ZnUrl is not parsing some URL

[Ducasse]

So we should bullet proof and catch the error during the conversion.
I do not really know what to do once I get the error caught

Typing this in microdown

Microdown parse: 'https://doc-si.inria.fr/spaces/SU/pages/109216287/Multi-Factor+Authentication#MultiFactorAuthentication-Présent
ation:'

'https://doc-si.inria.fr/spaces/SU/pages/109216287/Multi-Factor+Authentication#MultiFactorAuthentication-Présent
ation:' asUrl

[kasperosterbye]

The error is the that it ZnUrl does not speak French. If one changes MultiFactorAuthentication-Présentation: to MultiFactorAuthentication-Presentation: (é to e), it works.

So it is an issue about where one should change é to %C3%A9

RFC 3986 Standard: URLs should only contain ASCII characters in the basic set

a) The file name is not standard, so change the file name
b) We can change the ZnUrl to fix this. But should we do that for west european letters (the danish ø gives the same problem)
c) Add a test in ZnUrl that gives an error if the file name is not following RFC 3986 Standard.

Solution a) is the easy, c) is the good, and b) just points to many many alphabetic issues to be handled (Russian, Chinese,...)

a) - no code is needed
c) - should be relatively simple, but I have no clue where the GitHub issue is

[Ducasse]

At least we should catch the error because we do not know what people are accessing.