* Qual: Update spelling for pre-select variants
# Qual: Update spelling for pre-select variants
In English, preselect is without the hyphen. Update text and made some translations
related to preselect.
* Qual: Update composant to component and/or adequate translation.
# Qual: Update composant to component and/or adequate translation.
"Composant(s)" was mostly referenced in french file/class comments.
Updated
* Qual: Fix misspellings related to "criteria"
# Qual: Fix misspellings related to "criteria"
* Qual: Fix produt misspellings
# Qual: Fix produt misspellings
Change 'produt' to 'product'.
* Qual: Update French comments with "composants"
#Qual: Update French comments with "composants"
- Translating French comments to English (avoid codespell notice)
* Qual: Fixed typo 'bad practive' to 'bad practice'
# Qual: Fixed typo 'bad practive' to 'bad practice'
* Qual: Update phan.yml to exclude specific files from analysis
- Added file exclusion pattern to match phan configuration
- Added check for empty file list to avoid unnecessary phan execution
* Qual: Update file filtering in phan.yml workflow
The change updates the file filtering process in the phan.yml workflow to correctly redirect the output of the grep command to a temporary file.
* Qual: Ignore $systemfunction always exists
---------
Co-authored-by: Laurent Destailleur <eldy@destailleur.fr>
Recreate the OAuth service with its configured scopes when refreshing an access token (required by some providers like Microsoft v2).
Also keep the previous refresh token when the refreshed token does not include one (e.g. Google returns it only once).
Co-authored-by: caminotravelcenter <caminotravelcenter@localhost>
Co-authored-by: Laurent Destailleur <eldy@destailleur.fr>
* try to fix doc
* try to fix doc
* try to fix doc
* try to fix doc
* try to fix doc
* try to fix doc
* try to fix doc
* try to fix doc
* try to fix doc
The current solutions only uses time() which is a timestamp based on seconds. In some scenarios like automizing mails this is not enough. Even with the mt_rand there could be more likely collisions of the ids which is not allowed as it should be overall unique.
The newly used uniqid used a timestamp based on milliseconds and adds additionally a bit randomness which reduces the collision changes much more.
Co-authored-by: Laurent Destailleur <eldy@destailleur.fr>
