Ruby regex match hex string

ruby-talk
1

Hey guys,

I was wondering if someone could help me out.
I'm trying to validate a string with a regex but can't get it working.
I need to validate a string containing hex pairs divided by a comma,
like this

11,1,aa,a,1b,3b,55,b6

I got this regex:

/\A([0-9a-fA-F]{1,2},{1})*([0-9a-fA-F]{1,2}){1}\Z/

but it only validates the last two pairs, 55, and b6

It should't validate strings like this:

111 or strings like aaa,aaa,bob,ccc

Can anybody help me out with some pointers?
I'm new to using regex expressions.

Thanks a lot!

···

--
Posted via http://www.ruby-forum.com/.

2

Hi,

What makes you think that only the last two pairs are matched? This is
impossible, because you're using the \A and \Z anchors. Either the whole
string matches or it doesn't match at all.

The regex is actually correct and it does match the example string.

However, you can leave out the {1} quantifier, because it has no effect
(it means repeating the pattern exactly once). And you can make the
regex shorter by using the "i" modifier, which makes it case
insensitive:

/\A(?:[0-9a-f]{1,2},)*[0-9a-f]{1,2}\z/i

···

--
Posted via http://www.ruby-forum.com/.

3

I hope this help

puts "11,1,aa,a,1b,3b,55,b6,6l,mm".scan(/([0-9a-fA-F][0-9a-fA-F])/)

any better idea??

···

--
Posted via http://www.ruby-forum.com/.

4

You can use \h to match a hexadecimal digit as well, which match both
lower and upper case a-f.

  /\A(\h{1,2},)*\h{1,2}\z/

Regards,
Ammar

···

On Sat, Mar 31, 2012 at 9:58 PM, Niels Steves <lists@ruby-forum.com> wrote:

Hey guys,

I was wondering if someone could help me out.
I'm trying to validate a string with a regex but can't get it working.
I need to validate a string containing hex pairs divided by a comma,
like this

11,1,aa,a,1b,3b,55,b6

I got this regex:

/\A([0-9a-fA-F]{1,2},{1})*([0-9a-fA-F]{1,2}){1}\Z/

5

The confusion is that only two capturing groups are returned
(while the first part between brackets matches many times).

1.9.3p125 :001 > s = "11,1,aa,a,1b,3b,55,b6"
=> "11,1,aa,a,1b,3b,55,b6"
1.9.3p125 :002 > s.match(/\A([0-9a-fA-F]{1,2},{1})*([0-9a-fA-F]{1,2}){1}\Z/)
=> #<MatchData "11,1,aa,a,1b,3b,55,b6" 1:"55," 2:"b6">
1.9.3p125 :003 > s.match(/\A([0-9a-fA-F]{1,2},)*([0-9a-fA-F]{1,2})\Z/)
=> #<MatchData "11,1,aa,a,1b,3b,55,b6" 1:"55," 2:"b6">

And as mentioned by gabe, maybe scan is what the OP was after;

1.9.3p125 :004 > s.scan(/[0-9a-f]{1,2}/i).inspect
=> "[\"11\", \"1\", \"aa\", \"a\", \"1b\", \"3b\", \"55\", \"b6\"]"

1.9.3p125 :005 > s.scan(/([0-9a-f]{1,2}),/i).inspect # more selective
on the comma
=> "[[\"11\"], [\"1\"], [\"aa\"], [\"a\"], [\"1b\"], [\"3b\"], [\"55\"]]"

HTH,

Peter

···

On Sat, Mar 31, 2012 at 8:31 PM, Jan E. <lists@ruby-forum.com> wrote:

Hi,

What makes you think that only the last two pairs are matched? This is
impossible, because you're using the \A and \Z anchors. Either the whole
string matches or it doesn't match at all.

The regex is actually correct and it does match the example string.

6

Ammar Ali wrote in post #1054416:

···

On Sat, Mar 31, 2012 at 9:58 PM, Niels Steves <lists@ruby-forum.com> > wrote:

/\A([0-9a-fA-F]{1,2},{1})*([0-9a-fA-F]{1,2}){1}\Z/

You can use \h to match a hexadecimal digit as well, which match both
lower and upper case a-f.

  /\A(\h{1,2},)*\h{1,2}\z/

Just a tiny remark: this one should be more efficient

/\A\h{1,2}(?:,\h{1,2})*\z/

because it avoids backtracking of the NFA and does not use a capturing
group.

Kind regards

robert

--
Posted via http://www.ruby-forum.com/\.

7

Note that hexadecimal digits can be matched by a POSIX character class, or in 1.9 there is a special escape for them:

  /\A(?:[[:xdigit:]]{1,2},)*[[:xdigit:]]{1,2}\z/ # 1.8, 1.9

or

  /\A(?:\h{1,2},)*\h{1,2}\z/ # 1.9

would match a valid string.

Mike

···

On 2012-03-31, at 4:15 PM, Peter Vandenabeele wrote:

On Sat, Mar 31, 2012 at 8:31 PM, Jan E. <lists@ruby-forum.com> wrote:

Hi,

What makes you think that only the last two pairs are matched? This is
impossible, because you're using the \A and \Z anchors. Either the whole
string matches or it doesn't match at all.

The regex is actually correct and it does match the example string.

The confusion is that only two capturing groups are returned
(while the first part between brackets matches many times).

1.9.3p125 :001 > s = "11,1,aa,a,1b,3b,55,b6"
=> "11,1,aa,a,1b,3b,55,b6"
1.9.3p125 :002 > s.match(/\A([0-9a-fA-F]{1,2},{1})*([0-9a-fA-F]{1,2}){1}\Z/)
=> #<MatchData "11,1,aa,a,1b,3b,55,b6" 1:"55," 2:"b6">
1.9.3p125 :003 > s.match(/\A([0-9a-fA-F]{1,2},)*([0-9a-fA-F]{1,2})\Z/)
=> #<MatchData "11,1,aa,a,1b,3b,55,b6" 1:"55," 2:"b6">

And as mentioned by gabe, maybe scan is what the OP was after;

1.9.3p125 :004 > s.scan(/[0-9a-f]{1,2}/i).inspect
=> "[\"11\", \"1\", \"aa\", \"a\", \"1b\", \"3b\", \"55\", \"b6\"]"

1.9.3p125 :005 > s.scan(/([0-9a-f]{1,2}),/i).inspect # more selective
on the comma
=> "[[\"11\"], [\"1\"], [\"aa\"], [\"a\"], [\"1b\"], [\"3b\"], [\"55\"]]"

HTH,

Peter

--

Mike Stok <mike@stok.ca>
http://www.stok.ca/~mike/

The "`Stok' disclaimers" apply.