从多个阵列中删除常见项目

$array1 = array("a","b");

$array2 = array("b","c");

I want the following array from above arrays containing items which are not common in both arrays.

$output = array("a","c");

I have tried folloing

$output = array_diff($array1,$array2);

How can do it. Thanks.

Use:

$array1 = array("a","b");

$array2 = array("b","c");

$output = array("a","c");

$output = array_merge(array_diff($array1,$array2),array_diff($array2,$array1));

print_r($output);

Just make subtraction of union and intersection:

$array1 = array("a","b");    
$array2 = array("b","c");

$output = array_diff(array_merge($array1, $array2), array_intersect($array1, $array2));

So I read earlier today about how array_merge was considered slow. I also read the same about array_diff, specifically here. This got me to thinking that using both at once could be particularly expensive, so I put the following experiment together:

I expanded upon merlyn's function to come up with this:

function multiArrayUnique($array1, $array2)
{
    //leaves array1 intact
    $arrayFrom1=$array1;
    $arrayAgainst = array_flip($array2);

    foreach ($arrayFrom1 as $key => $value) {
        if(isset($arrayAgainst[$value])) {
            unset($arrayFrom1[$key]);
        }
    }
    //arrayFrom1=array("a");

    $arrayFrom2=$array2;//b,c
    $arrayAgainst = array_flip($array1);//a,b
    foreach ($arrayFrom2 as $key => $value) {
        if(isset($arrayAgainst[$value])) {
            unset($arrayFrom2[$key]);
        }
    }
    //arrayFrom2=array("c");

    foreach ($arrayFrom2 as $item) {
        array_push($arrayFrom1, $item);
    }

    return $arrayFrom1;
}

Using the OP's example, this function also returns array("a","c").

Now, that is a monstrous amount of code, and it it difficult to read... but, here is the efficiency it buys on a large data set:

$a1 = range(0,25000); 
$a2 = range(15000,50000);

$start=microtime(true);
$output = array_diff(array_merge($a1, $a2), array_intersect($a1, $a2));
$end=microtime(true);
echo $end-$start."<hr>";

7.5844340324402

$start=microtime(true);
$output = array_merge(array_diff($a1,$a2),array_diff($a2,$a1));
$end=microtime(true);
echo $end-$start."<hr>";

6.551374912262

$start=microtime(true);
$result=multiArrayUnique($a1,$a2);
$end=microtime(true);
echo $end-$start."<hr>";

0.21001195907593

So while the function's code is considerably more complex, and may have other limitations, it's implementation is still a one-liner, and the processing time is 14x faster.