How do I specify a valid character property using Oniguruma regexes?

Question

I'm using the oniguruma gem to get unicode-aware regexes in ruby 1.8. According to the syntax documentation, I should be able to use \p{M} or \p{Mark} to match code points with the Mark property.

However, when I do the following

ORegexp.new '\p{M}',
            :options => OPTION_MULTILINE | OPTION_SINGLELINE | OPTION_IGNORECASE | OPTION_EXTEND,
            :syntax => SYNTAX_JAVA, # so we can use character properties
            :encoding => ENCODING_UTF8

I get ArgumentError: Oniguruma Error: invalid character property name {M}. I get the same error if I use {Mark}, or if I use one of the other syntaxes that support \p.

What am I doing wrong? How do I specify a valid character property using Oniguruma regexes?

UPDATE - If I use one of the UTF16 encodings, the regex compiles; but since my strings are in UTF8 that doesn't help. So my question becomes: how do I specify a valid character property using UTF-8 Oniguruma regexes?

Answer 1

Try using

/\p{Mark}

I was reading on an old Ruby blog that using the forward slash would "try to find the value as encoding in the string"

http://www.ruby-forum.com/topic/154384