PHP select multiple Array tables - php

Question

Is it possible to SELECT with multiple Array tables. I know it sounds confusing but here is what I have done :
First of all, I've created a form, that has two checkboxes option as follows :
<form action="something.php" method="post">
<input type="checkbox" name="store[]" value="M1">
<input type="checkbox" name="store[]" value="M2">
<input type="submit" value="Go">
</form>
Now after submitting the form, I can view which store selected by doing foreach loop:
$allstore = $_POST['store'];
foreach ($allstore as $store=>$value) {
echo $value;
}
Everything till now works as needed !
However those two values in checkboxes are considered to be table names ! Now how can I find a way to let PHP select either one or two tables based on user selection $value ?
$query = "SELECT * from **{$allstore[0]},{$allstore[1]}** WHERE .....";
As you can see {$allstore[0]},{$allstore[1]} should be created based under foreach loop. I can't seem to find a way of doing it! Can I insert a function to do it for me?
Like this : $query = "SELECT * from ((( Function Here ))) WHERE .....";
If you have a different way of doing it, Please share it.
Edit :
M1 Table
id |item_no |qty |price
1 x1 10 20
2 x2 5 22
3 x3 3 5
M2 Table
id |item_no |qty |price
1 x1 11 20
2 x9 5 30
3 x10 6 26
The output table should be
item_no | price | M1 | M2
x1 20 10 11
x2 22 5 N/A
x3 5 3 N/A
x9 30 N/A 5
x10 26 N/A 6
That's what I am aiming for. I hope it can be solved !

Answer 1

here's the structure for 2 tables sqlfiddle
I think you can add more tables from here.
SELECT T1.item_no,
COALESCE(M1.price,M2.price) as price,
M1.qty as M1,
M2.qty as M2
FROM
(SELECT item_no FROM M1
UNION
SELECT item_no FROM M2
)T1
LEFT JOIN M1 ON T1.item_no = M1.item_no
LEFT JOIN M2 ON T1.item_no = M2.item_no
UPDATED: I am not too familiar with PHP but I looked up some syntax and was able to dynamically generate SQL based on array of either ["M1","M2"] or ["M1"] or ["M2"]
DynamicPHPtobuildSQL
<?php
//Enter your code here, enjoy!
$allstore = ["M2"];
$item = 0;
$sqlpart1 = "";
$sqlpart2 = "";
$sqlpart3 = "";
$sqlpart4 = "";
foreach ($allstore as $store=>$value) {
$item += 1;
if ($item > 1){
$sqlpart1 .= ",";
$sqlpart2 .= ",";
$sqlpart3 .= " UNION ";
}
$sqlpart1 .= $value . ".price ";
$sqlpart2 .= $value . ".qty as " . $value . " ";
$sqlpart3 .= "SELECT item_no FROM " . $value . " ";
$sqlpart4 .= "LEFT JOIN " . $value . " ON T1.item_no=" . $value . ".item_no ";
}
$SQL = "SELECT T1.item_no,COALESCE(" . $sqlpart1 . ") as price," . $sqlpart2;
$SQL .= "FROM (" . $sqlpart3 . ")T1 " . $sqlpart4;
echo $SQL;
?>

Answer 2

Be careful to avoid the risk of SQL injection: compare the posted values against a closed list of existing store table names and reject any other value.
Note also that not only the FROM clause is influenced by the user's choices, but also the SELECT clause. So you have two dynamic parts in your SQL statement.
You could use this code which makes use of array_intersect, implode and array_map:
$selected_stores = $_POST['store'];
// Protect against SQL-injection by only accepting known tables:
$all_stores = array("M1", "M2", "M3");
$selected_stores = array_intersect($selected_stores, $all_stores);
// Build dynamic part of the FROM clause
$from = implode("
UNION
", array_map(function ($store) {
return "SELECT '$store' as store, item_no, price, qty FROM $store";
}, $selected_stores));
// Build dynamic part of the SELECT clause
$cols = implode(",
", array_map(function ($store) {
return "CASE store WHEN '$store' THEN qty END AS $store";
}, $selected_stores));
$sql = "
SELECT item_no,
MAX(price) as price,
$cols
FROM ( $from ) data
GROUP BY item_no
";
The SQL generated looks like this:
SELECT item_no,
MAX(price) as price,
CASE store WHEN 'M1' THEN qty END AS M1,
CASE store WHEN 'M2' THEN qty END AS M2
FROM ( SELECT 'M1' as store, item_no, price, qty FROM M1
UNION
SELECT 'M2' as store, item_no, price, qty FROM M2 ) data
GROUP BY item_no
See also this SQL fiddle.
As a side comment: I would advise to combine all store tables into one table, which would have an additional column indicating the store. This is more in line with normalised database design and will give more advantages than disadvantages in terms of searching, sorting, performance, and simplicity.