Files

dwindown e8fbfb14c1 fix: prevent asset conflicts between React and Grid.js versions

Add coexistence checks to all enqueue methods to prevent loading
both React and Grid.js assets simultaneously.

Changes:
- ReactAdmin.php: Only enqueue React assets when ?react=1
- Init.php: Skip Grid.js when React active on admin pages
- Form.php, Coupon.php, Access.php: Restore classic assets when ?react=0
- Customer.php, Product.php, License.php: Add coexistence checks

Now the toggle between Classic and React versions works correctly.

Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>

2026-04-18 17:02:14 +07:00

2.7 KiB

Raw Blame History

Changelog

Version 3.0.1

Fixed bug with https: URLs defaulting to port 80 instead of 443 if no port is specified. Thanks to @dskvr for reporting

This affects comparing URLs with the default HTTPs port to URLs without it. For example, comparing https://example.com/ to https://example.com:443/ or vice versa.

They should be treated as equivalent but weren't due to the incorrect port being used for https:.

Version 3.0.0

Changed to using global URL object instead of importing. – Thanks to @brendankenny

Version 2.4.0:

Added Typescript definitions. – Thanks to @danhab99 for creating
Added SECURITY.md policy and CodeQL scanning

Version 2.3.0:

Fixed bug where if the user-agent passed to isAllowed() / isDisallowed() is called "constructor" it would throw an error.

Added support for relative URLs. This does not affect the default behavior so can safely be upgraded.

Relative matching is only allowed if both the robots.txt URL and the URLs being checked are relative.

For example:

var robots = robotsParser('/robots.txt', [
    'User-agent: *',
    'Disallow: /dir/',
    'Disallow: /test.html',
    'Allow: /dir/test.html',
    'Allow: /test.html'
].join('\n'));

robots.isAllowed('/test.html', 'Sams-Bot/1.0'); // false
robots.isAllowed('/dir/test.html', 'Sams-Bot/1.0'); // true
robots.isDisallowed('/dir/test2.html', 'Sams-Bot/1.0'); // true

Version 2.2.0:

Fixed bug that with matching wildcard patterns with some URLs – Thanks to @ckylape for reporting and fixing
Changed matching algorithm to match Google's implementation in google/robotstxt
Changed order of precedence to match current spec

Version 2.1.1:

Fix bug that could be used to causing rule checking to take a long time – Thanks to @andeanfog

Version 2.1.0:

Removed use of punycode module API's as new URL API handles it
Improved test coverage
Added tests for percent encoded paths and improved support
Added getMatchingLineNumber() method
Fixed bug with comments on same line as directive

Version 2.0.0:

This release is not 100% backwards compatible as it now uses the new URL APIs which are not supported in Node < 7.

Update code to not use deprecated URL module API's. – Thanks to @kdzwinel

Version 1.0.2:

Fixed error caused by invalid URLs missing the protocol.

Version 1.0.1:

Fixed bug with the "user-agent" rule being treated as case sensitive. – Thanks to @brendonboshell
Improved test coverage. – Thanks to @schornio

Version 1.0.0:

Initial release.

2.7 KiB Raw Blame History Unescape Escape