Regural expression strange behaviour

[Josip]

I have the following code (I need to extract everything between
slash):

$buff="0,x,R / 1,y,R / - / 1,y,R";

preg_match_all( "/\s*([0-9]\s*,\s*[a-z]\s*,\s*.\s*)\s*[\/]\s*|\s*(-)\s*[\/]/U",$buff,$izlaz );

var_dump($izlaz );


This outputs:

array(3) {
[0]=>
array(4) {
[0]=>
string(7) "0,x,R /"
[1]=>
string(8) " 1,y,R /"
[2]=>
string(4) " - /"
[3]=>
string(8) " 1,y,R"
}

[1]=>
array(4) {
[0]=>
string(5) "0,x,R"
[1]=>
string(5) "1,y,R"
[2]=>
string(0) ""
[3]=>
string(5) "1,y,R"
}

[2]=>
array(4) {
[0]=>
string(0) ""
[1]=>
string(0) ""
[2]=>
string(1) "-"
[3]=>
string(0) ""
}
}



Problem is in the 3rd row: why is it there? I would expect '-' sign in
2nd row...

[Josip]

Re: Regural expression strange behaviour

On Tue, 18 Jan 2005 22:20:09 +0100, Josip <njuzovi@gmail. com> wrote:
[color=blue]
>I have the following code (I need to extract everything between
>slash):
>
>$buff="0,x,R / 1,y,R / - / 1,y,R";
>
>preg_match_all ("/\s*([0-9]\s*,\s*[a-z]\s*,\s*.\s*)\s*[\/]\s*|\s*(-)\s*[\/]/U",$buff,$izlaz );
>[/color]

Hm, this is partial solution...

preg_match_all( "/(\s*[0-9]\s*,\s*[a-z]\s*,\s*[LR]\s*\s*|\s*[\-]\s*)[\/]\s*/U",$buff,$izlaz );

[John Dunlop]

Re: Regural expression strange behaviour

Josip wrote:
[color=blue]
> I have the following code (I need to extract everything between
> slash):
>
> $buff="0,x,R / 1,y,R / - / 1,y,R";
>
> preg_match_all( "/\s*([0-9]\s*,\s*[a-z]\s*,\s*.\s*)\s*[\/]\s*|\s*(-)\s*[\/]/U",$buff,$izlaz );[/color]

Is that really your pattern? I only ask because the output
below doesn't come from that.
[color=blue]
> var_dump($izlaz );
>
>
> This outputs:
>
> array(3) {
> [0]=>
> array(4) {[/color]

No it doesn't. There are only three matches to the full
pattern in your subject string.
[color=blue]
> [0]=>
> string(7) "0,x,R /"
> [1]=>
> string(8) " 1,y,R /"
> [2]=>
> string(4) " - /"
> [3]=>
> string(8) " 1,y,R"[/color]

The last string here doesn't match because there must be a
slash at the end. You should allow the last slash to be
optional in your pattern. Consider:

`(\d\s*,\s*[a-z]\s*,\s*.|-)\s*/?`

[ ... ]
[color=blue]
> Problem is in the 3rd row: why is it there? I would expect '-' sign in
> 2nd row...[/color]

Your subject string is '0,x,R / 1,y,R / - / 1,y,R'. The
hyphen is in the third segment, not the second. Move the
hyphen to the second segment and watch what happens.

--
Jock