Feature: Handle Arabic

[1saifj]

This Close #89

[1saifj]

@antonlindstrom @navruzm Your review please

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 100.00%. Comparing base (412e31a) to head (604410a).

Additional details and impacted files

@@            Coverage Diff            @@
##            master       #90   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            2         2           
  Lines          126       197   +71     
=========================================
+ Hits           126       197   +71     

Files with missing lines	Coverage Δ
languages_substitution.go	100.00% <100.00%> (ø)
slug.go	100.00% <100.00%> (ø)

[1saifj]

@antonlindstrom @navruzm @matrixik @kvalev ur review please

[matrixik]

Hello, thank you for this change.

Please revert change to BenchmarkMakeShortSymbols (there is already test for simple text BenchmarkMakeShortAscii).

Also please lower the number of examples in doc.go to 1 example (best would be the trickiest one, it's there to show it's possible, not to show all abilities for one language).

From my google search Arabic doesn't have upper letters but could you add one test for it like:

{"ar", "مرحبا بالعالم", "mrhba-balalm", false},