使用colums值作为参数计算所有字符串的出现次数-MySQL

Question

I'm gathering some data information for a project, I would like to see what words repeat more in the database using the column values as search parameters, like:

table_product
  id  | name         | description
  1   | blue shirt   | cool shirt 
  2   | yellow pants | pretty nice pants
  3   | red shirt    | cool red shirt

The expected result is something like:

field   |  count
shirt   |  4
cool    |  2
pants   |  2
...     |  ...

Is there a way to achieve this using only querys? Thanks.

EDIT: I can't pre-set a word list, the goal is to cross all the string and count the occurrences.

Answer 1

to do this type of result you will have to union each word as a separate query. like so..

SELECT 
'shirt' AS field,
SUM(
    CASE WHEN LOCATE('shirt', `name`) > 1 AND LOCATE('shirt', `description`) THEN 2 
         WHEN LOCATE('shirt', `name`) > 1 OR LOCATE('shirt', `description`) THEN 1
         ELSE 0 
    END ) AS 'Count'
FROM table_product

UNION

SELECT 
'cool' AS field,
SUM(
    CASE WHEN LOCATE('cool', `name`) > 1 AND LOCATE('cool', `description`) THEN 2 
         WHEN LOCATE('cool', `name`) > 1 OR LOCATE('cool', `description`) THEN 1
         ELSE 0 
    END ) AS 'Count'
FROM table_product

UNION

SELECT 
'pants' AS field,
SUM(
    CASE WHEN LOCATE('pants', `name`) > 1 AND LOCATE('pants', `description`) THEN 2 
         WHEN LOCATE('pants', `name`) > 1 OR LOCATE('pants', `description`) THEN 1
         ELSE 0 
    END ) AS 'Count'
FROM table_product

DEMO

RECOMMENDATION:

I would recommend you just make a new column for each count you want.. so in other words just use a pivoted result.. as it'll be much easier to handle and if you are trying to get the result in another programming language the key will be the name aka 'shirt', 'cool'... and the value will be the count.. demonstration

SELECT 
    SUM(
    CASE WHEN LOCATE('shirt', `name`) > 1 AND LOCATE('shirt', `description`) THEN 2 
         WHEN LOCATE('shirt', `name`) > 1 OR LOCATE('shirt', `description`) THEN 1
         ELSE 0 
    END ) AS 'shirt',

    SUM(
    CASE WHEN LOCATE('cool', `name`) > 1 AND LOCATE('cool', `description`) THEN 2 
         WHEN LOCATE('cool', `name`) > 1 OR LOCATE('cool', `description`) THEN 1
         ELSE 0 
    END ) AS 'cool',

    SUM(
    CASE WHEN LOCATE('pants', `name`) > 1 AND LOCATE('pants', `description`) THEN 2 
         WHEN LOCATE('pants', `name`) > 1 OR LOCATE('pants', `description`) THEN 1
         ELSE 0 
    END ) AS 'pants'
FROM table_product;

Answer 2

Thanks John for the information and your time.

For whom it may be usefull, I developed a quick script to do this trick in php.

PS: This is by no means the best way you can do it, but it'll work.

//search the strings
$result = mysql_query("SELECT field_a, field_b FROM table");

$arrayMaster = array();
while ($row = mysql_fetch_array($result)) {
    //explode the spaces to get each word
    $field_a = explode(' ', $row['field_a']);
    foreach ($field_a as $string) {
        $string = strtolower($string);
        //remove words with less then 4 charac
        if(strlen($string) < 4){
            continue;
        }
        //check if there is a position on the array with that string, if it exists ++
        if( isset($arrayMaster[$string]) ){         
            $arrayMaster[$string]++;
        } else {
            $arrayMaster[$string] = 1;
        }
    }

    $field_b = explode(' ', $row['field_b']);
    foreach ($field_b as $string) {
        $string = strtolower($string);
        if(strlen($string) < 4){
            continue;
        }
        if( isset($arrayMaster[$string]) ){         
            $arrayMaster[$string]++;
        } else {
            $arrayMaster[$string] = 1;
        }
    }
}
//sort the array in reverse order
arsort($arrayMaster);
//print the words and the amount that they show up
print_r($arrayMaster);