legitbot 1.2.0 → 1.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6cc8cb39a7c005fd9c8e97d095f482c98db48ebbb570408feb81782e03ce4e88
-  data.tar.gz: 200b76238f1cbd2239233bc82347e5d0f4c00b3856754a87beea7ecc3e80e54e
+  metadata.gz: d709839505242be7ed07c5f6413e607bd8eb127a011765b9f45c1a3e0bba2fbb
+  data.tar.gz: 53a19413068473299ce78887756a2ef50f59859e54d652e4314755178a30028c
 SHA512:
-  metadata.gz: 7f3fe97a9e42feaf5233ae5733428fa7d87f1f5e545860240a1990dcef30a5785adeb349858817dc39a494978a910686778b3bb601f6928890ea1078d488b194
-  data.tar.gz: d4af0fced1c8fc0d629b49bab5da65ecdb945905ce2e2b76f0c78a3ed2aa7c71aa978d048bc11b591bb3f73c4a7981f23f1b52b5a98531e626ec611f6d436aaa
+  metadata.gz: daaf7bbde5fae2ab3937b5e4f4cd936a61f54b80d5ec87eaca8db473977f633c993f267f6aeb101b94404119e58835878ee1433683cf613b476c6fc7f27b34b3
+  data.tar.gz: 1e51fe4ee02eb14bb0a95230e2ca3d4716a56cee199a40f3cfcd4d0999158526f151fa7915a9a3f0a21a9054ebc3862353e1a68a89057c2cf8bbcac4d082afb8

data/README.md

@@ -1,4 +1,4 @@
-# Legitbot ![](https://github.com/alaz/legitbot/workflows/build/badge.svg) [![Gem Version](https://badge.fury.io/rb/legitbot.svg)](https://badge.fury.io/rb/legitbot)
+# Legitbot ![](https://github.com/alaz/legitbot/workflows/build/badge.svg) ![](https://badge.fury.io/rb/legitbot.svg)
 
 Ruby gem to check that an IP belongs to a bot, typically a search
 engine. This can be of help in protecting a web site from fake search
@@ -55,6 +55,7 @@ end
 * [Pinterest](https://help.pinterest.com/en/articles/about-pinterest-crawler-0)
 * [Twitterbot](https://developer.twitter.com/en/docs/tweets/optimize-with-cards/guides/getting-started), the list of IPs is in the [Troubleshooting page](https://developer.twitter.com/en/docs/tweets/optimize-with-cards/guides/troubleshooting-cards)
 * [Yandex robots](https://yandex.com/support/webmaster/robot-workings/check-yandex-robots.xml)
+* [Petal robots (Huawei search)](http://aspiegel.com/petalbot)
 
 ## License
 

data/lib/legitbot.rb

@@ -15,3 +15,4 @@ require_relative 'legitbot/oracle'
 require_relative 'legitbot/pinterest'
 require_relative 'legitbot/twitter'
 require_relative 'legitbot/yandex'
+require_relative 'legitbot/petalbot'

data/lib/legitbot/petalbot.rb

@@ -0,0 +1,10 @@
+# frozen_string_literal: true
+
+module Legitbot # :nodoc:
+  # http://aspiegel.com/petalbot
+  class Petalbot < BotMatch
+    domains 'aspiegel.com.'
+  end
+
+  rule Legitbot::Petalbot, %w[PetalBot]
+end

data/lib/legitbot/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Legitbot
-  VERSION = '1.2.0'
+  VERSION = '1.3.0'
 end

data/test/petalbot_test.rb

@@ -0,0 +1,52 @@
+# frozen_string_literal: true
+
+require 'minitest/autorun'
+require 'legitbot'
+
+class PetalbotTest < Minitest::Test
+  def test_malicious_ip
+    ip = '149.210.164.47'
+    match = Legitbot::Petalbot.new ip
+    assert !match.valid?, msg: "#{ip} is not a real Petalbot IP"
+  end
+
+  def test_valid_ip
+    ip = '114.119.153.50'
+    match = Legitbot::Petalbot.new ip
+    assert match.valid?, msg: "#{ip} is a valid Petalbot IP"
+  end
+
+  def test_malicious_ua
+    bot = Legitbot.bot(
+      'Mozilla/5.0 (compatible;PetalBot; +https://aspiegel.com/petalbot)',
+      '149.210.164.47'
+    )
+    assert bot, msg: 'Petalbot detected from User-Agent'
+    assert !bot.valid?, msg: 'Not a valid Petalbot'
+  end
+
+  def test_valid_ua
+    bot = Legitbot.bot(
+      'Mozilla/5.0 (compatible;PetalBot; +https://aspiegel.com/petalbot)',
+      '114.119.153.50'
+    )
+    assert bot, msg: 'Petalbot detected from User-Agent'
+    assert bot.valid?, msg: 'Valid Petalbot'
+  end
+
+  def test_valid_name
+    bot = Legitbot.bot(
+      'Mozilla/5.0 (compatible;PetalBot; +https://aspiegel.com/petalbot)',
+      '66.249.64.141'
+    )
+    assert_equal :petalbot, bot.detected_as
+  end
+
+  def test_fake_name
+    bot = Legitbot.bot(
+      'Mozilla/5.0 (compatible; PetalBot/2.1; +http://www.google.com/bot.html)',
+      '81.1.172.108'
+    )
+    assert_equal :petalbot, bot.detected_as
+  end
+end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: legitbot
 version: !ruby/object:Gem::Version
-  version: 1.2.0
+  version: 1.3.0
 platform: ruby
 authors:
 - Alexander Azarov
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2020-09-27 00:00:00.000000000 Z
+date: 2020-12-16 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: augmented_interval_tree
@@ -160,6 +160,7 @@ files:
 - lib/legitbot/google.rb
 - lib/legitbot/legitbot.rb
 - lib/legitbot/oracle.rb
+- lib/legitbot/petalbot.rb
 - lib/legitbot/pinterest.rb
 - lib/legitbot/twitter.rb
 - lib/legitbot/validators/domains.rb
@@ -176,6 +177,7 @@ files:
 - test/legitbot/validators/ip_ranges_test.rb
 - test/legitbot_test.rb
 - test/oracle_test.rb
+- test/petalbot_test.rb
 - test/pinterest_test.rb
 - test/twitter_test.rb
 homepage: https://github.com/alaz/legitbot
@@ -198,7 +200,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.1.2
+rubygems_version: 3.1.4
 signing_key:
 specification_version: 4
 summary: 'Validate requests from Web crawlers: impersonating or not?'
@@ -212,6 +214,7 @@ test_files:
 - test/apple_test.rb
 - test/oracle_test.rb
 - test/google_test.rb
+- test/petalbot_test.rb
 - test/botmatch_test.rb
 - test/facebook_test.rb
 - test/twitter_test.rb