Don't send X-Content-Type-Options HTTP header for certain WebPage impl to workaround CORB blocking
To disable globally this new behavior introduced in 9865bf07, set the `security.enable_header_xcontent_type_options` config parameter to false
Thanks @Molkobain for the review !
N°2504 - Add feature to connect Office mail box with OAuth2 for Microsoft Graph
N°5102 - Allow to send emails (eg. notifications) using GSuite SMTP and OAuth
* Config messages
* Fix unit tests
N°2504 - Add feature to connect Office mail box with OAuth2 for Microsoft Graph
N°5102 - Allow to send emails (eg. notifications) using GSuite SMTP and OAuth
* 2.7 migration
See following constants PHPDoc for information :
* core/config.class.inc.php ITOP_VERSION
* core/config.class.inc.php ITOP_CORE_VERSION
* setup/itopdesignformat.class.inc.php ITOP_DESIGN_LATEST_VERSION
- new Jenkinsfile and .jenkins removal to launch phpunit/behat tests
triggered on both iTop build and push.
- N°3053 - Check XML conversion methods
- N°3057 - New build recipe
- N°3059 - Automatically set the documentation URLs
- N°3052 - Check community modules XML version against latest version
- N°3054 - Check community modules version against major version
- N°3062 - setup.css file integrity test
- N°3060 - Check consistency between the list of modules and installation.xml
- Add exclusion group for CI
- N°3061 - Automatically check the installation.xml consistency
Now log file name is unchanged : current log is still /log/error.log \o/
Rotation check (using file last modification time) is done :
* on each file write : we don't want to miss calls if session last from 23:59:59 to 00:01 for example ! Though the filemtime() call is done once per session to lower performance impacts
* using a new background process (LogFileRotationProcess)
File renaming on setup is therefore removed.
Also the interface is renamed (from ILogFileNameBuilder to iLogFileNameBuilder) to conform to iTop convention.
- add possibility to inject var using string patterns (ie: `'%env(DB_HOST)?:localhost%`)
- on WriteToFile, preserve the non interpreted value when the interpreted value is kept the same
- added unit tests for both behaviours
- minor bugfix (default value in comment was wrong) and code readability improvements
