Separate pre-classification checks for client-side malware and phishing

chrome/browser/safe_browsing/browser_feature_extractor.cc

Index: chrome/browser/safe_browsing/browser_feature_extractor.cc
diff --git a/chrome/browser/safe_browsing/browser_feature_extractor.cc b/chrome/browser/safe_browsing/browser_feature_extractor.cc
index 522c3fc07d69a96dc57f855518a5d59cd75e4bf5..8d0bd73577bf5b4cf6d8b1f6e26d56e607fae5fa 100644
--- a/chrome/browser/safe_browsing/browser_feature_extractor.cc
+++ b/chrome/browser/safe_browsing/browser_feature_extractor.cc
@@ -262,7 +262,6 @@ void BrowserFeatureExtractor::ExtractMalwareFeatures(
     ClientMalwareRequest* request,
     const MalwareDoneCallback& callback) {
   DCHECK(BrowserThread::CurrentlyOn(BrowserThread::UI));
-  DCHECK_EQ(0U, request->url().find("http:"));

[mattm, 2014/03/18 02:19:06]

why is this removed?

[noé, 2014/03/20 17:01:45]

For malware we also look at HTTPS websites.  This is similar to what we do for
malicious download protection. 

[mattm, 2014/03/20 22:51:40]

Ah, I had forgotten the HTTP check done in PhishingClassifier. (Seems a little
weird that's not part of the pre-classification checks here. ah well.)

[noé, 2014/03/21 00:08:38]

You're right.  Added a pre-classification check for that but didn't change the
classifier for now.

   DCHECK(!callback.is_null());
 
   // Grab the IPs because they might go away before we're done