Regular Expression Grouping

rubyonrails-talk
1

Hi!

I couldn't understand the behavior of this code:

match = 'Today is Feb 23rd, 2003'.match(/Feb 23(rd)?/) a = match.to_a puts a.size # 2 puts a.join(",") # Feb 23rd,rd puts a[0] # Feb 23rd puts a[1] # rd

In my understanding, /Feb 23(rd)?/ is equivalent to /Feb 23|Feb 23rd/ . So, match should not include 'rd'.

Thanks.

2

? + and * are greedy, ie they always try to match as much of the string as possible so rd is part of the match. If you want a non greedy quantifier you need to add ? to it, for example match = 'Today is Feb 23rd, 2003'.match(/Feb 23(rd)??/) match[0] #=> "Feb 23"

Fred

3

Yeah you're right,

match = 'Today is Feb 23rd, 2003'.match(/Feb 23(rd)??/) match[0] #=> "Feb 23" match[1] #=> nil

I expected:

match = 'Today is Feb 23rd, 2003'.match(/Feb 23(rd)?/) # ? is greedy here match[0] #=> "Feb 23rd" match[1] #=> nil

But what I got in ruby 1.8.6 (2007-09-24 patchlevel 111) [i386-mswin32] is:

match[0] #=> "Feb 23rd" match[1] #=> 'rd'

The behavior of '?' being greedy is correct since it matched "Feb 23rd" which is stored in match[0] . But should match[1] not be nil? The regular expression does not match "rd" alone.

thanks

4

match[1] is the first group, which in this example should be 'rd' (unless you're talking about something else).

Fred

5

thanks Fred. I think I misunderstood something.