How to get value array from matching two arrays?

You can use array_filter() function

$output = array_filter($text, function($e) use($inline){
    foreach ($inline as $search) {
        if (strpos($e, $search) > -1) return true;
    }
    return false;
});

print_r($output);

I have done with one foreach and i have used preg_quote and preg_grep for this :-

$finaloutput = array();
$text = array(                                                               
          "I had locked the door before I left my home",                 
          "I was late. The train had left",                            
          "She had already found the key, before I broke the door",  
          "n..."                                                         
     );
$inline =  array("locked","found");
foreach($inline as $find){
    $input = preg_quote($find, '~');
    $result = preg_grep('~' . $input . '~', $text); 
    $finaloutput[] = $result;
}
$finalArray = call_user_func_array('array_merge', $finaloutput);
echo "<pre>"; print_r($finalArray);

Even you serached with Locked or Loc or lo it will fetch your data. Hope it helps!

You can use strpos and have nested foreach

$text = array(                                                               
              "I had locked the door before I left my home",                 
              "I was late. The train had left",                            
              "She had already found the key, before I broke the door",  
              "n..."                                                         
         );

$inline =  array("locked","found");
$output = array();
//Loop our sentences
foreach($text as $key => $sentence){
    //Check the word
    foreach($inline as $find){
        //Check if the word is in sentence
        if(strpos($sentence,$find) !== false){
            $output[$key] = $sentence;
        }
    }
}
print_r($output);

RESULT

Array
(
    [0] => I had locked the door before I left my home
    [2] => She had already found the key, before I broke the door
)

Not an expert, but I like to use preg_match myself. Not sure if there is a performance hit for using it instead of strpos. But being able to use the OR operator between patterns makes it easy to just implode the patterns into a string and search all the patterns at once instead of having to make another loop.

$texts = [
    "I had locked the door before I left my home",
    "I was late. The train had left",
    "She had already found the key, before I broke the door",
    "n..."
];
$patterns = [ "locked", "found" ];
$output = [];

foreach( $texts as $text )
{
    if( preg_match( '#' . implode( '|', $patterns ) . '#', $text )) {
        $output[] = $text;
    }
}

print_r( $output );

Here is a modification as per @mickmackusa suggestion which would be needed in situations where you do not want any of the characters in the patterns to be misinterpreted as regex control characters (ex: a period acting as a wildcard ) and if you want to ensure that the pattern words you are looking for are single words (not part of another word).

foreach( $texts as $key => $text )
{
    if( preg_match( '#\b\Q' . implode( '\E\b|\b\Q', $patterns ) . '\E\b#', $text )) {
        $output[$key] = $text;
    }
}

preg_grep() is perfectly suited to filter your array elements with one function call. The only preparation needed is to add word boundaries and make the strings literal with (\Q..\E).

In truth, your sample $inline data doesn't require \Q and \E, but this is just as a precaution. In the same vein, \b are not required for your sample input data but this is best practice for future-proofing your script.

Code: (Demo)

$text = array(                                                               
    "I had locked the door before I left my home",                 
    "I was late. The train had left",                            
    "She had already found the key, before I broke the door",  
    "n..."                                                         
);

$inline =  array("locked","found");
$pattern='/\b\Q'.implode('\E\b|\b\Q',$inline).'\E\b/';  // renders as: /\b\Qlocked\E\b|\b\Qfound\E\b/
var_export(preg_grep($pattern,$text));

Output:

array (
  0 => 'I had locked the door before I left my home',
  2 => 'She had already found the key, before I broke the door',
)

For more reading on \Q and \E see the Special Characters section of: https://www.regular-expressions.info/characters.html

For more reading on \b see: https://www.regular-expressions.info/wordboundaries.html

For more reading on | (Alternation) see: https://www.regular-expressions.info/alternation.html

p.s. if you need case-insensitive matching use the i pattern modifier:

$pattern='/\b\Q'.implode('\E\b|\b\Q',$inline).'\E\b/i';