From: "felix.leg" <felix.leg@...>
Date: 2011-01-25T20:21:11+09:00
Subject: Re: [Ruby 1.9] New String rules?

W dniu 16.01.2011 18:09, Michael Brooks pisze:
>
> "felix.leg" <felix.leg@vp.pl> wrote in message
> news:igv4bb$a49$1@news.onet.pl...
>> I have just updated my Ruby into 1.9.0, and tried to run one of my
>> programs. But I got that response:
>> > zrzutnik.rb:19: invalid multibyte char (US-ASCII)
>> > zrzutnik.rb:19: invalid multibyte char (US-ASCII)
>> > zrzutnik.rb:19: syntax error, unexpected $end, expecting keyword_end
>> > print "Klucz #{kl.inspect} jest nieznany. Zdefiniować
>> > nowy [T/N]? "
>> > ^
>> (at Ruby 1.8 it works fine)
>>
>> Does it mean I can't use any longer any non-latin character in my
>> String-s? WTF?
>>
>> PS: The file ("zrzutnik.rb") is encoded all in utf-8 (without BOM).
>
> Hello:
>
> In Ruby 1.9.x you need to explicitly specify the encoding of your
> strings to fit your requirements. Your default string encoding appears
> to be US-ASCII (mine is CP850). If you want it to be UTF-8 use the
> following to force the string to that encoding:
>
> abc = "abc"
> puts abc.encoding.name # >> US-ASCII
>
> abc.force_encoding("UTF-8")
> puts abc.encoding.name # >> UTF-8
>
> James Edward Gray II wrote an excellent blog on the topic here
> http://blog.grayproductions.net/articles/ruby_19s_string .
>
> Michael
>
>
>
Your proposition will not work, because it first requires a valid 
characters between "". Also, what if they come from external file?

One thing which bugs me yet else is does Ruby have got some core 
settings? I use UTF-8 very often so it might be a better solution (as 
for my thought) to set it as a default encoding instead of US-ASCII. 
Most Ruby-Gems in the Internet are in English and it fits pretty good 
with UTF-8 (in addition if some Gem is not in English an author probably 
have do some preparation to tell Ruby a right encoding). As I read in 
other's responses, the default encoding varies so I don't think my 
setting for US-ASCII was done on propose for some convention.