I have this query:
$relevantwords = {"one" , "two" , "three" } ;
foreach ($relevantwords as $word)
{
$query .= "SELECT * FROM delhi_items WHERE heading like '%$word%' AND id!={$entry_row['id']} AND visible=1 UNION ALL " ;
}
$query = implode( " " , explode(" " , $query , -3) ) ;
$query .= " ORDER BY time_stamp DESC LIMIT 0, 20 " ;
$result_set = mysql_query($query, $connection);
This causes several duplicates in my resultset. Is there a way to detect and remove these duplicates from the resultset ? I know I should probably try to avoid the duplicates in the first place, but I am unable to figure that out.
Also I tried distinct keyword, it didn't work (because its a loop, the same entry is fetched again and again).
Laslty I am kind of an amateur so please tell me if I am doing something fundamentally uncool with such a long sql query in a for loop.
Thanks
This is not the right way to do this query; don't use UNION ALL with several queries.
Just use one query and use an OR between the relevant WHERE clause parts. It'll select each row just once, regardless of how many bits it matches.
I would try to have one SELECT and no UNION and DISTINCT. It will probably be a faster query:
$relevantwords = {"one" , "two" , "three" } ;
$querycondition = "" ;
foreach ($relevantwords as $word)
{
$querycondition .= " heading LIKE '%$word%' OR"
}
$querycondition = substr($querycondition ,0 ,strlen($querycondition)-2 ) ;
$query = " SELECT * "
. " FROM delhi_items "
. " WHERE ( "
. $querycondition
. " ) "
. " AND id!={$entry_row['id']} "
. " AND visible=1 "
. " ORDER BY time_stamp DESC "
. " LIMIT 0, 20 " ;
$result_set = mysql_query($query, $connection);
Use DISTINCT:
SELECT DISTINCT * FROM ...
UPDATE:
Actually DISTINCT doesn't work since the un-duping would happen before the records are merged with each other in the UNION ALL.
To do what you are trying to do, you want the SELECT DISTINCT * to happen outside the union-ing of all the records.
Do the selecting and union-ing inside a derived table:
SELECT DISTINCT * FROM
(SELECT * FROM TABLE WHERE ...
UNION ALL
SELECT * FROM TABLE WHERE ...
) t
ORDER BY ...
Related
I am new in CodeIgniter, I want to count all rows from database table but i use limit in query and i want all count without use limit how can i do ?
my code is below :
$sql = " SELECT intGlCode,fkCategoryGlCode,'C' as acctyp,varEmail,varContactNo as phone,CONCAT(varFirstName,' ',varLastName) as name,dtCreateDate,chrStatus,varMessage as message
FROM " . DB_PREFIX . "Customer WHERE varEmail='$userEmail'
UNION
SELECT intGlCode,'' as fkCategoryGlCode,'P' as acctyp,varEmail,varPhoneNo as phone,varName as name,dtCreateDate,chrStatus,txtDescription as message FROM
" . DB_PREFIX . "Power WHERE varEmail='$userEmail' ORDER BY intGlCode DESC
LIMIT $start, $per_page ";
$query = $this->db->query($sql)
i use limit for pagination but i want to get all record from table.
You can add new column in both above and below UNION queries. It will be like below.
select (select count(*) from your_query), your_columns from query_above_union
UNION
select (select count(*) from your_query), your_columns from query_below_union
your_query = your full actual query your are using currently.
Although I am not sure about Codeigniter. But sure about SQl.
* If you count all records with all data including limit, than you can use this code. please check it. I hope it will works for you.*
$countsql = " SELECT intGlCode,fkCategoryGlCode,'C' as acctyp,varEmail,varContactNo as phone,CONCAT(varFirstName,' ',varLastName) as name,dtCreateDate,chrStatus,varMessage as message
FROM " . DB_PREFIX . "Customer WHERE varEmail='$userEmail'
UNION
SELECT intGlCode,'' as fkCategoryGlCode,'P' as acctyp,varEmail,varPhoneNo as phone,varName as name,dtCreateDate,chrStatus,txtDescription as message FROM
" . DB_PREFIX . "Power WHERE varEmail='$userEmail' ORDER BY intGlCode DESC";
$sql = $countsql. " LIMIT $start, $per_page";
$totalRecords = $this->db->query($countsql);
$result["total_rows"] = $totalRecords->num_rows();
$query = $this->db->query($sql);
$result["list"] = $query->result_array();
SQLSTATE[42804]: Datatype mismatch: 7 ERROR: argument of WHERE must be
type boolean, not type integer LINE 1
$sql = "SELECT mod_modulegroupcode, mod_modulegroupname FROM module "
. " WHERE 1 GROUP BY `mod_modulegroupcode` "
. " ORDER BY `mod_modulegrouporder` ASC, `mod_moduleorder` ASC ";
$stmt = $DB->prepare($sql);
$stmt->execute();
$commonModules = $stmt->fetchAll();
$sql = "SELECT mod_modulegroupcode, mod_modulegroupname, mod_modulepagename, mod_modulecode, mod_modulename FROM module "
. " WHERE 1 "
. " ORDER BY `mod_modulegrouporder` ASC, `mod_moduleorder` ASC ";
$stmt = $DB->prepare($sql);
$stmt->execute();
$allModules = $stmt->fetchAll();
$sql = "SELECT rr_modulecode, rr_create, rr_edit, rr_delete, rr_view FROM role_rights "
. " WHERE rr_rolecode = :rc "
. " ORDER BY `rr_modulecode` ASC ";
$stmt = $DB->prepare($sql);
$stmt->bindValue(":rc", $_SESSION["rolecode"]);
$stmt->execute();
$userRights = $stmt->fetchAll();
You are getting a datatype mismatch error because the where clause of a SQL statement expects you to provide conditions, not integer values. Your third query has a condition for its where clause, but your first two try to just give an integer value. There are some programming languages where 1 treated like "true", but SQL is not one of those languages.
Given that you're assembling the SQL with your code, if there's no condition to provide for the where clause, then just leave the where clause out, like this:
$sql = "SELECT mod_modulegroupcode, mod_modulegroupname FROM module "
. " GROUP BY `mod_modulegroupcode` "
. " ORDER BY `mod_modulegrouporder` ASC, `mod_moduleorder` ASC ";
But then you're going to have a problem because mod_modulegroupname is not aggregated. Since you're not doing any aggregation, I suggest just taking the GROUP BY clause out, too. You should also take the backticks out of your order by. This would leave you with:
$sql = "SELECT mod_modulegroupcode, mod_modulegroupname FROM module "
. " ORDER BY mod_modulegrouporder ASC, mod_moduleorder ASC ";
If I can kindly say so, it sounds like you should do a SQL tutorial.
I have some code which generates a MySQL query string called $query:
$query = "select * from Surveys where surveylayoutid='$surveyid' and customerid='" . $_SESSION['login_customerid'] . "' and (";
$clue = $_POST['postcode'];
$onwhat="Postcode";
$query .= $onwhat . " like '%$clue%') order by id desc";
$result = mysql_query($query, $connection) or die(mysql_error());
This returns something like:
select * from Surveys where surveylayoutid='12' and customerid='1' and (Postcode like '%dn%') order by id desc
which works fine. I've then altered the code because I want to search on more fields so it now reads:
$remap = array("Postcode", "Street", "HouseNum", "District", "Town");
$query = "select * from Surveys where surveylayoutid='$surveyid' and customerid='" . $_SESSION['login_customerid'] . "' and (";
for ($i=0; $i<=4; $i++) {
if ($_POST[strtolower($remap[$i])]!="") {
$clue = $_POST[strtolower($remap[$i])];
$query .= $remap[$i] . " like '%$clue%') order by id desc";
break;
}
}
This also returns:
select * from Surveys where surveylayoutid='12' and customerid='1' and (Postcode like '%dn%') order by id desc
which on the face of it is identical but it generates this error:
You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'like '%dn%' order by id desc' at line 1
In both cases $query contains the same "text" but for some reason isn't treated as a valid MySQL query in the updated code, can anyone tell me why?
One possible problem could be the interpretation of the content here.
If you use:
$query .= $remap[$i] . " like '%$clue%') order by id desc";
All that is inside "" gets to be interpreted. Thus there could be unwanted side effects that you don't see at first glance and can explain what is happening. To avoid this it would have to be changed to:
$query .= $remap[$i] . ' like ' . "'" . '%' . $clue . '%' . "') order by id desc";
Even though more clunky in terms of how big it is, it makes sure that $lue and also the % are not interpreted as all in between ' ' is not interpreted.
See if this help you solve your problem?
$remap = array(
"Postcode",
"Street",
"HouseNum",
"District",
"Town"
);
for ($i = 0; $i <= 4; $i++)
{
if ($_POST[strtolower($remap[$i]) ] != "")
{
$query = "select * from Surveys where surveylayoutid='12' and customerid='1' and (";
$clue = $_POST[strtolower($remap[$i]) ];
$query.= $remap[$i] . " like '%$clue%') order by id desc";
$query_done[] = $query;
unset($query);
$result = mysql_query($query_done[$i], $connection) or die(mysql_error());
// Display your result here
}
}
I tried changing your code abit, and it seems the result is something like this
select * from Surveys where surveylayoutid='12' and customerid='1' and (Postcode like '%Postcode%') order by id descselect * from Surveys where surveylayoutid='12' and customerid='1' and (Street like '%Street%') order by id descselect * from Surveys where surveylayoutid='12' and customerid='1' and (HouseNum like '%HouseNum%') order by id descselect * from Surveys where surveylayoutid='12' and customerid='1' and (District like '%District%') order by id descselect * from Surveys where surveylayoutid='12' and customerid='1' and (Town like '%Town%') order by id desc
I have the following method to create a search query that scores the search term by occurrence:
public function findAll($search, array $data = []) {
$query = DB::query("
SELECT
SQL_CALC_FOUND_ROWS
*,
SUM(MATCH(text) AGAINST('{$search}' IN BOOLEAN MODE)) as score
FROM " . DB::prefix() . "search_index
WHERE MATCH(text) AGAINST('{$search}' IN BOOLEAN MODE)
OR text LIKE '%{$search}%'
GROUP BY language_id, type, object_id
ORDER BY score DESC
LIMIT " . (int)$data['start'] . ", " . (int)$data['limit'] . "
");
$count = DB::query("SELECT FOUND_ROWS() AS total");
return [
'count' => (int)$count->row['total'],
'query' => $query->rows
];
}
This works beautifully, but I need to add in a query to the tag table so that it will increment the score and add items to the array for content that may not already exist in the main query:
$tags = DB::query("
SELECT * FROM " . DB::prefix() . "tag
WHERE tag = '{$search}'
");
Any ideas on how to implement this into the main query so that I'm only executing a single query?
I was able to combine these with a UNION query:
$query = DB::query("
(SELECT
SQL_CALC_FOUND_ROWS
type, object_id, language_id,
SUM(MATCH(text) AGAINST('{$search}' IN BOOLEAN MODE)) as score
FROM " . DB::prefix() . "search_index
WHERE MATCH(text) AGAINST('{$search}' IN BOOLEAN MODE)
OR text LIKE '%{$search}%'
GROUP BY language_id, type, object_id
ORDER BY score DESC
LIMIT " . (int)$data['start'] . ", " . (int)$data['limit'] . ")
UNION DISTINCT
(SELECT
section, element_id, language_id, tag
FROM " . DB::prefix() . "tag
WHERE tag = '{$search}')
");
I'm using Mysql 5.5.16. I've a query where i combine 3 queries using union opeartor... Each query contains more than 10000 records.
My Query looks like...
$search_qry1 = " (SELECT PC_name as name FROM profile_category_tbl WHERE PC_status=1 and PC_parentid!=0 and (";
$search_qry2 = " (SELECT PROFKEY_name as name FROM profile_keywords_tbl WHERE PROFKEY_status=1 and (";
$search_qry3 = " (SELECT COM_name as name FROM ".$C."_company_profile_tbl WHERE COM_status=1 and (";
$order_by = "ORDER BY";
$word_length = explode(" ", $my_data);
for($i=0; $i <= count($word_length) - 1; $i++) {
$dt = $word_length[$i];
if ($dt) {
if ($i == 0) {
$or="";
}
else {
$or="OR";
}
$search_qry1 .= " $or regex_replace('[^a-zA-Z0-9\-]','',remove_specialCharacter(PC_name)) LIKE '%$dt%' ";
$search_qry2 .= " $or regex_replace('[^a-zA-Z0-9\-]','',remove_specialCharacter(PROFKEY_name)) LIKE '%$dt%' ";
$search_qry3 .= " $or regex_replace('[^a-zA-Z0-9\-]','',remove_specialCharacter(COM_name)) LIKE '%$dt%' ";
$order_by .= " IF(name LIKE '%$dt%',1,0) ";
}
if ($i == count($word_length) - 1) {
$search_qry1 .= ")";
$search_qry2 .= ")";
$search_qry3 .= ")";
}
if ($i != count($word_length) - 1) {
$order_by .= "+ ";
}
}
$search_qry1 .= " GROUP BY PC_name )";
$search_qry2 .= " GROUP BY PROFKEY_name )";
$search_qry3 .= " GROUP BY COM_name )";
$search_qry = "select name from ( $search_qry1 UNION ALL $search_qry2 UNION ALL $search_qry3 )a $order_by desc LIMIT 0,25";
It works perfectly... but it takes the time to execute it - more than 4 secs... for each search.... how possible to reduce its execution time?.... If anyone know an idea about this please let me know....
Actually I run this query for auto-complete search. If type "Rest" in search box, then
The Output Should be,
Family Restaurants
Everest Park
Fast Food Restaurants
Everest Park Residency
Fish Restaurant
Everest Power Solution
Fish Restaurants
Thanks in Advance,
Jeni
You are using regex_replace and like "%foobar%",
that actually slows down your query significat, since mySql has to do a fulltext search and the regex operation on every single row. Maybe it's faster if you do these operations after you fetched your results, using php instead of mysql.
The main problem is that you're doing a full scan of each table to check the like '%$dt%' condition. No matter which indices you have on the table, none of them will be used.
But if you post the final SQL sentence and what are you trying to do maybe we could help.