Preg_match Issue

[NZ_Kiwis]

I have a series of names like this

Duck/Donald
Mouse/Minnie

these names are in a longer string of data which I'm using the following Preg_match to find them, it works well for all names however some names are like this

F/Duck/Donald

These ones I can not find.

How can I pick these ones up?

if (preg_match('# [A-Z/]{11}#', $Name, $m)) {
		$Name = str_replace(" ", "", $m[0]);

} elseif (preg_match('# [A-Z]{11} #', $Name, $m)) {
		$Name = str_replace(" ", "",$m[0]);

} elseif (preg_match('([A-Z]+/[A-Z]+)', $Name, $m)) {
		$Name = str_replace(" ", "",$m[0]);

}

[bradgrafelman]

Here's one way of doing it:

$text=" Something like Duck/Donald
and then we've got a Mouse/Minnie or F/Duck/Donald";


preg_match_all
(
	'@(?:(?P<middle_init>[A-Z])/)?(?P<last>[A-z]+)/(?P<first>[A-z-]+)@',
	$text,
	$matches,
	PREG_SET_ORDER
);

foreach($matches as $name)
{
	echo "First name: $name[first]\n";
	echo "Middle Initial: $name[middle_init]\n";
	echo "Last name: $name[last]\n\n";
}

[NZ_Kiwis]

bradgrafelman;10980165 wrote:

Here's one way of doing it:

$text=" Something like Duck/Donald
and then we've got a Mouse/Minnie or F/Duck/Donald";


preg_match_all
(
	'@(?:(?P<middle_init>[A-Z])/)?(?P<last>[A-z]+)/(?P<first>[A-z-]+)@',
	$text,
	$matches,
	PREG_SET_ORDER
);

foreach($matches as $name)
{
	echo "First name: $name[first]\n";
	echo "Middle Initial: $name[middle_init]\n";
	echo "Last name: $name[last]\n\n";
}

I don't want to split part of the names I just to want to the entire name out

For example

I have

ID:7117 DUCK/DONALD CARTOON
ID 8747 F/MOUSE/MINNIE MOUSE

So I want DUCK/DONALD and i get that just not minnie

why and how can i?

[bradgrafelman]

Well in that case, just use the same pattern above but get rid of the named subpatterns and () groupings.

[NZ_Kiwis]

I get super lost in the pattern, can you explain that to me?

[bradgrafelman]

NZ_Kiwis;10980192 wrote:

I get super lost in the pattern

S'ok - I offered pretty much the beefiest version of a pattern to do what I thought you were trying to do. It's a side effect of staring at complex, 10+ lines, heavily-commented regexp's I've been dealing with lately.

If you're just wanting to search for strings that match the pattern you're talking about, get rid of the sub-groups in my pattern above, e.g.:

@(?:[A-Z])/)?[A-z]+/[A-z-]+@

Again, you haven't fully explained all of the constraints on the different fields (e.g. is the optional first field always a single letter?), so some tweaking may still need to be done.

[NZ_Kiwis]

I guess it's a string up to 11 characters long and may contain one or two /'s

Some might not be 11 long either if the name is short.
All will be upper case

so how about

([A-Z]/[A-Z]/[A-Z]/)

??

[NZ_Kiwis]

I'm still lost here, how about something

one letter / then up to 9 more letters and maybe another /

so F/DUCK/DONALD

There must be a simple way of finding this?

[johanafm]

NZ_Kiwis;10980167 wrote:

ID:7117 DUCK/DONALD CARTOON
ID 8747 F/MOUSE/MINNIE MOUSE

So I want DUCK/DONALD and i get that just not minnie

I understand this as you're after DUCK/DONALD or MINNIE/MOUSE, rather than F/MOUSE.

And if I've understood you correctly, then this

NZ_Kiwis;10980829 wrote:

one letter / then up to 9 more letters and maybe another /

so F/DUCK/DONALD

should rather be expressed as:
"Possibly 1 letter and / then capture up to 9 letters followed by a slash and more letters."
which you could write as

$p = '#([A-Z]/)?\K[A-Z]{1,9}/[A-Z]+#';

which you read out as
"Find 0 or 1 occurences of [A-Z]/ then reset pattern matching to this position in the string (thus disgarding a matched [A-Z]/), then match 1-9 characters A-Z followed by / and 1 ore more letters A-Z.