[#46413] - [6.1] Update robot detection logic and add new user agents
- Pending
- Medium
- Build: 6.1-dev
- # 46413
- Diff
- BrainforgeUK:patch-2
User tests: Successful: Unsuccessful:
Now uses strpos instead of preg_match.
Summary of Changes
In isRobot() use strpos() instead of preg_match() as that is faster.
See https://www.php.net/manual/en/function.preg-match.php#refsect1-function.preg-match-notes
Testing Instructions
A simple test available in this plugin.
bfisrobottest.zip
Install and on enabling the front-end displays the following on success:
Test (false):
Test (true):1
Test (true):1
Actual result BEFORE applying this Pull Request
Robots detected.
Expected result AFTER applying this Pull Request
Robots detected.
Its a big list of robots - possibly a detectable performance improvement,
Link to documentations
Please select:
-
Documentation link for docs.joomla.org:
-
No documentation changes for docs.joomla.org needed
-
Pull Request link for manual.joomla.org:
-
No documentation changes for manual.joomla.org needed
| Status | New | ⇒ | Pending |
joomla-cms-bot
- | Category | ⇒ | Libraries |
Agree I was wondering about the best way forward. See comment I've added to #45323
Would moving this big list into a plugin be a first start? e.g.
plg_browserbot_legacy
Can then investigate other bot detectors like this.
https://fingerprint.com/try/bot-detection/
plg_browserbot_fingerprint
Up to site admin to enable the one(s) they want to use.
| Labels |
Added:
PR-6.1-dev
|
||
Added support for isrobot event in browserbot plugin folder.
Falls through to legacy list if no plugin found.
In future can add browserbot plugins whenever convenient.
Install and enable this plugin to test:
plg_browserbot_test.zip
Invocation:
To emulate a normal user
yourtestsite/index.php?isrobot=0
And to emulate a robot
yourtestsite/index.php?isrobot=1
Legacy isRobot behaviour
yourtestsite/index.php
Could you use in_array()? How would it compare for speed?
This comment was created with the J!Tracker Application at issues.joomla.org/tracker/joomla-cms/46413.
As I wrote in the referenced issue we should not be trying to maintain a list of robots ourselves and should be using a well maintained library instead
As I wrote in the referenced issue we should not be trying to maintain a list of robots ourselves and should be using a well maintained library instead
... The plugin solution opens up the possibility to use any number of free / freemium / paid for libraries. Maybe one plugin could be distributed with Joomla Project like Google Captcha is incluided. Other possibilities left to 3rd party developers.
Not an intended outcome of this pull request I have a long-standing use case of the Browser class where it looks like using this plugin approach is worth investigating.
I've updated the phpstan-baseline.neon file to make PHPstan pass. Basically it is a false alarm by PHPstan, see #46054 (comment) .
| Labels |
Added:
Feature
PBF
|
||
I have tested this item ✅ successfully on 0d69f9a
I have tested this during PBF 2026, I 'believe' successfully, please @BrainforgeUK confirm.
BEFORE PR, Frontend of the website shows:
Test (false):
Test (true):
Test (true):
AFTER PR, with plugin from Testing Instructions, Frontend of the website shows:
Test (false):
Test (true): 1
Test (true): 1
AFTER PR, with plugin from this comment: #46413 (comment), while
trying: https://master222.pbf20.org/index.php?isrobot=1 or https://master222.pbf20.org/index.php?isrobot=0
Frontend of the website shows:
Test (false):
Test (true): 1
Test (true): 1
(does not seem like the plg_browserbot_test.zip did/does anything or I misunderstood something.
This comment was created with the J!Tracker Application at issues.joomla.org/tracker/joomla-cms/46413.
The list is massively out of date
See #45323