Replace eth-ens-namehash with ens-normalize.js

[adraffy]

Copied from other fork: This is my first-attempt at replacing the eth-ens-namehash repo. I essentially deleted everything and started fresh. I kept the existing functionality (normalize() and hash()) and exposed my new functionality via import renames. Except for the hash.js shim, /derive/, /src/ and /validate/ are direct-copies from @adraffy/ens-normalize.js.

For more information, check Project Readme, Derive Readme, and Validation Readme.

---

Note: This build still has a few areas that are still under development:

• I just completed a big review of Latin, but it probably needs another pass or two to fix mistakes.
• There is an excessive amount of arrows, crosses, and star-like characters that need simplified -- there might not be a good solution.
• I've already limited combining marks to (1), and limited what they can attach to (like emoji, isolates, etc.), but I'm thinking that's not enough, and what we really need is a whitelisted sequence of valid combinations, possibly per script, because even if you lock down what marks can be used, marks can still be applied to characters that shouldn't have them.

I will keep this repo in sync as a finish the last few things.

[adraffy]

@adraffy/ens-normalize.js has been pretty stable.

This repo now contains the minimal data necessary for operation and validation conformance.

It is a direct replacement for the prior library except it is ESM not CJS.

I removed all of the build code as the 99.9% use-case is normalize()/namehash()—not rebuilding the spec from scratch.

[adraffy]

Updated to match @adraffy/ens-normalize.js 1.9.0

[adraffy]

https://discuss.ens.domains/t/draft-approval-of-ens-name-normalization-standard/16957/7

This should now be a zero-change drop-in replacement for the old library.