Missing stdlib function: unicodedata.normalize

[eddieantonio]

The entire unicodedata module is missing; this includes one of the most useful functions in this module, unicodedata.normalize().

This function can be substituted with the ES6's String.prototype.normalize(), if partial module definitions are okay.

[JdeH]

Hi

I'd like to get an idea of what this entails. Apart from the normalize function, are there other parts of the unicodedata that you often use? Can you indicate which ones are most relevant?

Partial module definitions are quite alright. In fact the whole of Transcrypt is designed with Parato's rule for IT in mind: 90% of the result is achieved with 10% of the effort. So what should be in the unicodedata module to make it do the job in most cases?

KR
Jacques

[eddieantonio]

It really depends upon the localization requirements for each project, but for me, the most useful members of unicodedata are:

1. unicodedata.name
2. unicodedata.category
3. unicodedata.lookup — does the same thing as \N{...} escapes in strings, e.g., "\N{FATHER CHRISTMAS}" == '🎅' == unicodedata.lookup('FATHER CHRISTMAS')
4. unicodedata.numeric and similar str -> numeric type conversions

I whipped up a quick proof-of-concept stub implementation of unicodedata.normalize(); I'll open a pull request soon.