Merge branch 'master' of github.com:discourse/discourse

app/models/screened_email.rb

@@ -17,11 +17,43 @@ class ScreenedEmail < ActiveRecord::Base
   end
 
   def self.should_block?(email)
-    screened_email = ScreenedEmail.find_by(email: email)
+    screened_emails = ScreenedEmail.order(created_at: :desc).limit(100)
+
+    distances = {}
+    screened_emails.each { |se| distances[se.email] = levenshtein(se.email, email) }
+
+    max_distance = SiteSetting.levenshtein_distance_spammer_emails
+    screened_email = screened_emails.select { |se| distances[se.email] <= max_distance }
+                                    .sort   { |se| distances[se.email] }
+                                    .first
+
     screened_email.record_match! if screened_email
+
     screened_email && screened_email.action_type == actions[:block]
   end
 
+  def self.levenshtein(first, second)
+    matrix = [(0..first.length).to_a]
+    (1..second.length).each do |j|
+      matrix << [j] + [0] * (first.length)
+    end
+
+    (1..second.length).each do |i|
+      (1..first.length).each do |j|
+        if first[j-1] == second[i-1]
+          matrix[i][j] = matrix[i-1][j-1]
+        else
+          matrix[i][j] = [
+            matrix[i-1][j],
+            matrix[i][j-1],
+            matrix[i-1][j-1],
+          ].min + 1
+        end
+      end
+    end
+    return matrix.last.last
+  end
+
 end
 
 # == Schema Information

config/locales/server.en.yml

@@ -890,6 +890,8 @@ en:
     white_listed_spam_host_domains: "A pipe-delimited list of domains excluded from spam host testing, new users will be able to create an unrestricted count of posts with links to this domain"
     staff_like_weight: "Extra weighting factor given to likes when performed by staff."
 
+    levenshtein_distance_spammer_emails: "Number of characters different from a known spammer email."
+
     reply_by_email_enabled: "Enable replying to topics via email"
     reply_by_email_address: "Template for reply by email incoming email address, for example: %{reply_key}@reply.example.com or replies+%{reply_key}@example.com"
 

config/site_settings.yml

@@ -527,6 +527,10 @@ spam:
   white_listed_spam_host_domains:
     default: ''
     type: list
+  levenshtein_distance_spammer_emails:
+      default: 2
+      min: 0
+      max: 3
 
 rate_limits:
   unique_posts_mins:

spec/models/screened_email_spec.rb

@@ -3,33 +3,34 @@ require 'spec_helper'
 describe ScreenedEmail do
 
   let(:email) { 'block@spamfromhome.org' }
+  let(:similar_email) { 'bl0ck@spamfromhome.org' }
 
   describe "new record" do
     it "sets a default action_type" do
-      described_class.create(email: email).action_type.should == described_class.actions[:block]
+      ScreenedEmail.create(email: email).action_type.should == ScreenedEmail.actions[:block]
     end
 
     it "last_match_at is null" do
       # If we manually load the table with some emails, we can see whether those emails
       # have ever been blocked by looking at last_match_at.
-      described_class.create(email: email).last_match_at.should be_nil
+      ScreenedEmail.create(email: email).last_match_at.should be_nil
     end
   end
 
   describe '#block' do
     context 'email is not being blocked' do
       it 'creates a new record with default action of :block' do
-        record = described_class.block(email)
+        record = ScreenedEmail.block(email)
         record.should_not be_new_record
         record.email.should == email
-        record.action_type.should == described_class.actions[:block]
+        record.action_type.should == ScreenedEmail.actions[:block]
       end
 
       it 'lets action_type be overriden' do
-        record = described_class.block(email, action_type: described_class.actions[:do_nothing])
+        record = ScreenedEmail.block(email, action_type: ScreenedEmail.actions[:do_nothing])
         record.should_not be_new_record
         record.email.should == email
-        record.action_type.should == described_class.actions[:do_nothing]
+        record.action_type.should == ScreenedEmail.actions[:do_nothing]
       end
     end
 
@@ -37,22 +38,34 @@ describe ScreenedEmail do
       let!(:existing) { Fabricate(:screened_email, email: email) }
 
       it "doesn't create a new record" do
-        expect { described_class.block(email) }.to_not change { described_class.count }
+        expect { ScreenedEmail.block(email) }.to_not change { ScreenedEmail.count }
       end
 
       it "returns the existing record" do
-        described_class.block(email).should == existing
+        ScreenedEmail.block(email).should == existing
       end
     end
   end
 
   describe '#should_block?' do
-    subject { described_class.should_block?(email) }
+    subject { ScreenedEmail.should_block?(email) }
 
     it "returns false if a record with the email doesn't exist" do
       subject.should be_false
     end
 
+    it "returns true when there is a record with the email" do
+      ScreenedEmail.should_block?(email).should be_false
+      ScreenedEmail.create(email: email).save
+      ScreenedEmail.should_block?(email).should be_true
+    end
+
+    it "returns true when there is a record with a similar email" do
+      ScreenedEmail.should_block?(email).should be_false
+      ScreenedEmail.create(email: similar_email).save
+      ScreenedEmail.should_block?(email).should be_true
+    end
+
     shared_examples "when a ScreenedEmail record matches" do
       it "updates statistics" do
         Timecop.freeze(Time.zone.now) do
@@ -63,13 +76,13 @@ describe ScreenedEmail do
     end
 
     context "action_type is :block" do
-      let!(:screened_email) { Fabricate(:screened_email, email: email, action_type: described_class.actions[:block]) }
+      let!(:screened_email) { Fabricate(:screened_email, email: email, action_type: ScreenedEmail.actions[:block]) }
       it { should be_true }
       include_examples "when a ScreenedEmail record matches"
     end
 
     context "action_type is :do_nothing" do
-      let!(:screened_email) { Fabricate(:screened_email, email: email, action_type: described_class.actions[:do_nothing]) }
+      let!(:screened_email) { Fabricate(:screened_email, email: email, action_type: ScreenedEmail.actions[:do_nothing]) }
       it { should be_false }
       include_examples "when a ScreenedEmail record matches"
     end