Fix a possible security issue with spoofing

Thanks to MORI Shingo of DeNA Co., Ltd. for reporting this.

There is an issue where if an HTML file is uploaded with a .html
extension, but the content type is listed as being `image/jpeg`, this
will bypass a validation checking for images. But it will also pass the
spoof check, because a file named .html and containing actual HTML
passes the spoof check.

This change makes it so that we also check the supplied content type. So
even if the file contains HTML and ends with .html, it doesn't match the
content type of `image/jpeg` and so it fails.

NEWS

+New in 4.2.2:
+
+* Security fix: Fix a potential security issue with spoofing
+
 New in 4.2.1:
 
 * Improvement: Added `validate_media_type` options to allow/bypass spoof check

lib/paperclip/locales/en.yml

@@ -2,7 +2,7 @@ en:
   errors:
     messages:
       in_between: "must be in between %{min} and %{max}"
-      spoofed_media_type: "has an extension that does not match its contents"
+      spoofed_media_type: "has contents that are not what they are reported to be"
 
   number:
     human:

lib/paperclip/media_type_spoof_detector.rb

 module Paperclip
   class MediaTypeSpoofDetector
-    def self.using(file, name)
-      new(file, name)
+    def self.using(file, name, content_type)
+      new(file, name, content_type)
     end
 
-    def initialize(file, name)
+    def initialize(file, name, content_type)
       @file = file
       @name = name
+      @content_type = content_type || ""
     end
 
     def spoofed?
       if has_name? && has_extension? && media_type_mismatch? && mapping_override_mismatch?
-        Paperclip.log("Content Type Spoof: Filename #{File.basename(@name)} (#{supplied_file_content_types}), content type discovered from file command: #{calculated_content_type}. See documentation to allow this combination.")
+        Paperclip.log("Content Type Spoof: Filename #{File.basename(@name)} (#{supplied_content_type} from Headers, #{content_types_from_name} from Extension), content type discovered from file command: #{calculated_content_type}. See documentation to allow this combination.")
         true
+      else
+        false
       end
     end
 
@@ -27,35 +30,44 @@ module Paperclip
     end
 
     def media_type_mismatch?
-      ! supplied_file_media_types.include?(calculated_media_type)
+      supplied_type_mismatch? || calculated_type_mismatch?
+    end
+
+    def supplied_type_mismatch?
+      supplied_media_type.present? && !media_types_from_name.include?(supplied_media_type)
+    end
+
+    def calculated_type_mismatch?
+      !media_types_from_name.include?(calculated_media_type)
     end
 
     def mapping_override_mismatch?
       mapped_content_type != calculated_content_type
     end
 
-    def supplied_file_media_types
-      @supplied_file_media_types ||= MIME::Types.type_for(@name).collect(&:media_type)
+
+    def supplied_content_type
+      @content_type
     end
 
-    def calculated_media_type
-      @calculated_media_type ||= calculated_content_type.split("/").first
+    def supplied_media_type
+      @content_type.split("/").first
     end
 
-    def supplied_file_content_types
-      @supplied_file_content_types ||= MIME::Types.type_for(@name).collect(&:content_type)
+    def content_types_from_name
+      @content_types_from_name ||= MIME::Types.type_for(@name)
     end
 
-    def calculated_content_type
-      @calculated_content_type ||= type_from_file_command.chomp
+    def media_types_from_name
+      @media_types_from_name ||= content_types_from_name.collect(&:media_type)
     end
 
-    def mapped_content_type
-      Paperclip.options[:content_type_mappings][filename_extension]
+    def calculated_content_type
+      @calculated_content_type ||= type_from_file_command.chomp
     end
 
-    def filename_extension
-      File.extname(@name.to_s.downcase).sub(/^\./, '').to_sym
+    def calculated_media_type
+      @calculated_media_type ||= calculated_content_type.split("/").first
     end
 
     def type_from_file_command
@@ -65,5 +77,13 @@ module Paperclip
         ""
       end
     end
+
+    def mapped_content_type
+      Paperclip.options[:content_type_mappings][filename_extension]
+    end
+
+    def filename_extension
+      File.extname(@name.to_s.downcase).sub(/^\./, '').to_sym
+    end
   end
 end

lib/paperclip/validators/media_type_spoof_detection_validator.rb

@@ -5,7 +5,7 @@ module Paperclip
     class MediaTypeSpoofDetectionValidator < ActiveModel::EachValidator
       def validate_each(record, attribute, value)
         adapter = Paperclip.io_adapters.for(value)
-        if Paperclip::MediaTypeSpoofDetector.using(adapter, value.original_filename).spoofed?
+        if Paperclip::MediaTypeSpoofDetector.using(adapter, value.original_filename, value.content_type).spoofed?
           record.errors.add(attribute, :spoofed_media_type)
         end
       end

spec/paperclip/media_type_spoof_detector_spec.rb

@@ -43,4 +43,14 @@ describe Paperclip::MediaTypeSpoofDetector do
       Paperclip.options[:content_type_mappings] = {}
     end
   end
+
+  it "rejects a file if named .html and is as HTML, but we're told JPG" do
+    file = File.open(fixture_file("empty.html"))
+    assert Paperclip::MediaTypeSpoofDetector.using(file, "empty.html", "image/jpg").spoofed?
+  end
+
+  it "does not reject is content_type is empty but otherwise checks out" do
+    file = File.open(fixture_file("empty.html"))
+    assert ! Paperclip::MediaTypeSpoofDetector.using(file, "empty.html", "").spoofed?
+  end
 end

spec/paperclip/validators/media_type_spoof_detection_validator_spec.rb

@@ -30,7 +30,7 @@ describe Paperclip::Validators::MediaTypeSpoofDetectionValidator do
     Paperclip::MediaTypeSpoofDetector.stubs(:using).returns(detector)
     @validator.validate(@dummy)
 
-    assert_equal "has an extension that does not match its contents", @dummy.errors[:avatar].first
+    assert_equal I18n.t("errors.messages.spoofed_media_type"), @dummy.errors[:avatar].first
   end
 
   it "runs when attachment is dirty" do