.GetLeftPart does unicode character encoding under the hood #11253
Description
Hi
We learned that the .GetLeftPart method performs Unicode character encoding under the hood. This caused a subtle bug that went undetected for months. It should be called in https://learn.microsoft.com/en-us/dotnet/api/system.uri.getleftpart?view=net-9.0
Initially we used this method to remove SAS token later discovered it does more than removing SAS token.
We confirmed this manually and then decompiled the code to verify it.
Metadata
Metadata
Assignees
Type
Projects
Milestone
Relationships
Participants
Activity
[-].GetLeftPart does unicode encoding under the hood[/-][+].GetLeftPart does unicode character encoding under the hood[/+]dotnet-policy-service commentedon Apr 28, 2025
Tagging subscribers to this area: @dotnet/ncl
liveans commentedon Apr 28, 2025
/cc @MihaZupan
krwq commentedon May 8, 2025
Let me post offline comment on behalf of @MihaZupan:
GetLeftPartis just a shorthand for picking the set ofUriComponentflags to pass toGetComponents.E.g. calling
.GetLeftPart(UriPartial.Authority)is the same as calling.GetComponents(UriComponents.Scheme | UriComponents.UserInfo | UriComponents.Host | UriComponents.Port, UriFormat.UriEscaped)This is how basically all properties on Uri behave (Host, Query, PathAndQuery, AbsoluteUri, ...)