[MonologBridge] Added SwitchUserTokenProcessor to log the impersonator #37704
Conversation
862ad2b to
3ea8997
Compare
|
Hello, Thanks for your first contribution. I like a lot your idea 👍🏼 Thanks. I have a question about the implementation : Why did you choose the create two Processor instead of doing all the work in the existing processor? |
|
I have added the second processor in order to separate processing of tokens. In my opinion, tokens should be processed by their own processors. Moreover, some projects may not use user impersonation feature, and an abstract processor provides the ability to easily implement your own processor |
| */ | ||
| protected function getKey(): string | ||
| { | ||
| return 'original_token'; |
There was a problem hiding this comment.
I know it's the legacy naming, but I don't really like original_token which kind of misses some context. What about impersonating_token or impersonator_token instead?
There was a problem hiding this comment.
I have used the original_token key because of the original (impersonator) token's getter has name getOriginalToken. As for me, the impersonator_token key is more suitable, but I am not sure that is a good idea to change the original_token key to impersonator_token until the original token's getter is renamed to getImpersonatorToken
There was a problem hiding this comment.
As someone looking at the logs or Sentry reports for instance, impersonator_token is more meaningful than original_token though. The internal name in the code does not matter in this regard.
There was a problem hiding this comment.
Changed to impersonator_token
| */ | ||
| protected function getKey(): string | ||
| { | ||
| return 'original_token'; |
885a1d7 to
2f8651d
Compare
|
Thank you @IgorTimoshenko. |
This pull request adds the ability to log the impersonator user in case of user impersonation feature usage. The current TokenProcessor logs only the current token and there are no ability to simply log the original token, so we need to copy-paste the TokenProcessor and change a few lines to log the original token