I have a requirement where I need to insert user_id in following format
13310_userid_1
13310_userid_2
13310_userid_3
where
13310 = $_GET['userid'] //user id from session
userid = constant //constant text defined
1/2/3 = autoincrement value
Here the variation is when user_id is changed, the auto increment value will be inserted from beginning which would look like
13311_userid_1
13311_userid_2
13311_userid_2
and not
13311_userid_4
13311_userid_5
13311_userid_6
How can I check if the user_id is changed and insert auto increment value from 1 ?
Thanks
I'm not sure it's that you need, but you can use an array to store increment of each cont :
$const = 'userid';
$user_id = '13310';
$array_increment[$user_id] =1;
foreach(array('Franck','Robert','You','Me') as $index=>$test){
if($index==2)$user_id = '13311';
if(!isset($array_increment[$user_id])){
$array_increment[$user_id]=1;
}
$increment = $array_increment[$user_id];
echo $user_id.'_'.$const.'_'.$increment.'<br />';
$array_increment[$user_id]++;
}
Will show :
13310_userid_1
13310_userid_2
13311_userid_1
13311_userid_2
you could store the incremental value of each userid in a temporary session:
For each request check if $_GET['userid'] already exists in a session if not create it with value zero.
Increment by one and use this value to create your string for inserting into DB
//set up session if it is not already set
if(!isset($_SESSION['users'][$_GET['userid'])){
$_SESSION['users'][$_GET['userid']=0;
}
$_SESSION['users'][$_GET['userid']=$_SESSION['users'][$_GET['userid']+1;
$user=$_GET['userid'].'_userid_'.$_SESSION['users'][$_GET['userid'];
//insert $user into your DB or permanent storage.
$db->insert($user);
(this should be a comment but its a bit verbose)
Your only makes sense if a database figures in here somewhere - but you never explicitly stated that this is the case nor which database it is.
Relational database design (and to quite a large extent, non-relational database design) is subject to the rules of normalization. These are methods for describing the structure of your data and prevent you doing stupid things. This breaks the first rule.
Assuming you were designing the system properly, then you would keep the three attributes as seperate fields. But that does not answer the question of whether the id is nominal, cardinal or ordinal (and in the case of ordinal numbers whether there is a requirement for them to be consecutive).
Related
My idea is to make tooltip for new users on website. They will go through the tooltips and each time they complete a tooltip it inserts into DB.
However i've got a button which is skip all. So my idea is to insert all the reminder tooltips which they've not completed into the DB with for loop.
So it works if there has been no tooltips clicked already. So the tooltip would be equal to 1 because coming through the $data is equal to 0. However if the tooltip is equal to 1 when passed through $data it gets a +1 and the for loop doesn't seem to post anything into database
$NumberOfTooltips = 2;
(int)$data->tooltipLastID = ((int)$data->tooltipLastID === 0) ? 1 : (int)$data->tooltipLastID + 1;
for ($x = (int)$data->tooltipLastID; $x <= $NumberOfTooltips; $x++) {
$query = "";
$query = "INSERT INTO tooltips ";
$query .= "(tooltip_id, inserted) VALUES ($x, NOW())";
database::query($query);
$this->id = database::getInsertID();
}
On the broken loop the value of (int)$data->tooltipLastID is 2
Is it because the (int)$data->tooltipLastID is already equal to $NumberOfTooltips?
General improvements
These do not directly solve the question asked but they do give you a helping hand in clarifying your data and steaming out any secondary bugs and bloopers. Also pointing out some best (or at least, better) practise.
$x is a lazy and poorly defined counter. Prefer using descriptve veraibles such as $tooltipCounter
$data->tooltipLastID should not start at 1; use the same syntax as every other integer number system in PHP/programming and start at zero. If you need a one then add +1 only when it's needed (VALUES (".$x+1.")).
$NumberOfTooltips = 2; The number 2 is probably not high enough for adequate testing.
var_dump($data->tooltipLastID) and var_dump($NumberOfTooltips) to check both values are what you expect.
Rewrite the test code to take the variables out of the code so that you can check your Database connction works correctly (such as if you're trying to insert into a string field-type by mistake)
$query = ""; is redundant.
You should not need to type cast (int) your object variables ($data->whatever) all the time but type cast them when they're set.
Also by adding +1 to a variable PHP automatically recasts the variable as an int anyway.
Check that your $data->tooltipLastID is publicly accessible/writable.
You use $this ; so which class are you in? Are you self referencing the data class?
A bank holiday is just one day.
It is better the inserted Database column is set by the database automatically upon insert. You can use this SQL to alter your current table:
ALTER TABLE <yourTable> MODIFY COLUMN inserted timestamp DEFAULT CURRENT_TIMESTAMP
Check the type of $data->tooltipLastID? And plz use var_dump($data->tooltipLastID) and var_dump((int)$data->tooltipLastID) before the for loop to see what indeed the original value and the $x is.
Strange type casts will result in strange bugs...
I'm using PHP in order to create a website where managers have access and review forms that employees have submitted. In the reviewing PHP file, I have created two buttons which basically approve or disapprove the form. After they click on one of the buttons, they are being redirected to another PHP file which actually inserts into the MySQL Database a change in a column I named 'processed'. It changes 0 which is unprocessed to 1, which is processed. The table I am referring to has columns such as formid, fullname, department and other job related stuff, as well as the 'processed' column which allows the managers to see if there is a pending form to be reviewed.
My problem is that I have no idea how to actually allow MySQL to find the proper row and change only the cell with the name 'processed' from 0 to 1 without having to insert every cell again. Here's what I have tried till now:
$id = $_SESSION[id];
$fullname = $_SESSION[fullname];
$teamformid = $_SESSION[teamformid];
if (isset($_POST['approved'])) {
$sql = "INSERT INTO carforms (processed) where aboveid='$id' and processed='0' and teamformid=$teamformid
VALUES ('0')";
}
else if (isset($_POST['disapproved'])) {
//todo
}
How do I tell SQL to only find the specific row I want and change only one column which is processed?
Also, do I always have to type every column name when I use the INSERT INTO command?
Thanks in advance.
Use the Below code it'll work for you.
$id = $_SESSION[id];
$fullname = $_SESSION[fullname];
$teamformid = $_SESSION[teamformid];
if (isset($_POST['approved'])) {
$sql = "UPDATE `carforms` SET processed = '1' WHERE `aboveid` = '".$id."' AND `teamformid` = '".$teamformid."'";
}
Try:
"UPDATE carforms SET processed = 1 WHERE aboveid = $id AND teamformid = $teamformid"
From what I have interpreted from your question, it seems like you need to use the MySQL UPDATE command. This will update any existing rows.
For example, let's say you have a table called 'forms', consisting of a Primary Key 'form_id' and a field named 'processed'.
If we want to change the value of 'processed' to '1', we would run...
UPDATE forms SET processed = 1 WHERE form_id = [whatever number the form is];
Obviously this only works where the form (with a form_id) exists already
There is no "INSERT...WHERE" in SQL.
To change an existing record there are 2 options, REPLACE or UPDATE. The former will create the record if it does not already exist and has similar syntax to INSERT. UPDATE uses the WHERE clause to identify the record(s) to be changed.
Using REPLACE is tricky. It needs to work out whether it should INSERT a new record or UPDATE an existing one - it does this by checking if the data values presented already exist in a unique index on the table - if you don't have any unique indexes then it will never update a record. Even if you have unique indexes just now, the structure of these may change over time as your application evolves, hence I would recommend NOT using REPLACE for OLTP.
In your question you write:
where aboveid='$id' and processed='0' and teamformid=$teamformid
(it would have been helpful if you had published the relevant part of the schema)
'id' usually describes a unique identifier. So there shouldn't be multiple records with the same id, and therefore the remainder of the WHERE clause is redundant (but does provide an avenue for SQL injection - not a good thing).
If the relevant record in carforms is uniquely identifed by a value for 'id' then your code should be something like:
$id=(integer)$id;
$sql = "UPDATE carforms SET processed = $action WHERE aboveid=$id";
But there's another problem here. There are 3 possible states for a record:
not yet processed
declined
approved
But you've only told us about 2 possible states. Assuming the initial state is null, then the code should be:
$action=0;
if (isset($_POST['approved'])) {
$action=1;
}
$id=(integer)$id;
$sql = "UPDATE carforms SET processed = $action WHERE aboveid=$id";
if ($id &&
(isset($_POST['disapproved']) || isset($_POST['approved']))
) {
// apply the SQL to the database
} else {
// handle the unexpected outcome.
}
usersim interested how do i select a text field form my mysql database, i have a table named users with a text field called "profile_fields" where addition user info is stored. How do i access it in php and make delete it? I want to delete unvalidate people.
PHP code
<?php
//Working connection made before assigned as $connection
$time = time();
$query_unactive_users = "DELETE FROM needed WHERE profile_fields['valid_until'] < $time"; //deletes user if the current time value is higher then the expiring date to validate
mysqli_query($connection , $query_unactive_users);
mysqli_close($connection);
?>
In phpmyadmin the field shows (choosen from a random user row):
a:1:{s:11:"valid_until";i:1370695666;}
Is " ... WHERE profile_fields['valid_until'] ..." the correct way?
Anyway, here's a very fragile solution using your knowledge of the string structure and a bit of SUBSTRING madness:
DELETE FROM needed WHERE SUBSTRING(
profile_fields,
LOCATE('"valid_until";i:', profile_fields) + 16,
LOCATE(';}', profile_fields) - LOCATE('"valid_until";i:', profile_fields) - 16
) < UNIX_TIMESTAMP();
But notice that if you add another "virtual field" after 'valid_until', that will break...
You can't do it in a SQL command in a simple and clean way. However, the string 'a:1:{s:11:"valid_until";i:1370695666;}' is simply a serialized PHP array.
Do this test:
print_r(unserialize('a:1:{s:11:"valid_until";i:1370695666;}'));
The output will be:
Array ( [valid_until] => 1370695666 )
So, if you do the following, you can retrieve your valid_until value:
$arrayProfileData = unserialize('a:1:{s:11:"valid_until";i:1370695666;}');
$validUntil = arrayProfileData['valid_until'];
So, a solution would be to select ALL items in the table, do a foreach loop, unserialize each "profile_fields" field as above, check the timestamp, and store the primary key of each registry to be deleted, in a separate array. At the end of the loop, do a single DELETE operation on all primary keys you stored in the loop. To do that, use implode(',', $arrayPKs).
It's not a very direct route, and depending on the number of registers, it may not be slow, but it's reliable.
Consider rixo's comment: if you can, put the "valid_until" in a separate column. Serializing data can be good for storage of non-regular data, but never use it to store data which you may need to apply SQL filters later.
I have a view that needs updating with a list of id's. So I am storing the values that have been selected to remove from the view in a session variable that then goes into the mySQL query as below. Then when the form is reset the values are also reset out of the array.
But its not working... this is what I've got.
Any help would be appreciated.
if($_POST['flag']=='flag'){
//collect deleted rows
$_SESSION['delete-row'][] = $_POST['idval'];
//Split session array
$idavls = join(',' , $_session['delete-row'];
$sqlDelete = "CREATE OR REPLACE VIEW filtetbl AS SELECT * FROM `".$page['db-name']."`.`leads_tbl` WHERE ".$_SESSION['filter-view']." AND `lead_status` = '1' AND `lead_id` NOT IN (".$idvals.") ORDER BY `lead_added`";
$result = mysql_query($sqlDelete);
if($result){
echo true;
}
else{
echo mysql_error();
}
}
$_session isnt the same as $_SESSION for a start.
Also dont use mysql_query or similar (because it isnt safe) use PDO
This is hard to correct without more information (and there are several errors - probaby cut and paste) so I'll pull apart one by one and you can go from there.
1 - $_SESSION['delete-row'][] = $_POST['idval'];
If 'idval' comes from multiple inputs (i.e. ) then it is already an array, and you should have $_SESSION['delete-row'] = $_POST['idval']; If you are looping in an array of inputs (i.e. trying to append for many posts from then it is correct)
2 - $idavls = join(',' , $_session['delete-row'];
$_SESSION (you said this was a type) and you also need a bracket/bract ar the end
$sqlDelete = "CREATE OR REPLACE VIEW filtetbl AS SELECT * FROM ".$page['db-name'].".leads_tbl WHERE ".$_SESSION['filter-view']." AND lead_status = '1' AND lead_id NOT IN (".$idvals.") ORDER BY lead_added";
Firsly this is very insecure as pointed out by allen213. Even if you don't use PDO to make safe the variable, please cast all the inputs as (int) assuming the IDs are integers, or at least wrap the input in mysql_real_escape_string().
Secondly, the logic in the question doesn't quite make sense. You say you want to remove IDs from the view, but what you are doing is recreating the view with only those IDs in $_SESSION['delete-row'] removed - so this may re-introduce IDs previously removed from the view. You'd actually need to keep $_SESSION['delete-row'] and keep adding to it to ensure the next time the view was created, then all the IDs are removed.
I hope that helps. If not, more code may be required (i.e. the form you are using the send data, anythign else that affects sessions etc.
I am developing a mysql database.
I "need" a unique id for each user but it must not auto increment! It is vital it is not auto increment.
So I was thinking of inserting a random number something like mt_rand(5000, 1000000) into my mysql table when a user signs up for my web site to be. This is where I am stuck?!
The id is a unique key on my mysql table specific to each user, as I can not 100% guarantee that inserting mt_rand(5000, 1000000) for the user id will not incoherently clash with another user's id.
Is there a way in which I can use mt_rand(5000, 1000000) and scan the mysql database, and if it returns true that it is unique, then insert it as the user's new ID, upon returning false (somebody already has that id) generate a new id until it becomes unique and then insert it into the mysql database.
I know this is possible I have seen it many times, I have tried with while loops and all sorts, so this place is my last resort.
Thanks
You're better off using this: http://dev.mysql.com/doc/refman/5.0/en/miscellaneous-functions.html#function_uuid
Or using this: http://dev.mysql.com/doc/refman/5.0/en/insert-on-duplicate.html
But if you actually want to do what you are saying, you can just do something like:
$x;
do {
$x = random_number();
"SELECT count(*) FROM table WHERE id = $x"
} while (count != 0);
// $x is now a value that's not in the db
You could use a guid. That's what I've seen done when you can't use an auto number.
http://php.net/manual/en/function.com-create-guid.php
Doesn't this function do what you want (without verification): http://www.php.net/manual/en/function.uniqid.php?
I think you need to approach the problem from a different direction, specifically why a sequence of incrementing numbers is not desired.
If it needs to be an 'opaque' identifier, you can do something like start with a simple incrementing number and then add something around it to make it look like it's not, such as three random numbers on the end. You could go further than that and put some generated letters in front (either random or based on some other algorithm, such as the day of the month they first registered, or which server they hit), then do a simple checksuming algorithm to make another letter for the end. Now someone can't easily guess an ID and you have a way of rejecting one sort of ID before it hits the database. You will need to store the additional data around the ID somewhere, too.
If it needs to be a number that is random and unique, then you need to check the database with the generated ID before you tell the new user. This is where you will run into problems of scale as too small a number space and you will get too many collisions before the check lucks upon an unallocated one. If that is likely, then you will need to divide your ID generation into two parts: the first part is going to be used to find all IDs with that prefix, then you can generate a new one that doesn't exist in the set you got from the DB.
Random string generation... letters, numbers, there are 218 340 105 584 896 combinations for 8 chars.
function randr($j = 8){
$string = "";
for($i=0;$i < $j;$i++){
srand((double)microtime()*1234567);
$x = mt_rand(0,2);
switch($x){
case 0:$string.= chr(mt_rand(97,122));break;
case 1:$string.= chr(mt_rand(65,90));break;
case 2:$string.= chr(mt_rand(48,57));break;
}
}
return $string;
}
Loop...
do{
$id = randr();
$sql = mysql_query("SELECT COUNT(0) FROM table WHERE id = '$id'");
$sql = mysql_fetch_array($sql);
$count = $sql[0];
}while($count != 0);
For starters I always prefer to do all the randomization in php.
function gencode(){
$tempid=mt_rand(5000, 1000000);
$check=mysql_fetch_assoc(mysql_query("SELECT FROM users WHERE id =$tempid",$link));
if($check)gencode();
$reg=mysql_query("INSERT INTO users id VALUES ('$tempid')",$link);
//of course u can check for if $reg then insert successfull