Ruby 2.0正则表达式和西里尔文

Question

Before ruby 2.0, regex worked this way:

/\A[a-zа-я\d]+\z/i          =~ 'привет' # => 0
/\A[a-z\p{Cyrillic}\d]+\z/i =~ 'привет' # => 0

I updated ruby 2.0, and it has a bug:

/\A[a-zа-я\d]+\z/i          =~ 'привет' # => nil
/\A[a-z\p{Cyrillic}\d]+\z/i =~ 'привет' # => nil

How can I deal with this problem? Without \\d in the character class, it works correctly:

/\A[a-zа-я]+\z/i            =~ 'привет' # => 0

Answer 1

This bug looks similar and may be related to this bug that I asked about before. I reported it to ruby trunk , and it has been accepted as a bug. Hopefully, it will be fixed.

Answer 2

The bug seems to be fixed in ruby-head :

⮀ rvm use ruby-2.0.0-preview2
Using /home/am/.rvm/gems/ruby-2.0.0-preview2
⮀ irb
2.0.0dev :001 > regex = /\A[a-zа-я\d]+\z/i ; regex =~ 'привет'
# ⇒ nil 
⮀ rvm use ruby-2.0.0-preview1
Using /home/am/.rvm/gems/ruby-2.0.0-preview1
⮀ irb
2.0.0dev :001 > regex = /\A[a-zа-я\d]+\z/i ; regex =~ 'привет'
# ⇒ nil 
⮀ rvm use ruby-head
Using /home/am/.rvm/gems/ruby-head
⮀ irb
irb(main):001:0> regex = /\A[a-zа-я\d]+\z/i ; regex =~ 'привет'
# ⇒ 0