Using Ruby to scan through a string

Question

I am trying to create a regex to gather info from strings that look like this:

A22xB67-E34...

for any number.

I have the regex:

@spaceCode = "[A-Z]([A-Z0-9][0-9]|[0-9])"
@moveCode=/^(?<one>#{@spaceCode})((?<mode>x|\-)(?<two>#{@spaceCode}))+$/

However I get:

s="A11-A22xA33".scan(@moveCode)
=> [["A11", "11", "xA33", "x", "A33", "33"]]

which is most definatly NOT what I want.

The string could be any length of C22 etc, with either x or - as the seperator, and put it into an array like:

['A22','x',B22','-'.......]

Examples:

"A22xB23-D23xE25" => ['A22','x','B23','=','D23','E25;]
"AA2xA9-A1" => ['AA2','x','A9','-','A1']

Trying this in IRB, I got: "RegexpError: undefined (?...) sequence: /^(?<one>[A-Z]([A-Z0-9][0-9]|[0-9]))((?<mode>x|\-)(?<two>[A-Z]([A-Z0-9][0-9]|[0-9])))+$/" — John Dibling
– John Dibling, Commented Nov 9, 2012 at 15:50
@MartinVidner I was hoping I could get an array of something like ['A11',['-','A22','x','A33']] — KatGaea
– KatGaea, Commented Nov 9, 2012 at 15:53
I have to admit that I don't really know what "irb on rails" means. :( — John Dibling
– John Dibling, Commented Nov 9, 2012 at 15:56
that is expected since u r using named groups which itself contains a group — Anirudha
– Anirudha, Commented Nov 9, 2012 at 15:57

Kenny Grant · Accepted Answer · 2012-11-11 00:27:58Z

2

Presumably you want to find these three alphanumeric codes in isolation? Would this simpler regex (or a variant on it) do what you want?

def decode string
  puts "\nDecoding #{string}"
  code = "[A-Z0-9]{1,3}"
  sep = "[-x=]"
  r = /(?:(#{code})#{sep}?)?/
  string.scan(r)
end

puts decode("A22xA33")
puts decode("A11-A22xA33")
puts decode("A22xB67-E34")
puts decode("A22xC33xD44-E55")
puts decode("A22xB23-D23=E25")

or if you want to capture actions/separators too, something like:

r = /(?:#{code}#{sep}?)?/

or to capture actions/separators in the array as separate items between the codes:

r = /(#{code})?(#{sep})?/

So I think what you want is:

def decode string
  puts "\nDecoding #{string}"
  code = "[A-Z0-9]{1,3}"
  sep = "[-x=]"
  r = /(#{code})?(#{sep})?/
  string.scan(r)
end

or in its simplest form just:

string.scan(/([A-Z0-9]{1,3})?([-x=])?/)

edited Nov 11, 2012 at 0:27

answered Nov 9, 2012 at 15:56

Kenny Grant

9,7132 gold badges35 silver badges47 bronze badges

Sign up to request clarification or add additional context in comments.

3 Comments

KatGaea Over a year ago

Afraid not. basically this is for a variable sequence of moves, that can either be move, or captre, so it could look like AA2xA33 or AA2-A33xB44 or even A22xC33xD44-E55

Kenny Grant Over a year ago

Edited reply to deal with variable length patterns. I think this matches your examples. You should put a list of examples to match in your question to make it clearer.

KatGaea Over a year ago

let me do that, sorry for the trouble caused here, and thankyou

the Tin Man · Accepted Answer · 2012-11-11 05:20:13Z

1

I'd do it this way:

MOVE_REGEX = /[a-z]+\d+/i
REGEX = /(#{ MOVE_REGEX })([x-]?)/i

class String
  def parse_move
    self.scan(REGEX).flatten.reject(&:empty?)
  end
end

"A22xB23-D23xE25".parse_move
# => ["A22", "x", "B23", "-", "D23", "x", "E25"]

"AA2xA9-A1".parse_move
# => ["AA2", "x", "A9", "-", "A1"]

answered Nov 11, 2012 at 5:20

the Tin Man

161k44 gold badges222 silver badges308 bronze badges

Comments

Thilo · Accepted Answer · 2012-11-09 16:38:38Z

0

Seems like this should work, given that field names can be variable length:

def parse_moves(s)    
  s.scan(/([A-Z0-9]+)?([-x])?/).flatten.compact
end

Results for your examples:

1.9.3-p125 :027 > moves = ["A11-A22xA33", "A22xB23-D23xE25", "AA2xA9-A1"]
 => ["A11-A22xA33", "A22xB23-D23xE25", "AA2xA9-A1"] 
1.9.3-p125 :028 > moves.each { |s| puts parse_moves(s).to_s }
["A11", "-", "A22", "x", "A33"]
["A22", "x", "B23", "-", "D23", "x", "E25"]
["AA2", "x", "A9", "-", "A1"]

answered Nov 9, 2012 at 16:38

Thilo

17.8k5 gold badges70 silver badges86 bronze badges

Comments

the Tin Man · Accepted Answer · 2012-11-11 04:56:38Z

0

Use method like split with the regex as ([x-])

Your regex won't work for your input since this would always overwrite the groups spaceCode and mode, i.e. spaceCode and mode would contain only the last matched spaceCode and mode.

For example, the regex ^(\d)+$ for input 3664 would capture only 4 not 3,6,6,4 in the group.

edited Nov 11, 2012 at 4:56

the Tin Man

161k44 gold badges222 silver badges308 bronze badges

answered Nov 9, 2012 at 15:59

Anirudha

32.9k8 gold badges71 silver badges90 bronze badges

1 Comment

KatGaea Over a year ago

That gets it closer, but still only captures the first and last of the regex

Collectives™ on Stack Overflow

Using Ruby to scan through a string

4 Answers 4

3 Comments

Comments

Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

3 Comments

Comments

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related