the-moisrex/webpp

IDNA: Internationalizing Domain Names for Applications

Opened this issue · 2 comments

I hate dealing with Unicode. It's just too much work!

Link to the UTS

  • to ascii

Folder structure?

  • Inside strings
  • Inside uri/idna
  • Inside idna
  • Inside strings/unicode
  • Inside unicode
  • Inside unicode/idna

ChatGPT Response 😆

Based on the existing folders and the specific context of IDNA (Internationalizing Domain Names for Applications), here's a suggested folder structure:

- application/
- async/
- beast/
- cgi/
- common/
- concurrency/
- configs/
- convert/
- crypto/
- db/
- extensions/
- fcgi/
- http/
- hub/
- io/
- ip/
- json/
- libs/
- logs/
- main.hpp
- memory/
- socket/
- std/
- storage/
- strings/
  - unicode/
    - idna/
- time/
- traits/
- uri/
  - idna/
- utils/
- validators/
- version.hpp
- views/

In this structure, the folders are organized to accommodate the IDNA-related code and maintain consistency with the existing folder structure. The "strings" folder contains a subfolder "unicode," which further has a subfolder "idna" for IDNA-specific code related to string manipulation. Similarly, the "uri" folder has a subfolder "idna" for IDNA-related code specific to URI handling.

It's important to note that this is a suggested structure based on the limited information provided. The actual folder structure may vary depending on the specific needs and conventions of your project or organization.