Persian encoding

[bshafiei-ir]

It doesn't work for Unicode characters such as "نوشتار فارسی".
How should I do?

[ctolkien]

Hi,

Hrm, so this is really supposed to be language agnostic, it will strip diacratic to their base a-z character. I don't really have any idea on how to handle non-latin languages.

Can you advise what a slugged output should be for those characters?

[bshafiei-ir]

Yes. For example

input;
جای مردان سیاست بنشانید درخت تا هوا تازه شود.

output:
جای_مردان_سیاست_بنشانید_درخت_تا_هوا_تازه_شود.

I have seen one git project but the main problem is that, it wrote on .net framework 4.6 and has some conflict on .net core 2

[ctolkien]

This lib has a "known" character set based on the latin alphabet... Not really sure how we could handle this one.

[arisliang]

handle in the same way? not the most dynamic way, but it seem to works. another project for ref

[ctolkien]

Resurfacing this..... the linked project from @arisliang has a mapping from unicode characters from other languages to an ascii set.

With the source text of: نوشتار فارسی, that would result in nwshtr-frsy

Does that actually solve the issue that people have here?

[nnhhttmmjjyy]

As an example, there is another library used on the frontend https://github.com/pid/speakingurl, which also uses character mapping and can be taken as a reference. Character mapping can be found in this file

Input:
Наизусть

Output:
naizust

A character mapping can be created by comparing the characters in the other libraries to include all common characters.

[ctolkien]

Please see #82 for feedback.

[ctolkien]

This is currently in a prerelease v5, you can trial it via:

dotnet add package Slugify.Core --version 5.0.0-prerelease.4

cc @bshafiei-ir , @nnhhttmmjjyy