[RESOLVED] Searching array values - working just needs a tweak !

[Mad_T]

Hi
I'm using the following to filter results from specific lines of a text file.

But I have one slight issue.. The example below shows the issue :

<?php

     $line = "1=\"Device\",3=\"User\",7=\"ID123456\",27=\"Node\"";
     $q = explode(",",str_replace('"','',$line));
     $p = (search_array($q, "7=")); echo "ID : ".$p;

     
     
function stipos($haystack, $needle){
    return strpos($haystack, stristr( $haystack, $needle ));
}

function search_array ( $array, $term )
    {
       foreach ( $array as $key => $value )
            if ( stipos( $value, $term ) !== false )
             $val = str_replace('"',"",preg_replace("/[a-zA-Z0-9]=/","",$array[$key]));
               if (isset($val)) return $val;
          return false;
    } 
  ?>

I'm search for 7= and then expect to return the value associated to that. In this example, I'd expect : ID123456

However I have an additional entry in $line that is 27=\"Node\"
So the result I actually get is 2Node, not ID123456

I only want a match on 7= and NOT 17= or 27= which it's currently doing.
How do I change the to only return the value against 7= and not a mix of 27= & it's value ??

Thanks

[bradgrafelman]

As for your issue, you'd want to make sure the preceding character is either a comma or the beginning-of-string indicator (e.g. '^' in PCREs).

However, a bigger issue might be: Why aren't you using a DBMS (e.g. MySQL) for this? At the very least, why not use SQLite to manage your database?

[Mad_T]

Thanks for the reply.
I get a plain flat file and then read it in, I'm only selecting a few values from it so a DB is a bit of overkill.

I've tried updating this to :

     $p = (search_array($q, ",7=")); echo "ID : ".$p; 

ie: adding a comma before the 7= but this results in no values being returned.

Any ideas ?

[NogDog]

A somewhat different approach:

<?php

$line = "1=\"Device\",3=\"User\",7=\"ID123456\",27=\"Node\"";

$value_7 = null;
if(preg_match('#\b7="([^"]*)"#', $line, $matches)) {
    $value_7 = $matches[1];
}
var_dump($value_7);

[Mad_T]

NogDog thanks.
Please can you explain this ?

[bradgrafelman]

I get a plain flat file and then read it in, I'm only selecting a few values from it so a DB is a bit of overkill.

Well you already have a database - I was simply offering that using a DBMS might be a more efficient way of managing that database.

Regardless, the method I had in mind above is more or less exactly what NogDog came up with, although you could also have utilized the fact that it appears you're being given "key=value" pairs (e.g. the value on the left of the equal sign will never be repeated within a row - it's a unique identifier). Example:

// I'm assuming you actually have a file with many lines like the example string:
$file = [
    '1="Device",3="User",7="ID1",27="Node"',
    '1="Device",3="User",7="ID2",27="Node"',
    '1="Device",3="User",7="ID3",27="Node"',
    '1="Device",3="User",7="ID4",27="Node"'
];

// ... and then a loop to process all of them:
foreach($file as $line) {
    if(preg_match_all('/(?:,|)([^=]+)="([^"]+)"(?:,|)/', $line, $data)) {
        $data = array_combine($data[1], $data[2]);
        
        print "Found row with ID: $data[7]!\n";
    }
} 

And of course the loop could be modified to instead store the data for later use if you didn't want to (or are unable to) process the data immediately.

[NogDog]

NogDog thanks.
Please can you explain this ?

The \b sequence in a PCRE regex is a "word boundary" assertion, so \b7 will match on any "7" preceded by a white-space or any non-word character (I think the underscore is considered a word character in this case, but I'd have to check to be sure). The [^"]* is matching on 0 to n characters that are not a double quote, and wrapping that expression in parentheses makes it a sub-pattern that can be separately selected from the resulting $matches array.

[Mad_T]

Thanks NogDog, that works perfectly for what I need.

bradgrafelman - thanks I can see a use for this in another script I need.

Cheers