Breaking Change: don't try to normalize unicode in collision testing.

[nhz2]

Initially, I tried to check if two entries had the same name in a case-insensitive way using lowercase.

The function I should have used was Unicode.normalize

However, normalize fails on invalid UTF-8, and it is not clear that normalize is doing the same kind of normalization Windows does. In addition, on Windows it is possible to make a specific directory case sensitive.
https://learn.microsoft.com/en-us/windows/wsl/case-sensitivity#change-the-case-sensitivity-of-files-and-directories

Because of these complications, I am no longer attempting to normalize or lowercase entry names when checking for collisions.

I also am now checking if a new file is an existing implicit or explicit directory, and if a new implicit or explicit directory is an existing file.

This means you will now get an error if you try to create "a.txt/b.txt" and "a.txt" in the same archive.

[codecov-commenter]

Codecov Report

Patch coverage: 100.00% and project coverage change: +0.05 🎉

Comparison is base (ea9ef99) 98.04% compared to head (cb3e3a2) 98.10%.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #25      +/-   ##
==========================================
+ Coverage   98.04%   98.10%   +0.05%     
==========================================
  Files           6        6              
  Lines         820      843      +23     
==========================================
+ Hits          804      827      +23     
  Misses         16       16              

Impacted Files	Coverage Δ
src/types.jl	100.00% <ø> (ø)
src/filename-checks.jl	100.00% <100.00%> (ø)
src/writer.jl	100.00% <100.00%> (ø)

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.