Redesign combining table

[eschnett]

Closes #271.

[eschnett]

This is based on and supersedes #271.

[eschnett]

The previously failing normalization tests now succeed. However, others are failing.

It seems the failing ones are mentioned in the exclusions:

# ================================================
# (2) Post Composition Version precomposed characters
#
# These characters cannot be derived solely from the UnicodeData.txt file
# in this version of Unicode.
#
# Note that characters added to the standard after the
# Composition Version and which have canonical decomposition mappings
# are not automatically added to this list of Post Composition
# Version precomposed characters.
# ================================================

2ADC    #  FORKING
1D15E   #  MUSICAL SYMBOL HALF NOTE
1D15F   #  MUSICAL SYMBOL QUARTER NOTE
1D160   #  MUSICAL SYMBOL EIGHTH NOTE
1D161   #  MUSICAL SYMBOL SIXTEENTH NOTE
1D162   #  MUSICAL SYMBOL THIRTY-SECOND NOTE
1D163   #  MUSICAL SYMBOL SIXTY-FOURTH NOTE
1D164   #  MUSICAL SYMBOL ONE HUNDRED TWENTY-EIGHTH NOTE
1D1BB   #  MUSICAL SYMBOL MINIMA
1D1BC   #  MUSICAL SYMBOL MINIMA BLACK
1D1BD   #  MUSICAL SYMBOL SEMIMINIMA WHITE
1D1BE   #  MUSICAL SYMBOL SEMIMINIMA BLACK
1D1BF   #  MUSICAL SYMBOL FUSA WHITE
1D1C0   #  MUSICAL SYMBOL FUSA BLACK

[eschnett]

Yes, the only failing tests are normalization tests involving these characters. Ideas appreciated.

Edit: Solved.

[eschnett]

@stevengj Can you approve the CI run?

[eschnett]

@stevengj Can you approve the CI run?

[stevengj]

CI is running, but is failing with:

Error: This request has been automatically failed because it uses a deprecated version of actions/upload-artifact: v2. Learn more: https://github.blog/changelog/2024-02-13-deprecation-notice-v1-and-v2-of-the-artifact-actions/

This was part of the actions added by @inkydragon in #229 … I'm not actually sure why we have them. Is anyone actually using these artifacts?

[eschnett]

@stevengj I updated the actions. Can you approve the CI workflow again?

[eschnett]

@inkydragon Can you approve the CI workflows?

[inkydragon]

I'm not actually sure why we have them. Is anyone actually using these artifacts?

One of the problems here is that we lack a depbot to automatically upgrade the version of github action.

I would say that this is usually more useful for Windows users who only want compiled libraries, rather than downloading the source code and compiling it themselves.
Of course, since utf8proc is not a program that end-users can use directly, perhaps removing the upload artifacts step would be a good option.

Can you approve the CI workflows?

Sorry, I don't have committing privileges for this repo.

[stevengj]

@eschnett, can you add some documentation of the new combining tables, e.g. add a README.md to the data directory?

[eschnett]

@stevengj I added a comment in the definition of utf8proc_property_struct describing the table layout and lookup.

[stevengj]

In the future, please squash commits when merging.

[c42f]

Just a driveby comment of appreciation for how much nicer the table generator is in the section you've changed. Thank you @eschnett :-)