Regular Expression Matching - Tracks

Question

I have a string line that looks like

A GOMUP 59/20 61/30 63/40 64/50 64/60 MUSVA DUTUM

I am trying to write a Regex that matches this string, and returns each of the none spacial text in an array. It has to ensure the first letter is 1 digit.

The Regex I have tried doesn't work how I would expect

#^([A-Z])(?:\s(\S+))+#

Returns

array(3) {
  [0]=>
  array(1) {
    [0]=>
    string(49) "A GOMUP 59/20 61/30 63/40 64/50 64/60 MUSVA DUTUM"
  }
  [1]=>
  array(1) {
    [0]=>
    string(1) "A"
  }
  [2]=>
  array(1) {
    [0]=>
    string(5) "DUTUM"
  }
}

I expect/would like to return

array(10) {
  [0]=>
  array(1) {
    [0]=>
    string(49) "A GOMUP 59/20 61/30 63/40 64/50 64/60 MUSVA DUTUM"
  }
  [1]=>
  array(1) {
    [0]=>
    string(1) "A"
  }
  [2]=>
  array(1) {
    [0]=>
    string(5) "GOMUP"
  }
  [3]=>
  array(1) {
    [0]=>
    string(5) "59/20"
  }
  [4]=>
  array(1) {
    [0]=>
    string(5) "61/30"
  }
  [5]=>
  array(1) {
    [0]=>
    string(5) "63/40"
  }
  [6]=>
  array(1) {
    [0]=>
    string(5) "64/50"
  }
  [7]=>
  array(1) {
    [0]=>
    string(5) "64/60"
  }
  [8]=>
  array(1) {
    [0]=>
    string(5) "MUSVA"
  }
  [9]=>
  array(1) {
    [0]=>
    string(5) "DUTUM"
  }
}

How can this be achieved? I am using preg_match in PHP.

Answer 1

To split your string and check that the first item is a single letter at the same time, you can use this pattern:

$pattern = '~^[A-Z]\b|\G\s+\K\S+~';

$subject = 'A GOMUP 59/20 61/30 63/40 64/50 64/60 MUSVA DUTUM';

preg_match_all($pattern, $subject, $matches);

print_r($matches[0]);

You obtain:

Array
(
    [0] => A
    [1] => GOMUP
    [2] => 59/20
    [3] => 61/30
    [4] => 63/40
    [5] => 64/50
    [6] => 64/60
    [7] => MUSVA
    [8] => DUTUM
)

If I test the string ZZ A GOMUP 59/20 61/30 63/40 64/50 64/60 MUSVA DUTUM the pattern fails and no result is returned.

However you have the possibility to find the first substring which begins with a single letter using this pattern:

$pattern = '~^(?>\S{2,}\s+)*\K[A-Z]\b|\G\s+\K\S+~';

---

Pattern1 details: ~^[AZ]\\b|\\G\\s+\\K\\S+~

~          # pattern delimiter
^          # begining of the string anchor
[A-Z]\b    # single uppercase letter with a word boundary
|          # OR
\G         # contiguous match from the last
\s+        # one or more white characters (spaces, tab, newlines...)
           # which can be replaced by ' +' for your example string
\K         # reset the match before (remove the spaces from the result)
\S+        # all that is not a space
~          # pattern delimiter

Pattern2 details: ~^(?>\\S{2,}\\s+)*\\K[AZ]\\b|\\G\\s+\\K\\S+~

~          # pattern delimiter
^          # begining of the string anchor
(?>        # open a group (atomic here but you can use '(?:' instead)
  \S{2,}   # a non space character repeated at least two times
  \s+      # one or more spaces
)*         # repeat the group zero or more times
\K         # reset the begining of the match

and after it is like Pattern1.

Answer 2

Regular expressions in PHP don't allow a variable number of matching groups, so you'll have to write a group for every part of the string. See eg http://www.regular-expressions.info/captureall.html

It would be easier to split the string by whitespace with explode or preg_split, and only then do the additional checks.

Answer 3

if (preg_match_all('#([A-Z]+)|([\d]+/[\d]+)#', $text, $matches)){
    print_r($matches[0]);
}

Output:

Array
(
    [0] => A
    [1] => GOMUP
    [2] => 59/20
    [3] => 61/30
    [4] => 63/40
    [5] => 64/50
    [6] => 64/60
    [7] => MUSVA
    [8] => DUTUM
)