fn to highlight string differences?

sneakyimp

I was wondering if anybody could recommend a function that would highlight the differences in two strings. This may sound simple but is more complicated than it sounds. Ideally, it would just highlight the part that is different -- kind of like the 'diff' file command but for individual chars rather than lines. for instance:

$str1 = '111111111a1111111111b111111111';
$str2 = '11111111111111c11111111111111';

In that case, the 'a' and 'b' in $str would be highlighted and perhaps the 'c' in line 2?

Anyways, please forgive my addled thinking about this. There's an algorithm the name of which escapes me -- something like 'least differences something' which I'm thinking of. I'd like to adapt it to strings.

dagon

array_diff comes to mind

SunsetKnight

I'm not sure if this is exactly what your looking for, but I threw it together real quick, it's a start I guess:

<?php

$str1 = '111111111a1111111111b111111111';
$str2 = '11111111111111c111111111111111'; 

function highlight_different($str1, $str2)
{
	//Load an array with all values of $str1 as indivudal values
	$split_str1 = str_split($str1);
	//Load an array with all values of $str2 as indivudal values
	$split_str2 = str_split($str2);


//This prints out str1 with red values in HTML if they are different from str2 values
for($i = 0; $i < strlen($str1); $i = $i +1)
{
/*
strcasecmp returns = 0 if they are equal,
less then 0 if str1 is less then str2,
greater then 0 if str1 is greater then str2
*/
$compare = strcasecmp($split_str1[$i],$split_str2[$i]);
	if ($compare == 0)
		echo $split_str1[$i];
	else
		//I'm not sure how your using this, but I used it to display it in html
		echo "<font color='red'>" . $split_str1[$i] . "</font>";

}	


}
//Call the function on the two strings
highlight_different($str1,$str2);

?>

This would output the following html code:

111111111<font color='red'>a</font>1111<font color='red'>1</font>11111<font color='red'>b</font>111111111

which would display as:

111111111a1111111111b111111111

Let me know if this is something you were looking for, I'd be more then happy to further help!

Cheers!

sneakyimp

I appreciate the effort. Yours is great in that it takes length into account. But what if the strings are two different lengths like this?

$str1 = 'page.php?arg1=one&arg2=two&arg3=three';
$str = 'page.php?arg1=one&argx=X&arg3=three';

Despite the fact that arg3 is the same in both strings, it will be flagged as different.

I tried another approach which does very little too:

<?php
$str1 = '123456789a1234567890b123456789';
$str2 = '12345678901234c234567890123456';
//Call the function on the two strings
highlight_different($str1,$str2);

function highlight_different($str1, $str2) {
    //Load an array with all values of $str1 as indivudal values
    $s1 = str_split($str1);
    //Load an array with all values of $str2 as indivudal values
    $s2 = str_split($str2);

$only1 = array_diff($s1, $s2);
$only2 = array_diff($s2, $s1);

$len1 = count($s1);
echo "string 1 ($len1): ";
for($i=0; $i<$len1; $i++) {
    $c = $s1[$i];
    if (in_array($c, $only1)) {
            // make it blue
            echo '<span style="color:#0000ff;">' . $c . '</span>';
    } elseif (in_array($c, $only2)) {
            // make it orange
            echo '<span style="color:#ffdd00;">' . $c . '</span>';
    } else {
            echo $c;
    }
}
echo '<br>';

$len2 = count($s2);
echo "string 2 ($len2): ";
for($i=0; $i<$len2; $i++) {
    $c = $s2[$i];
    if (in_array($c, $only2)) {
            // make it blue
            echo '<span style="color:#0000ff;">' . $c . '</span>';
    } elseif (in_array($c, $only1)) {
            // make it orange
            echo '<span style="color:#ffdd00;">' . $c . '</span>';
    } else {
            echo $c;
    }
}
echo '<br>';

}

?>

Just try it on this to watch it fail:

$str1 = 'http://www.phpbuilder.com/board/showthread.php?p=10909748#post10909748';
$str2 = 'http://www.phpbuilder.com/board/showthread.php?p=10909874#post10909874';

This is what I mean, the diff algorithm is pretty tricky.

I looked on wikipedia and found the 'Longest Common Subsequence Problem'

I think that's what I'm after.

laserlight

sneakyimp wrote:
I looked on wikipedia and found the 'Longest Common Subsequence Problem'

I think that's what I'm after.

Yes, I think you have found it. I do not think that a function already exists in PHP for this, so implement the dynamic programming solution.

SunsetKnight

sneakyimp;10909846 wrote:
I appreciate the effort. Yours is great in that it takes length into account. But what if the strings are two different lengths like this?
$str1 = 'page.php?arg1=one&arg2=two&arg3=three';
$str = 'page.php?arg1=one&argx=X&arg3=three';

So just to see if I'm on track, are these the results your looking for with the above input:

page.php?arg1=one&arg2=two&arg3=three
page.php?arg1=one&argx=X&arg3=three