加入1个表中具有单个ID /行但在另一个表中链接到多个行（相同ID）的SQL查询？

Question

First sorry for the long question title.

My question/situation is as such.

1.) I have 2 tables in mysql

2.) In first table, each listing has a unique id(each listing is in 1 row)

3.) In the second table it has the name/tags for images linked to the listing id,from the first table

4.) Each listing can have multiple images(multiple row in the second table).

What i am trying to do is to pull all the listings from table 1 and then use the listing.id from table one to pull all the rows of images from table 2 that are linked to the listing.id. I am confused at the moment because there are multiple rows that has the same listing.id from table 2. ANd i tried query to display* but it only echo the last image(row) from table 2.

It doesnt seem to work when i join the 2 tables. And i am not sure if i query it twice then push array together.

Thanks for your time

 $result=mysqli_query($con,"SELECT * FROM Listing JOIN listingpic ON 
    (Listing.id = listingpic.listingid)

 WHERE date >= curdate() - INTERVAL DAYOFWEEK(curdate())+300 DAY GROUP BY Listing.id ORDER BY Listing.id DESC") or die( mysqli_error($con));

while ($row = mysqli_fetch_assoc($result))
        {
  $output[] = $row;
        }
        if (!empty($output)){

        echo json_encode( $output );}
        else{

            echo json_encode( [] );
        }

Answer 1

You only get one result per id, because you used GROUP BY Listing.id . If you do not group them, you get one result row for each table 2 row, including the Listing id each time.

If you wish to retrieve the data in one query in the form "one id: multiple data", you can use GROUP_CONCAT for example and then explode() the retrieved string result.

Otherwise get all ids from table 1 and then iterate over them in PHP and do one additional query per ID

Answer 2

Pro tip: Don't use the viciously confusing MySQL extension to GROUP BY . Read this: http://dev.mysql.com/doc/refman/5.6/en/group-by-handling.html

Pro tip: Don't use SELECT * , especially when you're joining tables. Instead, enumerate the columns you want in your result set.

Inherent to SQL is the idea that resultsets, like tables, are rectangular. They have rows and columns. Each row usually represents some real world item -- an "entity" -- and each column represents some attribute of that entity.

The result set you describe will, inherently, repeat information from your first table so it can show the info from the second table row by row.

What you want for a query is this, I think.

SELECT Listing.id, listing.date, 
       listingpic.id, listingpic.url, listingpic.caption
  FROM Listing
  JOIN listingpic ON  Listing.id = listingpic.listingid
  WHERE date >= curdate() - INTERVAL DAYOFWEEK(curdate())+300 DAY 
  ORDER BY Listing.id DESC, listingpic.id

This will give you one row per image.

If you're running out of memory it's because your result set is massive. You may want to limit it somehow, either using a first and last publication date:

  WHERE date >= curdate() - INTERVAL DAYOFWEEK(curdate())+300 DAY 
    AND date <  curdate() - INTERVAL DAYOFWEEK(curdate())+293 DAY

or with a LIMIT clause.

  ORDER BY Listing.id DESC, listingpic.id
  LIMIT 100