我会为你​​在文件中读取过程这一点。

I would process this as you're reading in the file.

// how many total columns
$total = count( $row ); 

// get the halfway point (and round up if a decimal)
$middle = ceil( $total/2 );

// Form a new row using the first (0), last ($total-1) and middle ($middle)
$new_row = array( $row[0], $row[ $middle ], $row[ $total-1 ] );

您code中嵌入：

function csv_to_array($filename='', $delimiter=',') {
   if(!file_exists($filename) || !is_readable($filename)){
      return FALSE;
    }

$header = NULL;
$data = array();
if (($handle = fopen($filename, 'r')) !== FALSE) {
    while (($row = fgetcsv($handle, 1000, $delimiter)) !== FALSE) {

        $total = count( $row ); 
        $middle = ceil( $total/2 );
        $new_row = array( $row[0], $row[ $middle ], $row[ $total-1 ] );

        if(!$header) {
            $header = $new_row;
        }else{
            $data[] = array_combine($header, $new_row);
        }
    }

    fclose($handle);
}

  return $data;
}

您可以减轻一些处理能力，通过在整个文件假定每行将具有相同的列数。如果是这样，你只需要计算一次，就像这样：

You could alleviate some processing power, by assuming each row in the entire file will have the same column count. If so, you need only count once, like so:

function csv_to_array($filename='', $delimiter=',') {
   if(!file_exists($filename) || !is_readable($filename)){
      return FALSE;
    }

$header = NULL;
$data = array();
if (($handle = fopen($filename, 'r')) !== FALSE) {
    while (($row = fgetcsv($handle, 1000, $delimiter)) !== FALSE) {

        if ( !$total ){ // Verify if we have the total yet, and if not:
            $total = count( $row ); 
            $middle = ceil( $total/2 );
        }
        $new_row = array( $row[0], $row[ $middle ], $row[ $total-1 ] );

        if(!$header) {
            $header = $new_row;
        }else{
            $data[] = array_combine($header, $new_row);
        }
    }

    fclose($handle);
}

  return $data;
}