如何从union all join查询中删除重复项

Question

I have a table which holds partnumbers and prices of those partnumbers. This is my main table with about 100000 entries. I also have 3 tables with possible suppliers. Not all partnumbers of the main table are in all 3 supplier tables, and the 3 supplier tables also have a different content.

What I want is to query the 3 tables for the partnumbers of my main table. If the partnumber is found in more then 1 table it should only return the lowest price one. If it is found in just 1 table, return the price of that part.

What I am doing now is the following query:

SELECT `table1.partnumber`,
    `table1`.`price`,
    `maintable`.`price`
FROM `table1`
    INNER JOIN `maintable`
        ON `table1`.`partnumber` = `maintable`.`partnumber`
UNION ALL
SELECT `table2`.`partnumber`,
    `table2`.`price`,
    `maintable`.`price`
FROM `table2`
    INNER JOIN `maintable`
        ON `table2`.`partnumber` = `maintable`.`partnumber`
UNION ALL
SELECT `table3.partnumber`,
    `table3`.`price`,
    `maintable`.`price`
FROM `table3`
    INNER JOIN `maintable`
        ON `table3`.`partnumber` = `maintable`.`partnumber`

This gives me all the data I need, and after that I check for the lowest price and remove the higher price partnumbers with foreach php code. This is very slow on large sets of data.

I am sure there is a way to only get the lowest priced partnumber if that partnumber exists in more than 1 table, and just return the price if it only exists in 1 table. I just don't have a clue how to query the mysql database for it. I have tried using combination of joins, unions, min and distinct, but to no avail.

Can someone help me out?

The answer from eggyal below send me on my way. I now have the following query which works for me:

SELECT `partnumber`,MIN(`price`) as `price`
FROM (
    SELECT `table1`.`partnumber`,
        `table1`.`price`,
        `maintable`.`price`
    FROM `table1`
        INNER JOIN `maintable`
        ON `table1`.`partnumber` = `maintable`.`partnumber`
    UNION ALL
    SELECT `table2`.`partnumber`,
        `table2`.`price`,
        `maintable`.`price`
    FROM `table2`
        INNER JOIN `maintable`
        ON `table2`.`partnumber` = `maintable`.`partnumber`
    UNION ALL
    SELECT `table3.partnumber`,
        `table3`.`price`,
        `maintable`.`price`
    FROM `table3`
        INNER JOIN `maintable`
        ON `table3`.`partnumber` = `maintable`.`partnumber`
) AS `part` GROUP BY `partnumber`

Perhaps there is a better way, in which case i'm all ears. But for know I am happy with the above.

Answer 1

SELECT partnumber, MIN(price) FROM (
  SELECT 1 AS tn, partnumber, table1.price
  FROM   table1 JOIN maintable USING (partnumber)
UNION ALL
  SELECT 2 AS tn, partnumber, table2.price
  FROM   table2 JOIN maintable USING (partnumber)
UNION ALL
  SELECT 3 AS tn, partnumber, table3.price
  FROM   table3 JOIN maintable USING (partnumber)
) t
GROUP BY partnumber
HAVING COUNT(DISTINCT tn) >= 2