计算结果集中一个字段中包含的唯一词总数

Question

how can I count single words appearing in a field of a resultset?

example

id| myfield
1 | pear apple peach watermelon
2 | lime grapes pear watermelon

I want to get 6 because there are 6 unique words

I don't need a fast query, it is just a statistic calculation that will be executed rarely

thanks!

Answer 1

You can split the results on a space and then add them to an array, such as

foreach($results as $result)
{
    $words = explode(" ", $result['myfield']);

    $array = array();
    foreach($words as $word)
       $array[$word] = true;

}   
    echo count($array);

Probably a better way, but this is quick and dirty

Answer 2

function uniqueWords ($result, $field) {
    $words = array();
    while ($row = mysql_fetch_assoc($result)) { /* or something else if you use PDO/mysqli/some ORM */
        $tmpWords = preg_split('/[\s,]+/', $row[$field]);
        foreach ($tmpWords as $tmpWord) {
            if (!in_array($tmpWord, $words))
                $words[] = $tmpWord;
        }
    }

    return count($words);
}

Answer 3

I cannot think of a pure SQL solution - MySQL dosen't really like to split a single row nto many.

A PHP version is easy though:

$sql="CREATE TEMPORARY TABLE words (word VARCHAR(100) PRIMARY KEY)";
//run this according to your DB access framework

$sql="SELECT myfield FROM mytable";
//run this according to your DB access framework
while (true) {
  //fetch a row into $row according to your DB access framework
  if (!$row) break;
  $words=preg_split('/[\s,]+/',$row['myfield']);
  foreach ($words as $word) {
    //Escape $word according to your DB access framework or use a prepared query
    $sql="INSERT IGNORE INTO words VALUES('$word')";
    //run this according to your DB access framework
  }
}

$sql="SELECT count(*) AS wordcount FROM words";
//run this according to your DB access framework, the result is what you want


$sql="DROP TABLE words";
//run this according to your DB access framework