Match just once with regex

Question

I'm using this regex to mach some words without numbers and it works well

(?:searchForThis|\G).+?(\b[^\d\s]+?\b)

The problem that Regex searching the entire document and not only in the line that contains searchForThis

So if I have 2 times searchForThis it will take them twice

I want to stop it only on that 1st line so it will not search the other lines after Any help please?

I'm using Regex with php

Example of the problem here: http://www.rubular.com/r/vPhk8VbqZR

In the example you will see :

Match 1
1.  word
Match 2
1.  worldtwo
Match 3
1.  wordfive
Match 4
1.  word
Match 5
1.  worldtwo
Match 6
1.  wordfive

But I need only :

Match 1
1.  word
Match 2
1.  worldtwo
Match 3
1.  wordfive

You will see that it's doing twice

===========Edit for more details as asked ===========================

In my php I have :

define('CODE_REGEX', '/(?:searchForThis|\G(?<!^)).*?(\b[a-zA-Z]+\b)/iu')

Output :

if (preg_match_all(CODE_REGEX, $content, $result))
            return trim($result[1][0].' '.$result[1][1].' '.$result[1][2].' '.$result[1][3].' '.$result[1][4].' '.$result[1][5]);

Thank you

I think you're looking for this: [Link][1] [1]: http://stackoverflow.com/questions/2503413/regular-expression-to-stop-at-first-match — Frederik Voordeckers, Dec 06 '13 at 13:36
If I use preg_much it doesn't what I want It's capturing names without numbers Just I want that it searched on the first line found and not going ahead for the hole document — amorino, Dec 06 '13 at 13:59
@amorino: Rather than rubular link provide your input and expected match in the question itself. — anubhava, Dec 06 '13 at 14:00
@anubhava I'm using cakePhp as this : `public static function getDoctorName($content) { if (PROVIDER == '1') { if (preg_match(CodeA_REGEX, $content, $result)) return trim($result[0]); } else if (PROVIDER == '2') { if (preg_match_all(CODE_REGEX, $content, $result)) return trim($result[1][0].' '.$result[1][1].' '.$result[1][2].' '.$result[1][3].' '.$result[1][4].' '.$result[1][5]); } return null; }` — amorino, Dec 06 '13 at 14:22
@amorino: Not the code I was asking that does string `$content` hold? — anubhava, Dec 06 '13 at 14:24

Casimir et Hippolyte · Accepted Answer · 2013-12-06T15:14:56.063

3

You can use this pattern instead:

(?:\A[\s\S]*?searchForThis|\G).*?(\b[a-z]+\b)/iu

or

(?:\A(?s).*?searchForThis|\G)(?-s).*?(\b[a-z]+\b)/iu

To deal with multiple line between the first "searchForThis" and others or the end of the string, you can use this: (with your example string you will obtain "After" and "this".)

(?:\A.*?searchForThis|\G)(?>[^a-z]++|\b[a-z]++\S)*?(?!searchForThis)(\b[a-z]+\b)/ius

Note: in all the three pattern you can replace \A with ^ since the multiline mode is not used. Be carefull with rubular that is designed for ruby regexes: m in ruby = s in php (that is the dotall/singleline mode), m in php is the multiline mode (each start of the line can be matched with ^)

edited Dec 06 '13 at 15:14

answered Dec 06 '13 at 14:25

Casimir et Hippolyte

88,009
5
94
125

Thank youuuuuuuuu a lot I was searching for this for 7 days and nights Thank you thank you Casimir et Hippolyte – amorino Dec 06 '13 at 14:31
Great regex, but it won't work if we've got another useless line before. – zessx Dec 06 '13 at 14:31
\A[\s\S]*? tell to search only on that line? – amorino Dec 06 '13 at 14:32
No, it means `Start of the input string, followed by any char, x time`. Note `\A` differs from `^` as it match the start of the **input string**, not the **line**. – zessx Dec 06 '13 at 14:36
@zessx: Yes however, you can use `^` instead, since the multiline mode is not set, it has the same meaning. – Casimir et Hippolyte Dec 06 '13 at 14:44
1

Interesting... @amorino You can use `(?s:^.*?searchForThis|\G).*?(?\b[a-z]+\b)/iu` to get all words in `$results['words']` – zessx Dec 06 '13 at 14:48
@Casmir et Hippolyte and @zessx Thank you a lot for all that precious help The solution 1 worked like a charm the 2 and 3 no I added the suggestion of zessx so : `(?:\A[\s\S]*?searchForThis|\G).*?(?\b[a-z]+\b)/iu` and `$results['words']` You saved my life ;) – amorino Dec 06 '13 at 15:03
@amorino: Test 2 and 3 directly in php, or with http://regex.larsolavtorvik.com . Rubular is for Ruby – Casimir et Hippolyte Dec 06 '13 at 15:07
@Casimir et Hippolyte directly in php and in the rubular site – amorino Dec 06 '13 at 15:11

zessx · Answer 2 · 2013-12-06T14:50:13.583

You can make it in two stages :

// get the first line with 'searchForThis'
preg_match('/searchForThis(?<line>.*)\n/m', $text, $results);
$line = $results['line'];

// get every word from this line
preg_match_all('/\b[a-z]+\b/i', $line, $results);
$words = $results[0];

Another way, based on the great Casimir's answer (just for readibility) :

preg_match_all('/(?s:^.*?searchForThis|\G).*?(?<words>\b[a-z]+\b)/iu', $str, $results);
$words = $results['words'];

Match just once with regex

2 Answers2