Adds detection for Jaunt, PhantomJS and improves detection for other …

…libraries (#7768) * Adds detection for Jaunt * Improves detection * Move Cypress to libraries * Remove Dezor from this PR * Adds detection for PhantomJS * Move back browser before library * Improves detection for Android Browser * Improves detection for Cypress and PhantomJS
matomo-org · Nov 14, 2024 · 8ad6d49 · 8ad6d49
1 parent aab8aa0
commit 8ad6d49
Show file tree

Hide file tree

Showing 8 changed files with 47 additions and 41 deletions.
diff --git a/Parser/Client/Browser.php b/Parser/Client/Browser.php
@@ -1084,7 +1084,7 @@ public function parse(): ?array
             }
         }
 
-        if (empty($name)) {
+        if (empty($name) || 1 === \preg_match('/Cypress|PhantomJS/', $this->userAgent)) {
             return [];
         }
 

diff --git a/Tests/Parser/Client/fixtures/library.yml b/Tests/Parser/Client/fixtures/library.yml
@@ -695,3 +695,27 @@
     type: library
     name: quic-go
     version: ""
+-
+  user_agent: Jaunt/1.5
+  client:
+    type: library
+    name: Jaunt
+    version: "1.5"
+-
+  user_agent: Android.LVLDM
+  client:
+    type: library
+    name: Android License Verification Library
+    version: ""
+-
+  user_agent: Mozilla/5.0 (X11; Linux aarch64) AppleWebKit/537.36 (KHTML, like Gecko) Cypress/12.1.0 Chrome/106.0.5249.51 Electron/21.0.0 Safari/537.36
+  client:
+    type: library
+    name: Cypress
+    version: 12.1.0
+-
+  user_agent: Mozilla/5.0 (Unknown; Linux x86_64) AppleWebKit/538.1 (KHTML, like Gecko) PhantomJS/2.1.1 Safari/538.1
+  client:
+    type: library
+    name: PhantomJS
+    version: 2.1.1
diff --git a/Tests/fixtures/bots.yml b/Tests/fixtures/bots.yml
@@ -7050,15 +7050,6 @@
     producer:
       name: Ghost Inspector, Inc.
       url: https://www.ghostinspector.com/
--
-  user_agent: Mozilla/5.0 (X11; Linux aarch64) AppleWebKit/537.36 (KHTML, like Gecko) Cypress/12.1.0 Chrome/106.0.5249.51 Electron/21.0.0 Safari/537.36
-  bot:
-    name: Cypress
-    category: Site Monitor
-    url: https://github.com/cypress-io/cypress
-    producer:
-      name: Cypress.io, Inc.
-      url: https://www.cypress.io/
 -
   user_agent: Google-Area120-PrivacyPolicyFetcher
   bot:

diff --git a/Tests/fixtures/podcasting.yml b/Tests/fixtures/podcasting.yml
@@ -494,24 +494,6 @@
     model: ""
   os_family: Unknown
   browser_family: Unknown
--
-  user_agent: Android.LVLDM
-  os:
-    name: Android
-    version: ""
-    platform: ""
-  client:
-    type: browser
-    name: Android Browser
-    version: ""
-    engine: WebKit
-    engine_version: ""
-  device:
-    type: ""
-    brand: ""
-    model: ""
-  os_family: Android
-  browser_family: Android Browser
 -
   user_agent: PRDownloader
   os: [ ]

diff --git a/regexes/bots.yml b/regexes/bots.yml
@@ -4299,14 +4299,6 @@
     name: 'Ghost Inspector, Inc.'
     url: 'https://www.ghostinspector.com/'
 
-- regex: 'Cypress/[\d.]+'
-  name: 'Cypress'
-  category: 'Site Monitor'
-  url: 'https://github.com/cypress-io/cypress'
-  producer:
-    name: 'Cypress.io, Inc.'
-    url: 'https://www.cypress.io/'
-
 - regex: 'Google-Apps-Script'
   name: 'Google Apps Script'
   category: 'Service Agent'

diff --git a/regexes/client/browsers.yml b/regexes/client/browsers.yml
@@ -3290,7 +3290,7 @@
   version: '$1'
 
 #Android Browser
-- regex: '(?<! like )Android'
+- regex: '(?<! like )Android(?!\.)'
   name: 'Android Browser'
   version: ''
   engine:

diff --git a/regexes/client/libraries.yml b/regexes/client/libraries.yml
@@ -5,6 +5,11 @@
 # @license http://www.gnu.org/licenses/lgpl.html LGPL v3 or later
 ###############
 
+- regex: 'PhantomJS(?:/(\d+[.\d]+))?'
+  name: 'PhantomJS'
+  version: '$1'
+  url: 'https://phantomjs.org/'
+
 - regex: 'IPinfoClient/.*/(\d+[.\d]+)'
   name: 'IPinfo'
   version: '$1'
@@ -71,7 +76,7 @@
   url: 'https://azure.microsoft.com/en-us/products/data-factory/'
 
 # Dart
-- regex: 'Dart(?:/(\d+[.\d]+))?'
+- regex: 'Dart/(\d+[.\d]+)'
   name: 'Dart'
   version: '$1'
   url: 'https://dart.dev/'
@@ -124,6 +129,18 @@
   version: '$1'
   url: 'https://www.slimerjs.org/'
 
+# Jaunt
+- regex: 'Jaunt/(\d+[.\d]+)'
+  name: 'Jaunt'
+  version: '$1'
+  url: 'https://jaunt-api.com/'
+
+# Cypress
+- regex: 'Cypress/(\d+[.\d]+)'
+  name: 'Cypress'
+  version: '$1'
+  url: 'https://github.com/cypress-io/cypress'
+
 - regex: 'Wget(?:/(\d+[.\d]+))?'
   name: 'Wget'
   version: '$1'
@@ -162,7 +179,7 @@
   name: 'Java HTTP Client'
   version: '$1'
 
-- regex: 'Java(?:/?(\d+[.\d]+))?'
+- regex: 'Java/?(\d+[.\d]+)'
   name: 'Java'
   version: '$1'
 
@@ -215,7 +232,7 @@
   version: '$1'
   url: 'https://github.com/lostisland/faraday'
 
-- regex: '(?:Go-http-client|Go )/?(?:(\d+[.\d]+))?(?: package http)?'
+- regex: '(?:Go-http-client|^Go )/?(?:(\d+[.\d]+))?(?: package http)?'
   name: 'Go-http-client'
   version: '$1'
 

diff --git a/regexes/client/mobile_apps.yml b/regexes/client/mobile_apps.yml
@@ -2571,7 +2571,7 @@
   version: '$1'
 
 # Electron generic apps
-- regex: ' (?!(?:AppleWebKit|brave|Franz|Mailspring|Notion|Basecamp|Evernote|catalyst|ramboxpro|BlueMail|BeakerBrowser|Dezor|TweakStyle|Colibri|Polypane|Singlebox|Skye|VibeMate|(?:d|LT|Glass|Sushi|Flash|OhHai)Browser|Sizzy))([a-z0-9]*)(?:-desktop|-electron-app)?/(\d+\.[\d.]+).*Electron/'
+- regex: ' (?!(?:AppleWebKit|brave|Cypress|Franz|Mailspring|Notion|Basecamp|Evernote|catalyst|ramboxpro|BlueMail|BeakerBrowser|Dezor|TweakStyle|Colibri|Polypane|Singlebox|Skye|VibeMate|(?:d|LT|Glass|Sushi|Flash|OhHai)Browser|Sizzy))([a-z0-9]*)(?:-desktop|-electron-app)?/(\d+\.[\d.]+).*Electron/'
   name: '$1'
   version: '$2'
-Original file line number
+Diff line change
@@ Expand Up / @@ -1084,7 +1084,7 @@ public function parse(): ?array @@
                 }
             }
-            if (empty($name)) {
+            if (empty($name) || 1 === \preg_match('/Cypress|PhantomJS/', $this->userAgent)) {
                 return [];
             }
@@ Expand Down @@