Deleting a row in certain conditions with mysql - php

I want to automatically delete rows when the table (shown below) gets a new insert, if certain conditions are met.
When:
There are rows referring to the same 'field' with the same 'user_id'
Their 'field', 'display' and 'search' columns are the same
Simply, when the rows would become duplicates (except the 'group_id' column) the non null 'group_id' should be deleted, otherwise a row should be updated or inserted.
Is there a way to set this up in mysql (in spirit of "ON DUPLICATE do stuff" combined with unique keys etc.), or do I have to explicitly check for it in php (with multiple queries)?
Additional info:
There should always be a row with NULL 'group_id' for every possible 'field' (there's a limited set, defined elsewhere). On the other hand there might not be one with a non null 'group_id'.
CREATE TABLE `Views` (
`user_id` SMALLINT(5) UNSIGNED NOT NULL,
`db` ENUM('db_a','db_b') NOT NULL COLLATE 'utf8_swedish_ci',
`field` VARCHAR(40) NOT NULL COLLATE 'utf8_swedish_ci',
`display` TINYINT(1) UNSIGNED NOT NULL,
`search` TINYINT(1) UNSIGNED NOT NULL,
`group_id` SMALLINT(6) UNSIGNED NULL DEFAULT NULL,
UNIQUE INDEX `user_id` (`field`, `db`, `user_id`),
INDEX `Views_ibfk_1` (`user_id`),
INDEX `group_id` (`group_id`),
CONSTRAINT `Views_ibfk_1` FOREIGN KEY (`user_id`) REFERENCES `User` (`id`) ON
UPDATE CASCADE ON DELETE CASCADE
)
COLLATE='utf8_swedish_ci'
ENGINE=InnoDB;

I think you need to revise your logic. It makes no sense to Insert a row only to delete another row. Why not just update the Group_ID field in the duplicate row to what is being inserted? Below is a rough idea of how I would go about it.
N.b. I haven't done much work with MySQL and cannot get the below to run on SQLFiddle, but based on the MySQL docs I can't work out why. Perhaps someone more versed in MySQL can correct me?
SET #User_ID = 1;
SET #db = 'db_a';
SET #Field = 'Field';
SET #Display = 1;
SET #Search = 1;
SET #Group_ID = 1;
IF EXISTS
( SELECT 1
FROM Views
WHERE User_ID = #User_ID
AND DB = #DB
AND Field = #Field
AND Group_ID IS NOT NULL
)
THEN
UPDATE Views
SET Group_ID = #Group_ID,
Display = #Display,
Search = #Search
WHERE User_ID = #User_ID
AND DB = #DB
AND Field = #Field
AND Group_ID IS NOT NULL
ELSE
INSERT INTO Views (User_ID, DB, Field, Display, Search, Group_ID)
VALUES (#User_ID, #DB, #Field, #Display, #Search, #Group_ID)
END IF;
Alternatively (and my preferred solution), add a Timestamp field to your table and create a view as follows:
SELECT v.User_ID, v.DB, v.Field, v.Display, v.Search, v.Group_ID
FROM Views v
INNER JOIN
( SELECT User_ID, DB, Field, MAX(CreatedDate) AS CreatedDate
FROM Views
WHERE Group_ID IS NOT NULL
GROUP BY User_ID, DB, Field
) MaxView
ON MaxView.User_ID = v.User_ID
AND MaxView.DB = v.DB
AND MaxView.Field = v.Field
AND MaxView.CreatedDate = v.CreatedDate
WHERE v.Group_ID IS NOT NULL
UNION ALL
SELECT v.User_ID, v.DB, v.Field, v.Display, v.Search, v.Group_ID
FROM Views v
WHERE v.Group_ID IS NULL
This would allow you to track changes to your data properly, without compromising the need to be able to view unique records.

delete group_id from Views where group_id != 'NUll'

Your question is not very good to understand, so I'm not sure this is what you want:
DELETE FROM Views WHERE # delete from the table views
group_id IS NOT NULL AND # first condition delete only rows with not null group_id
(SELECT count(*) as tot FROM Views GROUP BY group_id) = 1 # second condition count the difference in group id
If that's not what you want, please update your question with more details...

Related

Yii2 ActiveQuery join keep returns distinct values

I have two tables as below
table halte :
CREATE TABLE `halte` (
`id` int(11) NOT NULL AUTO_INCREMENT,
`nama` varchar(255) NOT NULL,
`lat` float(10,6) DEFAULT NULL,
`lng` float(10,6) DEFAULT NULL,
PRIMARY KEY (`id`)
)
table stops :
CREATE TABLE `stops` (
`id` int(11) NOT NULL AUTO_INCREMENT,
`id_halte` int(11) DEFAULT NULL,
`sequence` int(2) DEFAULT NULL,
PRIMARY KEY (`id`),
KEY `id_halte` (`id_halte`)
)
I also have some other tables which don't cause any problems.
Halte table has many to one relation to stop. The problem is when i try to get rows from halte table using right join to table stops, Yii only returns unique rows. Yii won't return same halte's row more once even stop table has more than one record related to same row in halte table.
Here's my code
$haltes = $modelHalte->find()
->rightJoin('stops', 'halte.id = stops.id_halte')
->where(['stops.id_rute'=>Yii::$app->request->get('rute')])
->orderBy('sequence')
->all();
I have tried distinct(false) but no result.
I've also check debugger and it run right query i want :
SELECT `halte`.* FROM `halte` RIGHT JOIN `stops` ON halte.id = stops.id_halte WHERE `stops`.`id_rute`='1' ORDER BY `sequence`
I tried to run that query manually and it returned 29 rows which is what what i want. But in Yii, it only returned 27 rows because 2 rows is same record in halte table.
I know i can achieve this using yii\db\Query, but i want to use ActiveRecord.
Are there any way to work around this?
I would really appreciate your opinion/help.
Thanks.
Check the sql command generated by you active query
$haltes = $modelHalte->find()
->rightJoin('stops', 'halte.id = stops.id_halte')
->where(['stops.id_rute'=>Yii::$app->request->get('rute')])
->orderBy('sequence')
->all();
echo $haltes->createCommand()->sql;
or to get the SQL with all parameters included try:
$haltes->createCommand()->getRawSql();
And compare the code generated by ActiveQuery with your created manually ..

MySQL INSERT IGNORE Adding 1 to Non-Indexed column

I'm building a small report in a PHP while loop.
The query I'm running inside the while() loop is this:
INSERT IGNORE INTO `tbl_reporting` SET datesubmitted = '2015-05-26', submissiontype = 'email', outcome = 0, totalcount = totalcount+1
I'm expecting the totalcount column to increment every time the query is run.
But the number stays at 1.
The UNIQUE index composes the first 3 columns.
Here's the Table Schema:
CREATE TABLE `tbl_reporting` (
`datesubmitted` date NOT NULL,
`submissiontype` varchar(20) COLLATE utf8mb4_unicode_ci NOT NULL,
`outcome` tinyint(1) unsigned NOT NULL DEFAULT '0',
`totalcount` mediumint(5) unsigned NOT NULL DEFAULT '0',
UNIQUE KEY `datesubmitted` (`datesubmitted`,`submissiontype`,`outcome`)
) ENGINE=MyISAM DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci
When I modify the query into a regular UPDATE statement:
UPDATE `tbl_reporting` SET totalcount = totalcount+1 WHERE datesubmitted = '2015-05-26' AND submissiontype = 'email' AND outcome = 1
...it works.
Does INSERT IGNORE not allow adding numbers? Or is my original query malformed?
I'd like to use the INSERT IGNORE, otherwise I'll have to query for the original record first, then insert, then eventually update.
Think of what you're doing:
INSERT .... totalcount=totalcount+1
To calculate totalcount+1, the DB has to retrieve the current value of totalcount... which doesn't exist yet, because you're CREATING a new record, and there is NO existing data to retrieve the "old" value from.
e.g. you're trying eat your cake before you ever went to the store to buy the ingredients, let alone mix/bake them.

How would I be able to duplicate my auto_increment? DATABASE PHP

Hey how would I be able to duplicate my only auto increment key to another key, basically I want my (' id ') to display the same information on my (' user_id '), here is the code:
CREATE TABLE IF NOT EXISTS `".$db_table_prefix."users` (
`id` int(11) NOT NULL AUTO_INCREMENT,
`user_id` int(10) NOT NULL,
`user_name` varchar(50) NOT NULL,
`username` varchar(50) NOT NULL,
PRIMARY KEY (`id`),
KEY `user_id` (`id`)
How would I be able to input the same information from my id to my user_id?
Not sure what you mean but if you want to have the same value repeted two times in the same record It's pointless and redundant.
You can use the SQL aliases to achive what you want:
SELECT id as user_id FROM ...
If you really need to sync up the two field of your table you can do:
UPDATE table SET user_id = id WHERE user_id != id
Not sure why you would want to do this, but if you want to duplicate the information after an INSERT you would need to fetch the new ID and then perform an UPDATE
// get the newly inserted ID
$new_id = $db->insert_id;
// perform the update on the table
$db->query("UPDATE users SET user_id=".$db->escape($new_id)." WHERE id=".$db->escape($new_id));
Also, in your table definition the fields don't match: int(11) vs. int(10).

what's wrong with this SQL statement causing column count doesn't match value count at row 1?

Mysql table (migration_terms) fields are as follows
oldterm count newterm seed
I used the following create table statment.
CREATE TABLE `migration_terms`
(
`oldterm` varchar(255) DEFAULT NULL,
`count` smallint(6) DEFAULT '0',
`newterm` varchar(255) DEFAULT NULL,
`seed` int(11) NOT NULL AUTO_INCREMENT, PRIMARY KEY (`seed`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8
And It works, no problems there.
but then when I used the following insert into statement to populate it;
"INSERT INTO migration_terms
SELECT looseterm as oldterm,
COUNT(seed) AS count
FROM looseterms
GROUP BY looseterm
ORDER BY count DESC "
I get this error;
Column count doesn't match value count at row 1
I cannot figure out why?
If you need the table structure of the looseterms table, it was created by the following create table statement.
CREATE TABLE looseterms
(
`seed` INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
`looseterm` varchar(255)
)
You need to specify the columns if your select statement has fewer columns than the table
"INSERT INTO migration_terms
(oldterm,
count)
SELECT looseterm AS oldterm,
Count(seed) AS count
FROM looseterms
GROUP BY looseterm
ORDER BY count DESC "
From MySql docs on Insert Syntax
If you do not specify a list of column names for INSERT ... VALUES or
INSERT ... SELECT, values for every column in the table must be
provided by the VALUES list or the SELECT statement. If you do not
know the order of the columns in the table, use DESCRIBE tbl_name to
find out.
Your insert is adding 2 columns of data, whereas your table's definition has 4 columns

Developing a Tagging System - How to update present tags with new ones?

I am designing a Tagging system for my site(a customized blog). These are my source codes
<b>Please enter your tags, separating each one with a comma.<b><br>
<input type="text" name="tags" size=50>
$taginput = $_POST["tags"];
$tagarray = explode(",",$taginput);
for($i=0;$i<count($tagarray);$i++){
$usetag = mysql_real_escape_string(stripslashes(ltrim(rtrim($tagarray[$i]))));
if($usetag == "") continue;
$query = "INSERT INTO tags (item_id,tag) VALUES ($itemid,'$usetag')";
mysql_query($query);
}
CREATE TABLE `tags` (
`uid` INT( 11 ) UNSIGNED NOT NULL AUTO_INCREMENT PRIMARY KEY ,
`item_id` INT( 11 ) NOT NULL,
`tag` VARCHAR( 50 ) NOT NULL
) ENGINE = MYISAM ;
Where item_id is the blog post id and tag is the tag name
Now Im looking for a solution on how if I update a post with tags.. how can i delete the present tags and replace it with the new tags? and how can i delete some specific present tags and replacing it with new ones?
Like this,
Tags: shirt, apparel, arrivals replaced by Tags: t-shirts, old, sale
and
Tags: shirt, apparel, arrivals to Tags: shirt, old, arrivals
I also want to prevent overpopulating the tags tables with duplicate tag names if possible. Please help..
You could add a new column to your table called modifiedOn. Then whenever you insert or update a row in that table, set the modifiedOn column to the current time.
This will allow you to use INSERT ... ON DUPLICATE KEY UPDATE ... and then after you have inserted all of your new tags, simply delete the tags for the given item ID that were modified before the time you used in your insert queries. You should probably also do this inside of a transaction.
To prevent duplicate tags you need another table for the many to many relationship
create table `items_tags` (
`item_id` int(10) unsigned not null default 0,
`tag_id` int(10) unsigned not null default 0,
primary key (`post_id`, `tag_id`),
)
create table `tags` (
`id` int(10) unsigned not null auto_increment,
`tag` varchar(100) not null default '',
primary key (`id`),
unique key (`tag`)
)
ps - an unsigned int is only 10 characters in length, not 11.
I'm not gonna type out all the code for you but when the form is submitted you "INSERT IGNORE" on the tags table, select the id's where tag IN( ), delete from items_tags where item_id = and finally insert into items_tags a record for each tag id with the current item id.
I suggest you start learning pdo. It make things much easier. You should also have foreign key constraints with cascade on delete so when an item or tag is deleted it's references in the items_tags table are also deleted.
To retrieve all tags for a post is simple:
select i.*, t.* from items as i left join items_tags as it on (it.item_id = i.id) left join tags as t on (it.tag_id = t.id) where i.id = <current item id>
Hope that helps.
You could just
DELETE FROM `tags` WHERE `item_id` = $itemid;
before you run your loop.
NB: I'd look at escaping (at the least) or switching to stored procedures (best case) to secure your application against the risk of SQL injection.

Categories