Site virtual host doesn't work in case it has UTF-8 characters

• Type: Bug
• Status: Closed
• Resolution: Fixed
• Affects Version/s: 7.1.X, Master
• Fix Version/s: 7.0.0 DXP FP89, 7.0.10.13 DXP SP13, 7.0.X, 7.1.10 DXP FP16, 7.1.X, 7.2.10 DXP FP4, 7.2.X, 7.3.0 CE GA1, 7.3.10 DXP GA1, Master
• Component/s: Core Infrastructure
• Labels:
  • ee-ts
  • liferay-dxp-70-sp13
  • liferay-fixpack-de-89-7010
  • liferay-fixpack-dxp-16-7110
  • liferay-fixpack-dxp-4-7210
• Branch Version/s:
  7.2.x, 7.1.x, 7.0.x
• Backported to Branch:
  Committed
• Fix Priority: 3
• Git Pull Request: https://github.com/brianchandotcom/liferay-portal/pull/80726

Sites virtual hosts doesn't work in case they have UTF-8 characters.

Root cause of this issue is:

• We are storing virtual hosts in virtualhost table as UTF-8
• But "Host:" header that we receive from browser is encoden using Punycode (see https://www.punycoder.com/ )

As a workaround you can configure them encoded using Punycode, but the problem is all links are also written to web page using that encoding.

Solution: Create a fallback code that in case the virtualhost is not found and it starts with "xn--" try decoding it using punycode and search again using the decoded UTF-8 string

Steps to reproduce

1. Create a new site
2. Add a test page to that site with a portlet
3. Configure its public virtual host with ññ.example (it has UTF-8 characters)
4. Edit your /etc/hosts (Linux) or C:\Windows\System32\drivers\etc\hosts (Windows) file and add following line:

   127.0.0.1	xn--idaa.example

5. Open http://ññ.example:8080 in a browser:
   • Expected behavior: Home page of created site is created
   • Wrong behavior: Created site is not displayed
6. Additional check: open page source code and check the links:
   • Expected behavior: All links are created with "ññ.example"
   • Wrong behavior: Links are created with other domain: localhost", "127.0.0.1" or "xn--idaa.example"