I was recently trying to do a project*, which caused me to ask this question. Although since then I've found an alternative solution, I am still curious if what I envisioned doing is, in any way, possible.
Essentially, I am wondering if there is anyway to perform a MySQL query on a MySQL query result in php. For example:
$result = mysql_query("SELECT * FROM foo WHERE bar=".$barValue);
AND THEN, be able to perform multiple queries on $result:
$newResult = mysql_query("SELECT * FROM $result WHERE otherBar=".$barValue);
OR
$otherNewResult = mysql_query("SELECT * FROM $result WHERE otherOtherBar=".$barValue." ORDER BY foobar ASC");
AND so on and so forth...
I realize that I could append the original query with my new WHERE statements and ORDER BYs, but that causes my to query the database unnecessarily and it prevents me from writing more objected oriented code (because I can't pass around a result to be queried, but rather have to requery the database in every function...)
Any advice, pieces of code, frameworks, or ramblings appreciated.
*BTW, my project was having to query a large database of people for people born in certain age groups and then query those age groups for different demographics.
Edit
No, writing a custom function to query the database is not worth the object-orientation (and modifiability) it would give me
You could do a nested query in the same SQL query and keep PHP out of it:
'SELECT * FROM (SELECT * FROM foo WHERE bar="something") AS q1 WHERE q1.bar2 = "something else"'
The question has already been answered. However following explanation will help someone who might be interested in knowing the details of it.
What are Nested query / subquery:
Subqueries are also known as nested queries. A subquery is a SELECT statement within another statement. MySQL supports all SQL standards and additionally provides MySQL specific features.
Why should I use Subquery:
Subquery is structured and it is possible to isolate each parts of statement
Subquery is more readable that complex joins and unions
Subquery provides alternative means to perform action which otherwise would require complex joins and unions
What Subquery returns:
A subquery can return a single value, a single row, a single column, or a table. These are called scalar, column, row, and table subqueries.
Reference: http://dev.mysql.com/doc/refman/5.7/en/subqueries.html
http://www.w3resource.com/sql/subqueries/nested-subqueries.php
Related
I am trying to display the data from 'table' if a key inputted by the user is found in the database. Currently I have it set up so that the database checks if the key exists, like so:
//Select all from table if a key entry that matches the user specified key exists
$sql = 'SELECT * FROM `table` WHERE EXISTS(SELECT * FROM `keys` WHERE `key` = :key)';
//Prepare the SQL query
$query = $db->prepare($sql);
//Substitute the :key placeholder for the $key variable specified by the user
$query->execute(array(':key' => $key));
//While fetched data from the query exists. While $r is true
while($r = $query->fetch(PDO::FETCH_ASSOC)) {
//Debug: Display the data
echo $r['data'] . '<br>';
}
These aren't the only SQL statements in the program that are required. Later, an INSERT query along with possibly another SELECT query need to be made.
Now, to my understanding, using WHERE EXISTS isn't always efficient. However, would it be more efficient to split the query into two separate statements and just have PHP check if any rows are returned when looking for a matching key?
I took a look at a similar question, however it compares multiple statements on a much larger scale, as opposed to a single statement vs a single condition.
#MarkBaker Join doesn't have to be faster than exists statement. Query optymalizer is able to rewrite the query live if it sees better way to accomplish query. Exists statement is more readable than join.
Fetching all the data and making filtering directly in PHP is always bad idea. What if your table grow up to milions of records? MySQL is going to find the best execute plan for you. It will automaticaly cache the query if it is going to improve performance.
In other words, your made everything correctly as far as we can see your code now. For futher analyse show us all of your queries.
This question already has answers here:
Why is SELECT * considered harmful?
(16 answers)
Closed 9 years ago.
I am developing an application and I was reading about how queries work. I read somewhere that you should avoid SELECT * FROM... where blah = blah
Why is that? And what's the workaround if you're trying to select pretty much everything?
Initially need to know what data you will need. Although, you can select all at once, if such requests will not be much. The difference in performance, you'll see only in heavy projects.
This is not really a direct answer to your question "Why is that?" (so downvote the answer if you need to.) It's an answer to the "what's a workaround if you need to" question.
The only workaround to avoid SELECT *, when I need all of the columns in the table, is to get a list of all the columns. And that's just extra busy I work I don't need when I'm already busy.
To put a backwards twist on a line from Office Space charaacter Peter Gibbons: "The thing is, Bob, it's not that I don't care, it's just that I'm lazy."
With MySQL I make for less busy work by using the SQLyog right click menu option to generate a skeleton SELECT statement that contains all the columns.
For a SQL statement that references multiple tables, I want every column reference to be qualified with a table alias, so I'll just use a SQL statement to retrieve a ready-to-use list of columns for me:
SELECT GROUP_CONCAT(CONCAT('t.',c.column_name)
ORDER BY c.ordinal_position
) AS col_list
FROM information_schema.columns c
WHERE c.table_schema = 'mydatabase'
AND c.table_name = 'mytable'
if I only need a few out a long list, it's easier for me to get them out of a vertical list
SELECT CONCAT(',s.`',c.column_name,'`') AS col_names
FROM information_schema.columns c
WHERE c.table_schema = 'mydatabase'
AND c.table_name = 'mytable'
ORDER BY c.ordinal_position
When the column references are qualified, the backticks are only needed for "special" characters in column names (or maybe some weird case sensitive setting.)
I can start with that list, and whittle out the columns I know I don't need.
Again, I apologize that this doesn't answer the question "Why?" There's several good reasons, given in answers to similar questions. For me, a big reason is that a future reader of the statement isn't going to have to go look somewhere else to find out what columns are being returned. Sure they can copy the statement, and go run it in a different environment, to see the list. But if the statement has variable substitutions, bind variables, and the dots and double quotes and calls to mysql_real_escape_string (in the case of mysql_ interface), that's a bigger hassle than it needs to be. Sure, the code can be modified to echo out the SQL text before its executed, and the reader may need to do that. But someone just reviewing the code shouldn't have to do that. And having the list of columns and expressions being returned by the statement, in an order more appropriate than the ordinal position of the columns in the table, I think that just makes for more readable code. (If it's important for the column to be returned by the statement, then I think it's reasonable that the column name is shown in the query.)
(This was in terms of application code, statements that are going to be included in an application. For ad hoc queries and development and such, I use the SELECT c.* freely. But when a statement is going into an application, that * gets replaced.
I'm using a PHP webservice where I have performed a simple SELECT query, and stored it
$result = run_query($get_query);
I now need to perform further querying on the data based on different parameters, which I know is possible via MySQL in the form:
SELECT *
FROM (SELECT *
FROM customers
WHERE CompanyName > 'g')
WHERE ContactName < 'g'
I do know that this performs two Select queries on the table. However, what I would like to know is if I can simply use my previously saved query in the FROM section of the second section, such as this, and if my belief that it helps performance by not querying the entire database again is true:
SELECT *
FROM ($result)
WHERE ContactName < 'g'
You can make a temp table to put the initial results and then use it to select the data and in the second query. This will work faster only if your 1-st query is slow.
PHP and SQL are different languages and very different platforms. They often don't even run in the same computer. Your PHP variables won't interact at all with the MySQL server. You use PHP to create a string that happens to contain SQL code but that's all. In the end, the only thing that counts is the SQL code you sent to the server—how you manage to generate it is irrelevant.
Additionally, you can't really say how MySQL will run a query unless you obtain an explain plan:
EXPLAIN EXTENDED
SELECT *
FROM (SELECT *
FROM customers
WHERE CompanyName > 'g')
WHERE ContactName < 'g'
... but I doubt it'll read the table twice for your query. Memory is much faster than disk.
Thanks for the responses, everyone. Turns out what I was looking for was a "query of query", which isn't supported directly by PHP but I found a function over here which provides the functionality: http://www.tom-muck.com/blog/index.cfm?newsid=37
That was found from this other SO question: Can php query the results from a previous query?
I still need to do comparisons to determine whether it improves speed.
If I understand your question correctly you want to know whether saving the "from" part of your SQL query in a php variable improves the performance of you querying your SQL server, then the answer is NO. Simply because the variable keeping the value is inserted into the query.
Whether performance is gained in PHP, the answer is most probable yes; but depends on the length of the variable value (and how often you repeat using the variable instead of building a new complete query) whether the performance will be notable.
Why not just get this data in a single query like this?
SELECT *
FROM customers
WHERE CompanyName > 'g'
AND ContactName < 'g'
Until recently I've been using mysql_real_escape_string() to fix most of my variables before making SQL queries to my database. A friend said that I should be using PDO's prepared statements instead, so after reading a bit about them I'm now switching over to them.
I've only encountered one problem so far in switching over, and that's counting the rows to returned by a SELECT statement. On occasion in my code, I'd run an SQL query and then count the number of rows returned from the SELECT statement. Depending on whether a result set returned, I would take different actions. Sometimes I do need to use the result set from it. MySQL let me go straight to mysql_fetch_assoc() after mysql_num_rows() with no problem. However, PDO doesn't seem to have anything like mysql_num_rows().
I've been reading some responses on SO that gave me a solution, to either use COUNT() in the SQL statement or to use the PHP function count() on the result set. COUNT() would work fine in the SQL statement if I didn't need the result set in some places, however, several people have mentioned that using count() on the result set is fairly inefficient.
So my question is, how should I be doing this if I need to count the number of rows selected (if any), then run a script with the result set? Is using count() on the result set the only way in this case, or is there a more efficient way to do things?
Below is a short example of something similar to my previous SQL code:
$query=mysql_query('SELECT ID FROM Table WHERE Name='Paul' LIMIT 1);
if(mysql_num_rows($query)>0)
{
print_r(mysql_fetch_assoc($query));
}
else
{
//Other code.
}
Thanks.
EDIT
I do know that you use fetchAll() on the statement before counting the result set (which gives me what I need), but I'm just trying to figure out the most efficient way to do things.
$stmt->rowCount();
http://php.net/manual/en/pdostatement.rowcount.php
the rows must be fetched(buffered into memory, or iterated) for it to work. It's not uncommon for your pdo driver to be configured to do this automatically.
You will have to use Count(). You can run two queries like
SELECT COUNT(ID) FROM Table WHERE Name='Paul'
one you have get the count, then run the query with select clause
SELECT ID FROM Table WHERE Name='Paul' LIMIT 1
Count() function is not inefficient at all if you are using it like COUNT(ID), because most probably id is primary key and have an index. MYSQL wont even have to access the table.
I've recently started work on a new project using PHP5 and want to use their PDO classes for it. The problem is that the MySQL PDO Driver doesn't support rowCount() so there's no way to run a query and then get the number of affected rows, or rows returned, which is a pretty big issue as far as I'm concerned. I was wondering if anyone else has dealt with this before and what you've done to work around it. Having to do a fetch() or fetchAll() to check if any rows were affected or returned seems like a hack to me, I'd rather just do $stmt->numRows() or something similar.
You can issue a SELECT FOUND_ROWS() query right after the original SELECT query to get row count.
$pdo->query("SELECT * FROM users");
$foundRows = $pdo->query("SELECT FOUND_ROWS()")->fetchColumn();
See also: MySQL Docs on FOUND_ROWS()
For those of you who are using MySQL stored procedures, this solution isn't really feasible. What I would suggest that you do is have your stored procedure create two rowsets. The first one will contain one row and one column, containing the number of records. The second will be the recordset you will use for fetching that number of rows.
The number of unlimited rows can be a SELECT COUNT(*) with the exact same WHERE clause as the second rowset without the LIMIT/OFFSET clauses.
Another idea could be to create a temporary table. Use your SELECT statement to populate the temporary table. Then you can use SELECT COUNT(*) FROM tmpTable for your first rowset and SELECT * FROM tmpTable for your second.
This question is based on several false assumptions and one outdated statement.
First of all, do not confuse number of affected and selected rows. PDO supported the former even back in '09.
Speaking of number of rows returned by SELECT statement - you just don't need that number. The data you have is enough.
And yeah, nowadays rowCount() supports number of rows selected from mysql as well. But again - you don't need that number in an average web-application anyway.