miteru 2.1.1 → 2.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 509957b241020f76fcd9c85a6acbb9e80c0f8b3224fe5c1b4af41dcd2faa6ca5
-  data.tar.gz: 4e507a9974264a9b98d1ded16c09496a06ed4ca83409f75fc49719c37f934f77
+  metadata.gz: 3767b21d7f569a43215e2fdc0f83f9fc4f29fe2a4bb11fc3548a5d1657812774
+  data.tar.gz: 7877fb357548c619210ec63f3bd26c61c3fda68dc6d21910aa22fc0d7144ebf1
 SHA512:
-  metadata.gz: 3dc8e2b4569555f5551a1b7fd2407162949cc7a459a5a7f83c5a787206b5cf39b34f16235466b6bcd74f34bac1878e1e4df6797daebe407eda2c03a9e7379e3a
-  data.tar.gz: 69d83a4c964ebbcd87b416675ed03bc18e270c0d604a19ef98b8988a0f344388eb49d0fb290edaa2f8637fa5f100608a3e9a76a1571503bda529611aea332bc0
+  metadata.gz: '02480c9a5604e5c7959ae4adb9e3a657d4b59f5c668be51b7d1fe120bd46e5548ecc9a26404d88d001729bf19f4f3d1b1545be80820c0717c0e027b7995030cd'
+  data.tar.gz: f2c1adc686393f3979ccf77e87c1bc5dba9cad9023903620ab61bbf6e6a2a4406dbbdd3ffb9ca73f5297c2135436dfd995f9572085fd9fde02fb5b72eac6032b

data/lib/miteru/cache.rb

@@ -0,0 +1,45 @@
+require "redis"
+
+module Miteru
+  class Cache < Service
+    # @return [String]
+    attr_reader :url
+
+    #
+    # @param [String] url
+    #
+    def initialize(url)
+      super()
+      @url = url
+    end
+
+    #
+    # @param [String] key
+    # @param [String] value
+    # @param [Integer. nil] ex
+    #
+    def set(key, value, ex:)
+      value = redis.set("#{cache_prefix}:#{key}", value, ex:)
+      logger.info("Cache:#{key} is set.") if verbose?
+      value
+    end
+
+    #
+    # @param [String] key
+    #
+    def cached?(key)
+      value = redis.exists?("#{cache_prefix}:#{key}")
+      logger.info("Cache:#{key} found.") if verbose?
+      value
+    end
+
+    private
+
+    #
+    # @return [Redis]
+    #
+    def redis
+      @redis ||= Redis.new(url:)
+    end
+  end
+end

data/lib/miteru/concerns/url_truncatable.rb

@@ -0,0 +1,24 @@
+# frozen_string_literal: true
+
+module Miteru
+  module Concerns
+    module UrlTruncatable
+      extend ActiveSupport::Concern
+
+      def decoded_url
+        @decoded_url ||= URI.decode_www_form_component(url)
+      end
+
+      #
+      # @return [String]
+      #
+      def truncated_url
+        @truncated_url ||= decoded_url.truncate(64)
+      end
+
+      def defanged_truncated_url
+        @defanged_truncated_url ||= truncated_url.to_s.gsub(".", "[.]")
+      end
+    end
+  end
+end

data/lib/miteru/config.rb

@@ -22,6 +22,9 @@ module Miteru
       sentry_dsn: nil,
       sentry_trace_sample_rate: 0.25,
       sidekiq_redis_url: nil,
+      cache_redis_url: nil,
+      cache_ex: nil,
+      cache_prefix: "miteru:cache",
       slack_channel: "#general",
       slack_webhook_url: nil,
       threads: Parallel.processor_count,
@@ -39,6 +42,15 @@ module Miteru
     #   @return [Float]
 
     # @!attribute [r] sidekiq_redis_url
+    #   @return [String, nil]
+
+    # @!attribute [r] cache_redis_url
+    #   @return [String, nil]
+
+    # @!attribute [r] cache_ex
+    #   @return [Integer, nil]
+
+    # @!attribute [r] cache_prefix
     #   @return [String]
 
     # @!attribute [r] http_timeout

data/lib/miteru/crawler.rb

@@ -11,49 +11,46 @@ module Miteru
       Try[OpenSSL::SSL::SSLError, ::HTTP::Error, Addressable::URI::InvalidURIError] do
         info = "Website:#{website.info}."
         info = info.colorize(:red) if website.kits?
-
-        Miteru.logger.info(info)
-        return unless website.kits?
-
-        notify website
-
-        return unless auto_download?
+        logger.info(info)
 
         website.kits.each do |kit|
           downloader = Downloader.new(kit)
           result = downloader.result
-          if result.success?
-            Miteru.logger.info("Kit:#{kit.truncated_url} downloaded as #{result.value!}.")
-          else
-            Miteru.logger.warn("Kit:#{kit.truncated_url} failed to download - #{result.failure}.")
+          unless result.success?
+            logger.warn("Kit:#{kit.truncated_url} failed to download - #{result.failure}.")
+            next
           end
+          destination = result.value!
+          logger.info("Kit:#{kit.truncated_url} downloaded as #{destination}.")
+          # Remove downloaded file if auto_download is not allowed
+          FileUtils.rm(destination, force: true) unless auto_download?
+          # Notify the kit
+          notify(kit)
         end
+
+        # Cache the website
+        cache.set(website.url, website.source, ex: cache_ex) if cache?
       end.recover { nil }.value!
     end
 
     private
 
-    def auto_download?
-      Miteru.config.auto_download
-    end
-
     #
-    # @param [Miteru::Website] website
+    # @param [Miteru::Kit] kit
     #
-    def notify(website)
+    def notify(kit)
       notifiers.each do |notifier|
-        result = notifier.result(website)
+        result = notifier.result(kit)
         if result.success?
-          Miteru.logger.info("Notifier:#{notifier.name} succeeded.")
+          logger.info("Notifier:#{notifier.name} succeeded.")
         else
-          Miteru.logger.warn("Notifier:#{notifier.name} failed - #{result.failure}.")
+          logger.warn("Notifier:#{notifier.name} failed - #{result.failure}.")
         end
       end
     end
 
-    #
-    # @return [Array<Miteru::Notifiers::Base>]
-    #
+    private
+
     def notifiers
       @notifiers ||= Miteru.notifiers.map(&:new)
     end

data/lib/miteru/downloader.rb

@@ -43,7 +43,7 @@ module Miteru
 
       # Record a kit in DB
       Record.create_by_kit_and_hash(kit, sha256: sha256(destination))
-      Miteru.logger.info "Download #{kit.url} as #{destination}"
+      logger.info "Download #{kit.url} as #{destination}"
 
       destination
     end

data/lib/miteru/kit.rb

@@ -2,6 +2,8 @@
 
 module Miteru
   class Kit < Service
+    include Concerns::UrlTruncatable
+
     # @return [String]
     attr_reader :url
 
@@ -85,17 +87,6 @@ module Miteru
       @hostname ||= URI(url).hostname
     end
 
-    def decoded_url
-      @decoded_url ||= URI.decode_www_form_component(url)
-    end
-
-    #
-    # @return [String]
-    #
-    def truncated_url
-      url.truncate(64)
-    end
-
     private
 
     def filename_to_download

data/lib/miteru/notifiers/base.rb

@@ -3,7 +3,7 @@
 module Miteru
   module Notifiers
     class Base < Service
-      def call(website)
+      def call(kit)
         raise NotImplementedError
       end
 

data/lib/miteru/notifiers/slack.rb

@@ -77,13 +77,13 @@ module Miteru
       #
       # Notifiy to Slack
       #
-      # @param [Miteru::Website] website
+      # @param [Miteru::Kit] kit
       #
-      def call(website)
+      def call(kit)
         return unless callable?
 
-        attachment = SlackAttachment.new(website.url)
-        notifier.post(text: website.info, attachments: attachment.to_a) if website.kits?
+        attachment = SlackAttachment.new(kit.url)
+        notifier.post(text: kit.defanged_truncated_url, attachments: attachment.to_a)
       end
 
       def callable?

data/lib/miteru/notifiers/urlscan.rb

@@ -4,12 +4,12 @@ module Miteru
   module Notifiers
     class UrlScan < Base
       #
-      # @param [Miteru::Website] website
+      # @param [Miteru::Kit] kit
       #
-      def call(website)
+      def call(kit)
         return unless callable?
 
-        website.kits.each { |kit| submit(kit.url, source: website.source) }
+        submit(kit.decoded_url, source: kit.source)
       end
 
       def callable?

data/lib/miteru/orchestrator.rb

@@ -3,30 +3,37 @@
 module Miteru
   class Orchestrator < Service
     def call
-      Miteru.logger.info("#{websites.length} websites loaded in total.") if verbose?
+      logger.info("#{non_cached_websites.length} websites loaded in total.") if verbose?
 
-      if Miteru.sidekiq?
-        websites.each do |website|
-          Jobs::CrawleJob.perform_async(website.url, website.source)
-          Miteru.logger.info("Website:#{website.truncated_url} crawler job queued.") if verbose?
-        end
+      if sidekiq?
+        sidekiq_call
       else
-        Miteru.logger.info("Use #{threads} thread(s).") if verbose?
-        Parallel.each(websites, in_threads: threads) do |website|
-          Miteru.logger.info("Website:#{website.truncated_url} crawling started.") if verbose?
+        parallel_call
+      end
+    end
 
-          result = Crawler.result(website)
-          if result.success?
-            Miteru.logger.info("Crawler:#{website.truncated_url} succeeded.")
-          else
-            Miteru.logger.info("Crawler:#{website.truncated_url} failed - #{result.failure}.")
-          end
+    def sidekiq_call
+      non_cached_websites.each do |website|
+        Jobs::CrawleJob.perform_async(website.url, website.source)
+        logger.info("Website:#{website.truncated_url} crawler job queued.") if verbose?
+      end
+    end
+
+    def parallel_call
+      logger.info("Use #{threads} thread(s).") if verbose?
+      Parallel.each(non_cached_websites, in_threads: threads) do |website|
+        logger.info("Website:#{website.truncated_url} crawling started.") if verbose?
+        result = Crawler.result(website)
+        if result.success?
+          logger.info("Crawler:#{website.truncated_url} succeeded.")
+        else
+          logger.info("Crawler:#{website.truncated_url} failed - #{result.failure}.")
         end
       end
     end
 
     #
-    # @return [Array<Miteru::Websites>]
+    # @return [Array<Miteru::Website>]
     #
     def websites
       @websites ||= [].tap do |out|
@@ -34,30 +41,33 @@ module Miteru
           result = feed.result
           if result.success?
             websites = result.value!
-            Miteru.logger.info("Feed:#{feed.source} has #{websites.length} websites.") if verbose?
+            logger.info("Feed:#{feed.source} has #{websites.length} websites.") if verbose?
             out << websites
           else
-            Miteru.logger.warn("Feed:#{feed.source} failed - #{result.failure}")
+            logger.warn("Feed:#{feed.source} failed - #{result.failure}")
           end
         end
       end.flatten.uniq(&:url)
     end
 
     #
-    # @return [Array<Miteru::Feeds::Base>]
+    # @return [Array<Miteru::Website>]
     #
-    def feeds
-      Miteru.feeds.map(&:new)
-    end
-
-    private
-
-    def threads
-      Miteru.config.threads
+    def non_cached_websites
+      @non_cached_websites ||= [].tap do |out|
+        out << if cache?
+          websites.reject { |website| cache.cached?(website.url) }
+        else
+          websites
+        end
+      end.flatten.uniq(&:url)
     end
 
-    def verbose?
-      Miteru.config.verbose
+    #
+    # @return [Array<Miteru::Feeds::Base>]
+    #
+    def feeds
+      @feeds ||= Miteru.feeds.map(&:new)
     end
   end
 end

data/lib/miteru/service.rb

@@ -24,5 +24,49 @@ module Miteru
         new.result(...)
       end
     end
+
+    private
+
+    def auto_download?
+      Miteru.config.auto_download
+    end
+
+    #
+    # @return [SemanticLogger]
+    #
+    def logger
+      Miteru.logger
+    end
+
+    def cache?
+      Miteru.cache?
+    end
+
+    def sidekiq?
+      Miteru.sidekiq?
+    end
+
+    #
+    # @return [Miteru::Cache]
+    #
+    def cache
+      Miteru.cache
+    end
+
+    def threads
+      Miteru.config.threads
+    end
+
+    def verbose?
+      Miteru.config.verbose
+    end
+
+    def cache_prefix
+      Miteru.config.cache_prefix
+    end
+
+    def cache_ex
+      Miteru.config.cache_ex
+    end
   end
 end

data/lib/miteru/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Miteru
-  VERSION = "2.1.1"
+  VERSION = "2.2.1"
 end

data/lib/miteru/website.rb

@@ -4,6 +4,8 @@ require "oga"
 
 module Miteru
   class Website < Service
+    include Concerns::UrlTruncatable
+
     # @return [String]
     attr_reader :url
 
@@ -41,17 +43,6 @@ module Miteru
       (href_links + possible_file_links).compact.uniq
     end
 
-    #
-    # @return [String]
-    #
-    def truncated_url
-      url.truncate(64)
-    end
-
-    def defanged_truncated_url
-      truncated_url.to_s.gsub(".", "[.]")
-    end
-
     def info
       "#{defanged_truncated_url} has #{kits.length} kit(s) (Source: #{source})"
     end

data/lib/miteru.rb

@@ -31,8 +31,12 @@ require "miteru/errors"
 # Concerns
 require "miteru/concerns/database_connectable"
 require "miteru/concerns/error_unwrappable"
+require "miteru/concerns/url_truncatable"
 
 # Core classes
+require "miteru/service"
+
+require "miteru/cache"
 require "miteru/config"
 require "miteru/http"
 
@@ -93,6 +97,14 @@ module Miteru
       env == "development"
     end
 
+    def cache?
+      !Miteru.config.cache_redis_url.nil?
+    end
+
+    def cache
+      @cache ||= Cache.new(Miteru.config.cache_redis_url)
+    end
+
     #
     # @return [Boolean]
     #
@@ -117,8 +129,6 @@ module Miteru
 end
 
 # Services
-require "miteru/service"
-
 require "miteru/crawler"
 require "miteru/downloader"
 require "miteru/kit"

data/miteru.gemspec

@@ -1,67 +1,68 @@
 # frozen_string_literal: true
 
-lib = File.expand_path('lib', __dir__)
+lib = File.expand_path("lib", __dir__)
 $LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
-require 'miteru/version'
+require "miteru/version"
 
 Gem::Specification.new do |spec|
-  spec.name = 'miteru'
+  spec.name = "miteru"
   spec.version = Miteru::VERSION
-  spec.authors = ['Manabu Niseki']
-  spec.email = ['manabu.niseki@gmail.com']
-  spec.metadata['rubygems_mfa_required'] = 'true'
+  spec.authors = ["Manabu Niseki"]
+  spec.email = ["manabu.niseki@gmail.com"]
+  spec.metadata["rubygems_mfa_required"] = "true"
 
-  spec.summary = 'A phishing kit collector for scavengers'
-  spec.description = 'A phishing kit collector for scavengers'
-  spec.homepage = 'https://github.com/ninoseki/miteru'
-  spec.license = 'MIT'
+  spec.summary = "A phishing kit collector for scavengers"
+  spec.description = "A phishing kit collector for scavengers"
+  spec.homepage = "https://github.com/ninoseki/miteru"
+  spec.license = "MIT"
 
   # Specify which files should be added to the gem when it is released.
   # The `git ls-files -z` loads the files in the RubyGem that have been added into git.
   spec.files = Dir.chdir(File.expand_path(__dir__)) do
     `git ls-files -z`.split("\x0").reject { |f| f.match(%r{^(test|spec|features)/}) }
   end
-  spec.bindir = 'exe'
+  spec.bindir = "exe"
   spec.executables = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
-  spec.require_paths = ['lib']
+  spec.require_paths = ["lib"]
 
-  spec.add_development_dependency 'bundler', '~> 2.5'
-  spec.add_development_dependency 'capybara', '~> 3.40'
-  spec.add_development_dependency 'coveralls_reborn', '~> 0.28'
-  spec.add_development_dependency 'fuubar', '~> 2.5'
-  spec.add_development_dependency 'mysql2', '~> 0.5'
-  spec.add_development_dependency 'pg', '~> 1.5'
-  spec.add_development_dependency 'rake', '~> 13.1'
-  spec.add_development_dependency 'rspec', '~> 3.13'
-  spec.add_development_dependency 'simplecov-lcov', '~> 0.8'
-  spec.add_development_dependency 'standard', '~> 1.33'
-  spec.add_development_dependency 'test-prof', '~> 1.3'
-  spec.add_development_dependency 'vcr', '~> 6.2'
-  spec.add_development_dependency 'webmock', '~> 3.19'
+  spec.add_development_dependency "bundler", "~> 2.5"
+  spec.add_development_dependency "capybara", "~> 3.40"
+  spec.add_development_dependency "coveralls_reborn", "~> 0.28"
+  spec.add_development_dependency "fuubar", "~> 2.5"
+  spec.add_development_dependency "mysql2", "~> 0.5"
+  spec.add_development_dependency "pg", "~> 1.5"
+  spec.add_development_dependency "rake", "~> 13.1"
+  spec.add_development_dependency "rspec", "~> 3.13"
+  spec.add_development_dependency "simplecov-lcov", "~> 0.8"
+  spec.add_development_dependency "standard", "~> 1.33"
+  spec.add_development_dependency "test-prof", "~> 1.3"
+  spec.add_development_dependency "vcr", "~> 6.2"
+  spec.add_development_dependency "webmock", "~> 3.19"
 
-  spec.add_dependency 'activerecord', '7.1.3'
-  spec.add_dependency 'addressable', '2.8.6'
-  spec.add_dependency 'anyway_config', '2.6.2'
-  spec.add_dependency 'colorize', '1.1.0'
-  spec.add_dependency 'dotenv', '2.8.1'
-  spec.add_dependency 'down', '5.4.1'
-  spec.add_dependency 'dry-files', '1.1.0'
-  spec.add_dependency 'dry-monads', '1.6.0'
-  spec.add_dependency 'http', '5.2.0'
-  spec.add_dependency 'memo_wise', '1.8.0'
-  spec.add_dependency 'oga', '3.4'
-  spec.add_dependency 'parallel', '1.24.0'
-  spec.add_dependency 'puma', '6.4.2'
-  spec.add_dependency 'rack', '3.0.9'
-  spec.add_dependency 'rack-session', '2.0.0'
-  spec.add_dependency 'rackup', '2.1.0'
-  spec.add_dependency 'semantic_logger', '4.15.0'
-  spec.add_dependency 'sentry-ruby', '5.16.1'
-  spec.add_dependency 'sentry-sidekiq', '5.16.1'
-  spec.add_dependency 'sidekiq', '7.2.1'
-  spec.add_dependency 'slack-notifier', '2.4.0'
-  spec.add_dependency 'sqlite3', '1.7.2'
-  spec.add_dependency 'thor', '1.3.0'
-  spec.add_dependency 'thor-hollaback', '0.2.1'
-  spec.add_dependency 'uuidtools', '2.2.0'
+  spec.add_dependency "activerecord", "7.1.3"
+  spec.add_dependency "addressable", "2.8.6"
+  spec.add_dependency "anyway_config", "2.6.3"
+  spec.add_dependency "colorize", "1.1.0"
+  spec.add_dependency "dotenv", "2.8.1"
+  spec.add_dependency "down", "5.4.1"
+  spec.add_dependency "dry-files", "1.1.0"
+  spec.add_dependency "dry-monads", "1.6.0"
+  spec.add_dependency "http", "5.2.0"
+  spec.add_dependency "memo_wise", "1.8.0"
+  spec.add_dependency "oga", "3.4"
+  spec.add_dependency "parallel", "1.24.0"
+  spec.add_dependency "puma", "6.4.2"
+  spec.add_dependency "rack", "3.0.9"
+  spec.add_dependency "rack-session", "2.0.0"
+  spec.add_dependency "rackup", "2.1.0"
+  spec.add_dependency "redis", "5.0.8"
+  spec.add_dependency "semantic_logger", "4.15.0"
+  spec.add_dependency "sentry-ruby", "5.16.1"
+  spec.add_dependency "sentry-sidekiq", "5.16.1"
+  spec.add_dependency "sidekiq", "7.2.1"
+  spec.add_dependency "slack-notifier", "2.4.0"
+  spec.add_dependency "sqlite3", "1.7.2"
+  spec.add_dependency "thor", "1.3.0"
+  spec.add_dependency "thor-hollaback", "0.2.1"
+  spec.add_dependency "uuidtools", "2.2.0"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: miteru
 version: !ruby/object:Gem::Version
-  version: 2.1.1
+  version: 2.2.1
 platform: ruby
 authors:
 - Manabu Niseki
 autorequire: 
 bindir: exe
 cert_chain: []
-date: 2024-02-07 00:00:00.000000000 Z
+date: 2024-02-09 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
@@ -226,14 +226,14 @@ dependencies:
     requirements:
     - - '='
       - !ruby/object:Gem::Version
-        version: 2.6.2
+        version: 2.6.3
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - '='
       - !ruby/object:Gem::Version
-        version: 2.6.2
+        version: 2.6.3
 - !ruby/object:Gem::Dependency
   name: colorize
   requirement: !ruby/object:Gem::Requirement
@@ -416,6 +416,20 @@ dependencies:
     - - '='
       - !ruby/object:Gem::Version
         version: 2.1.0
+- !ruby/object:Gem::Dependency
+  name: redis
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 5.0.8
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - '='
+      - !ruby/object:Gem::Version
+        version: 5.0.8
 - !ruby/object:Gem::Dependency
   name: semantic_logger
   requirement: !ruby/object:Gem::Requirement
@@ -566,6 +580,7 @@ files:
 - exe/miteru
 - lefthook.yml
 - lib/miteru.rb
+- lib/miteru/cache.rb
 - lib/miteru/cli/application.rb
 - lib/miteru/cli/base.rb
 - lib/miteru/cli/database.rb
@@ -575,6 +590,7 @@ files:
 - lib/miteru/commands/web.rb
 - lib/miteru/concerns/database_connectable.rb
 - lib/miteru/concerns/error_unwrappable.rb
+- lib/miteru/concerns/url_truncatable.rb
 - lib/miteru/config.rb
 - lib/miteru/crawler.rb
 - lib/miteru/database.rb