Which is the Fastest method to check Rows existence while fetching? - php

Question

From the following Queries, Which one is the most optimal and fastest to use?
[COUNT(id)]
$SQL = "SELECT name, COUNT(id) as Count FROM names WHERE name = :name";
$row = $stmt->fetch();
if ($data['count'] > 0) {
while ($row) {
$name = $row['name'];
}
} else {
return;
}
OR [rowCount()]
$SQL = "SELECT name FROM names WHERE name = :name";
if ($stmt->rowCount() > 0) {
while ($row = $stmt->fetch()) {
$name = $row['name'];
}
} else {
return;
}
OR [EXISTS]
$SQLEX = "SELECT EXISTS (SELECT name FROM names WHERE name = :name LIMIT 1)";
if ($stmt->fetchColumn == 1) {
$SQL = "SELECT name FROM names WHERE name = :name";
while (row = $stmt->fetch()){
$name = $row['name'];
}
} else {
return;
}
OR [RAW]
$SQL = "SELECT name FROM names WHERE name = :name";
$row = $stmt->fetch();
if ($row) {
while($row) {
$name = $row['name'];
}
} else {
return;
}
Also i wanted to know, Why does using $stmt->fetch() with $stmt->rowCount() allows me to fetch data, But using it with $stmt->fetchColumn doesn't?

Answer 1

First, if you have an index on names(name), then all should be quite comparable in speed.
Second, it is always worth trying such performance tests on your own system.
Third, if names are declared as unique (or primary key) in the names table, then all should be quite fast.
In general, though, the fastest way to determine if a row is available is:
SELECT EXISTS (SELECT name FROM names WHERE name = :name)
The LIMIT 1 in the subquery is unnecessary -- EXISTS stops at the first row (whether the database uses an index or a table scan).
In general, the first method using an aggregation is the worst solution. Without an index, it is going to result in a full table scan that reads the entire table. The second might or might not read the entire table, depending on whether the database starts returning matching rows as they are available. It also has the downside of returning more data.

Answer 2

Ok, it seems this question needs more than one answer...
f you need to check the existence only,
if there is an unique index for the field, all methods are equal, but some of them just make no sense.
if there is no unique index, then go for EXISTS
If you need to fetch the actual data and see if there was anything returned, then just select your data and fetch it:
if only one column from a single row is expected, then use fetchColumn()
if only one row is expected, then use fetch()
if multiple rows are expected, then use fetchAll()
and then use the resulting value to see whether your query returned any data.
So if you finally made your mind as to what you're asking about, here is the most optimal code for you:
$SQL = "SELECT name FROM names WHERE name = :name";
$data = $stmt->fetchAll(PDO::FETCH_COLUMN);
if (!$data) {
return;
}
foreach ($data as $name) ...
And there is nothing wrong with fetchColumn() other than your idea to use it.