visit(absoluteUrl) @@ -201,9 +211,11 @@ func Crawler(target string, txt string, html string, delayTime int, concurrency // Only those links are visited which are in AllowedDomains if (!intensive && utils.SameDomain(protocolTemp+"://"+target, absoluteUrl)) || (intensive && intensiveOk(targetTemp, absoluteUrl)) { - if ignoreBool && !IgnoreMatch(link, ignoreSlice) { - FinalResults = append(FinalResults, absoluteUrl) - c.Visit(absoluteUrl) + if ignoreBool { + if !IgnoreMatch(link, ignoreSlice) { + FinalResults = append(FinalResults, absoluteUrl) + c.Visit(absoluteUrl) + } } else { FinalResults = append(FinalResults, absoluteUrl) c.Visit(absoluteUrl)