[SOLVED] Regex condition not working

[knox203]

Hello all, I'm trying to separate names in a list. I'm having a hard time telling the code to do the same thing if the name format is either "John M. Doe" or "John Middle Doe". Here's the code:

 

<?php
$odbc = odbc_connect('sms','','') or die(odbc_errormsg());
$strsql= 'SELECT * FROM cardmstr.DBF';
$query = odbc_exec($odbc, $strsql) or die (odbc_errormsg());
while($row = odbc_fetch_array($query))
{
if ( ereg("^([A-Za-z]+ ?[A-Za-z]?.?) ([A-Za-z]+)", $row['item_desc'], $new) ) {
	echo '<b>'.$new[1]." ".$new[2].'</b><br />'."\n";
   }
}
odbc_close($odbc);
?>

Results returned with this code looks like:

"John M. Doe"

"John Middle"

 

I tried

<?php ereg("^([A-Za-z]+ ?([A-Za-z]|[A-Za-z]+)?.?) ([A-Za-z]+)", $row['item_desc'], $new) ?>

but fails all together.

 

Any sort of help would be great, thanks!

 

- Adam

[effigy]

Are the names together in one field and you're trying to separate the names, or are you trying to break a single name into pieces?

[knox203]

Hey effigy, I'm trying to break the names into pieces.

[effigy]

How about this?

 

$pieces = preg_split('/\S+/', $name);

[nrg_alpha]

$pieces = preg_split('/\S+/', $name);

 

Effigy.. I'm rather curious.. As I examine your pattern, I can't help but wonder.. Since you are matching any non-space character one or more consecutive times.. is this pretty much a 'beefier' version of:

'/\b\w+\b/' ? (since \w is basically a-zA-Z0-9_ (locale issues aside), I suppose your pattern anticipates a broader range of characters and let's the spaces act as 'word bounderies' or sorts)

 

[knox203]

You know, I'm not too sure, I'm new to regex. I need to group "John M." and "John Middle" as one group (first name) and "Doe" as last name. The names are in all sorts of different formats, so I'll include the rest of the code to hopefully give you a feel for what I'm working with.

<?php
$odbc = odbc_connect('sms','','') or die(odbc_errormsg());
$strsql= 'SELECT * FROM cardmstr.DBF';
$query = odbc_exec($odbc, $strsql) or die (odbc_errormsg());
while($row = odbc_fetch_array($query))
{
  // LastName, FirstName
if ( ereg("^([A-Za-z]+)(,) ([A-Za-z]+) (J(unio)?r(.)?|S(enio)?r(.)?)?", $row['item_desc'], $new) ) {
	echo '<b>'.$new[3]." ".$new[1]." ".$new[4].'</b><br />'."\n";
}
 // FirstName (2nd FirstName) LastName - Emp#
elseif ( ereg("^([A-Za-z]+( *)?([A-Za-z]+)?) ([A-Za-z]+)( *)?- (C?[0-9]+)", $row['item_desc'], $new) ) {
	echo '<b>'.$new[1]." ".$new[4].'</b><br />'."\n";
}
 // Emp# - FirstName LastName
elseif ( ereg("^(C?[0-9]+) (-)? ?([A-Za-z.?]+) ([A-Za-z]+)", $row['item_desc'], $new) ) {
	echo '<b>'.$new[3]." ".$new[4].'</b><br />'."\n";
}
 // "Truck #0-9" - Vehicle Type
elseif ( ereg("^Truck (#[1-9]+) - ([A-Za-z]+)", $row['item_desc'], $new) ) {
	echo '<b>'.$row['item_desc'].'</b><br />'."\n";
}
 // FirstName LastName
elseif ( ereg("^([A-Za-z]+ ?[A-Za-z]?.?) ([A-Za-z]+)", $row['item_desc'], $new) ) {
	echo '<b>'.$new[1]." ".$new[2].'</b><br />'."\n";
}
elseif ( trim($row['item_desc']) ) {
	echo '<b>'.$row['item_desc'].'</b><br />'."\n";
}
}
odbc_close($odbc);
?>

 

Thanks!

[DarkWater]

I think Effigy's solution is relatively optimal, if you're basically splitting by spaces anyway (that's what you're really doing).  Also, don't use ereg().  Use preg_* functions because ereg() and the POSIX extended regular expressions will be removed from PHP 6 (only to be available by means of PECL).

[effigy]

$pieces = preg_split('/\S+/', $name);

Effigy.. I'm rather curious.. As I examine your pattern, I can't help but wonder.. Since you are matching any non-space character one or more consecutive times.. is this pretty much a 'beefier' version of:

'/\b\w+\b/' ? (since \w is basically a-zA-Z0-9_ (locale issues aside), I suppose your pattern anticipates a broader range of characters and let's the spaces act as 'word bounderies' or sorts)

 

Essentially, yes. In many cases preg_split is better because you're being specific about what you do not want, rather than attempting to make a pattern that's versatile enough to match what you do want.

 

The names are in all sorts of different formats

 

What problems are you having with the given code? Do you need each variation to look for a middle name?

 

 

[knox203]

Thanks for your advice everyone. I'll see what I can come up with. (PREG looks like it's the way to go)