Recursive query in MySQL( find direct and indirect supervisor of Bob )? - mysql

There is one problem that I encountered and that can we solved using recursive query but my code is giving error in MySQL.
Problem is as follow:
person supervisor
Mary Susan
David Mary
Bob Alice
Alice David
Find all supervisor(direct and indirect) of Bob?
Here is my solution for above question:
create recursive view find_all(person,supervisor,depth) as
(
select person,supervisor,0 as depth from employee as e
where e.person = 'Bob'
union all
select in1.person,out1.supervisor,in1.depth+1
from find_all as in1,employee as out1
and in1.supervisor = out1.person
and in1.depth<=100
);
And for output:
select * from find_all;

Here's one approach. Just create a normal view. The view contains the recursive WITH clause.
CREATE VIEW find_all(person,supervisor,depth) AS
WITH RECURSIVE cte1 AS (
SELECT person, supervisor, 0 AS depth
FROM employee AS e
WHERE e.person = 'Bob'
UNION ALL
SELECT in1.person, out1.supervisor, in1.depth+1
FROM cte1 AS in1, employee AS out1
WHERE in1.supervisor = out1.person
AND in1.depth <= 100
)
SELECT * FROM cte1
;
Result:
select * from find_all;
+--------+------------+-------+
| person | supervisor | depth |
+--------+------------+-------+
| Bob | Alice | 0 |
| Bob | David | 1 |
| Bob | Mary | 2 |
| Bob | Susan | 3 |
+--------+------------+-------+
Alternate form:
The <from clause> contains a <table reference list> which is a comma separated list of <table reference>, <table reference>, .... This produces cross join behavior. We then use a WHERE clause to effectively inner join these table references.
It's commonly suggested to use a single <table reference>, for several good reasons, which can contain a <joined table>. The second UNION term below contains an example of a <joined table>.
A couple of advantages are ease of identifying (and not accidentally forgetting) the corresponding <join condition> and easier to change between join types.
CREATE VIEW find_all(person,supervisor,depth) AS
WITH RECURSIVE cte1 AS (
SELECT person, supervisor, 0 AS depth
FROM employee AS e
WHERE e.person = 'Bob'
UNION ALL
SELECT in1.person, out1.supervisor, in1.depth+1
FROM cte1 AS in1
JOIN employee AS out1
ON in1.supervisor = out1.person
WHERE in1.depth <= 100
)
SELECT * FROM cte1
;
In this case, the depth logic is not really part of the <join condition>. It's really a subsequent filter of the rows produced by the join.
The difference is seen when, for example, we have a left outer join. We don't want that logic to just generate more nulls for the right table columns. We want that logic to produce no further rows when depth is exceeded.

Related

mysql "and" logic within result set

Say I have a data set like the following:
table foo
id | employeeType | employeeID
-------------------------
1 | Developer | 1
2 | Developer | 2
3 | Developer | 3
4 | Manager | 1
5 | Manager | 4
6 | Manager | 5
7 | CEO | 1
8 | CEO | 6
and I wanted to run a query that would return all the employeeids (along with the employeeTypes) where there is a common employee id between all employeeTypes (that's the 'and' logic. ONly employeeIDs that have all employeeTypes will return. employeeType = Developer and employeeType=Manager and employeeType=CEO). For the data above the example output would be
result table
id | employeeType | employeeID
-------------------------
1 | Developer | 1
4 | Manager | 1
7 | CEO | 1
I was able to do this when I only had only TWO employeeTypes by self joining the table like this.
select * from foo as fooOne
join foo as fooTwo
on fooOne.employeeID = fooTwo.employeeID
AND
fooOne.employeeType <> fooTwo.employeeType
that query returns a result set with values from fooTwo when the 'and' logic matches, but again, only for two types of employees. My real use case scenario dictates that I need to be able to handle a variable number of employeeTypes (3, 4, 5, etc...)
Any thoughts on this would be greatly appreciated.
This should return the rows that you want:
SELECT foo.*
FROM
foo
WHERE
employeeID IN (
SELECT employeeID
FROM foo
GROUP BY employeeID
HAVING COUNT(DISTINCT employeeType) =
(SELECT COUNT(DISTINCT employeeType)
FROM foo)
)
Please see a fiddle here.
The inner query will return the number of distinct employee types:
(SELECT COUNT(DISTINCT employeeType) FROM foo)
The middle query will return all the employee IDs that have the maximum number of employee types:
SELECT employeeID
FROM foo
GROUP BY employeeID
HAVING COUNT(DISTINCT employeeType) =
(SELECT COUNT(DISTINCT employeeType) FROM foo)
and the outer query will return the whole rows.
You can try a subquery to make it dynamic
SELECT employeeID, employeeType
FROM foo
WHERE employeeID IN (
SELECT employeeID
FROM foo
GROUP BY employeeID
HAVING COUNT(DISTINCT employeeType) = (SELECT COUNT(DISTINCT employeeType) FROM foo)
)
I agree that this might be looked down as a very inefficient/hacky way of doing things, but this should still get the job done. And frankly, I can't see any other way out of this.
SELECT * FROM (
SELECT EMPLOYEE_ID, GROUP_CONCAT(DISTINCT EmployeeType ORDER BY EmployeeType) AS Roles
FROM EMPLOYEES GROUP BY EMPLOYEE_ID
) EMPLOYEE_ROLES
WHERE EMPLOYEE_ROLES.Roles = 'CEO,Developer,Manager';
Note that the comma separated list of roles provided in the end is in the alphabetical order.

SQL "chained" queries?

I have two tables.
I am a total newbie to SQL. Using mysql at the moment.
I have the following setup for a school-related db:
Table A contains students records.
Student's id, password,name, lastname and so on.
Table B contains class attendancy records.
Each record goes like this: date, student id, grade
I need to gather all the student info of students that attended classes in a certain date range.
Now, the stupid way would be
first I SELECT all classes from Table B with DATE IN BETWEEN the range
then for each class, I get the student id and SELECT * FROM students WHERE id = student id
What I can't wrap my mind around is the smart way.
How to do this in one query only.
I am failing at understanding the concepts of JOIN, UNION and so on...
my best guess so far is
SELECT students.id, students.name, students.lastname
FROM students, classes
WHERE classes.date BETWEEN 20140101 AND 20150101
AND
classes.studentid = students.id
but is this the appropriate way for this case?
Dont add the join statement in the where clause. Do it like this:
SELECT s.id, s.name, s.lastname,c.date,c.grade
FROM classes c
inner join students s
on c.studentid=s.id
WHERE c.date BETWEEN '01/01/2014' AND '01/01/2015'
This sounds like an assignment so I will attempt to describe the problem and give a hint to the solution.
An example of a union would be;
SELECT students.name, students.lastname
FROM students
WHERE students.lastname IS NOT NULL
UNION
SELECT students.name, 'N/A'
FROM students
WHERE students.lastname IS NULL;
+--------------+--------------+
| name | lastname |
+--------------+--------------+
| John | Doe | <- First two rows came from first query
| Jill | Smith |
| Bill | N/A | <- This came from the second query
+--------------+--------------+
The usual use case for a union is to display the same columns, but munge the data in a different way - otherwise you can usually achieve similar results through a WHERE clause.
An example of a join would be;
SELECT authors.id, authors.name, books.title
FROM authors LEFT JOIN books ON authors.id = books.authors_id
+--------------+--------------+------------------+
| id | name | title |
+--------------+--------------+------------------+
| 1 | Mark Twain | Huckleberry Fin. |
| 2 | Terry Prat.. | Good Omens |
+--------------+--------------+------------------+
^ First two columns from ^ Third column appended
from authors table from books table linked
by "author id"
Think of a join as appending columns to your results, a union is appending rows with the same columns.
In your situation we can rule out a union as you don't want to append more student rows, you want class and student information side by side.

SQL - select rows that have the same value in two columns

The solution to the topic is evading me.
I have a table looking like (beyond other fields that have nothing to do with my question):
NAME,CARDNUMBER,MEMBERTYPE
Now, I want a view that shows rows where the cardnumber AND membertype is identical. Both of these fields are integers. Name is VARCHAR. Name is not unique, and duplicate cardnumber, membertype should show for the same name, as well.
I.e. if the following was the table:
JOHN | 324 | 2
PETER | 642 | 1
MARK | 324 | 2
DIANNA | 753 | 2
SPIDERMAN | 642 | 1
JAMIE FOXX | 235 | 6
I would want:
JOHN | 324 | 2
MARK | 324 | 2
PETER | 642 | 1
SPIDERMAN | 642 | 1
this could just be sorted by cardnumber to make it useful to humans.
What's the most efficient way of doing this?
What's the most efficient way of doing this?
I believe a JOIN will be more efficient than EXISTS
SELECT t1.* FROM myTable t1
JOIN (
SELECT cardnumber, membertype
FROM myTable
GROUP BY cardnumber, membertype
HAVING COUNT(*) > 1
) t2 ON t1.cardnumber = t2.cardnumber AND t1.membertype = t2.membertype
Query plan: http://www.sqlfiddle.com/#!2/0abe3/1
You can use exists for this:
select *
from yourtable y
where exists (
select 1
from yourtable y2
where y.name <> y2.name
and y.cardnumber = y2.cardnumber
and y.membertype = y2.membertype)
SQL Fiddle Demo
Since you mentioned names can be duplicated, and that a duplicate name still means is a different person and should show up in the result set, we need to use a GROUP BY HAVING COUNT(*) > 1 in order to truly detect dupes. Then join this back to the main table to get your full result list.
Also since from your comments, it sounds like you are wrapping this into a view, you'll need to separate out the subquery.
CREATE VIEW DUP_CARDS
AS
SELECT CARDNUMBER, MEMBERTYPE
FROM mytable t2
GROUP BY CARDNUMBER, MEMBERTYPE
HAVING COUNT(*) > 1
CREATE VIEW DUP_ROWS
AS
SELECT t1.*
FROM mytable AS t1
INNER JOIN DUP_CARDS AS DUP
ON (T1.CARDNUMBER = DUP.CARDNUMBER AND T1.MEMBERTYPE = DUP.MEMBERTYPE )
SQL Fiddle Example
If you just need to know the valuepairs of the 3 fields that are not unique then you could simply do:
SELECT concat(NAME, "|", CARDNUMBER, "|", MEMBERTYPE) AS myIdentifier,
COUNT(*) AS count
FROM myTable
GROUP BY myIdentifier
HAVING count > 1
This will give you all the different pairs of NAME, CARDNUMBER and MEMBERTYPE that are used more than once with a count (how many times they are duplicated). This doesnt give you back the entries, you would have to do that in a second step.

MySQL join query or sub query

I'm trying to do a query that selects mike if it isn't in the three highest bids for a keyword. Rows 4 and 7 should be selected.
So in final, if mike isn't in the three highest bids for a keyword, then select.
How do I solve this? With a sub query?
$construct = "SELECT child.* FROM `temp-advertise` child
LEFT JOIN `temp-advertise` parent on child.keyword=parent.keyword
WHERE child.name='mike'
ORDER BY child.id DESC";
id | name| keyword | bid |
1 | mike| one | 7 |
2 | tom | one | 4 |
3 | ced | one | 6 |
4 | mike| two | 1 |
5 | tom | two | 5 |
6 | har | two | 5 |
7 | mike| one | 3 |
8 | har | two | 3 |
SELECT *
FROM `temp-advertise` ta
WHERE ta.keyword = 'one'
AND ta.name = 'mike'
AND ta.bid <
(
SELECT bid
FROM `temp-advertise` tai
WHERE tai.keyword = 'one'
ORDER BY
bid DESC
LIMIT 2, 1
)
Your structure doesn't look too promising, nor your sample data. However, that said, you want to know if "Mike" was in the top 3 per keyword... and that he has 3 bids.... 2 for "one", 1 for "two". From the raw data, it looks like Mike is in 1st place and 4th place for the "one" keyword, and 4th place for "two" keyword.
This should get you what you need with SOME respect to not doing a full query of all keywords. The first innermost query is to just get keywords bid on by "mike" (hence alias "JustMike"). Then join that to the temp-advertise on ONLY THOSE keywords.
Next, by using MySQL variables, we can keep track of the rank PER KEYWORD. The trick is the ORDER BY clause needs to return them in the order that represents proper ranking. In this case, each keyword first, then within each keyword, ordered by highest bid first.
By querying the records, then using the #variables, we increase the counter, start at 1 every time the keyword changes, then preserve the keyword into the #grpKeyword variable for comparison of the next record. Once ALL bids are processed for the respective keywords, it then queries THAT result but ONLY for those bid on by "mike". These records will have whatever his rank position was.
select RankPerKeyword.*
from
( SELECT ta.*,
#grpCnt := if( #grpKeyword = ta.Keyword, #grpCnt +1, 1 ) as KWRank,
#grpKeyword := ta.Keyword as carryForward
FROM
( select distinct ta1.keyword
from `temp-advertise` ta1
where ta1.name = "mike" ) as JustMike
JOIN `temp-advertise` ta
on JustMike.Keyword = ta.Keyword,
( select #grpCnt := 0,
#grpKeyword := '' ) SqlVars
ORDER BY
ta.Keyword,
ta.Bid DESC" ) RankPerKeyword
where
RankPerKeyword.name = "mike"
(Run above to just preview the results... should show 3 records)
So, if you want to know if it was WITHIN the top 3 for a keyword you could just change to
select RankPerKeyword.keyword, MIN( RankPerKeyword.KWRank ) as BestRank
from (rest of query)
group by RankPerKeyword.Keyword
Try this:
Select ID, name, keyword from temp-advertise e
where 3 <= (select count(name) from temp-advertise
where e.keyword = keyword and bid > e.bid)
Try
SELECT .. ORDER BY bid LIMIT 3,999

MYSQL join results set wiped results during IN () in where clause?

Editted heavily!
The original question was based on a misunderstanding of how IN() treats a column from a results set from a join. I thought IN( some_join.some_column ) would treat a results column as a list and loop through each row in place. It turns out it only looks at the first row.
So, the adapted question: Is there anything in MySQL that can loop through a column of results from a join from a WHERE clause?
Here's the super-simplified code I'm working with, stripped down from a complex crm search function. The left join and general idea are relics from that query. So for this query, it has to be an exclusive search - finding people with ALL specified tags, not just any.
First the DB
Table 1: Person
+----+------+
| id | name |
+----+------+
| 1 | Bob |
| 2 | Jill |
+----+------+
Table 2: Tag
+-----------+--------+
| person_id | tag_id |
+-----------+--------+
| 1 | 1 |
| 1 | 2 |
| 2 | 2 |
| 2 | 3 |
+-----------+--------+
Nice and simple. So, naturally:
SELECT name, GROUP_CONCAT(tag.tag_id) FROM person LEFT JOIN tag ON person.id = tag.person_id GROUP BY name;
+------+--------------------------+
| name | GROUP_CONCAT(tag.tag_id) |
+------+--------------------------+
| Bob | 1,2 |
| Jill | 2,3 |
+------+--------------------------+
So far so good. So what I'm looking for is something that would find only Bob in the first case and only Jill in the second - without using HAVING COUNT(DISTINCT ...) because that doesn't work in the broader query (there's a seperate tags inheritance cache and a ton of other stuff).
Here's my original sample queries - based on the false idea that IN() would loop through all rows at once.
SELECT DISTINCT name FROM person LEFT JOIN tag ON person.id = tag.person_id
WHERE ( ( 1 IN (tag.tag_id) ) AND ( 2 IN (tag.tag_id) ) );
Empty set (0.00 sec)
SELECT DISTINCT name FROM person LEFT JOIN tag ON person.id = tag.person_id
WHERE ( ( 2 IN (tag.tag_id) ) AND ( 3 IN (tag.tag_id) ) );
Empty set (0.00 sec)
Here's my new latest failed attempt to give an idea of what I'm aiming for...
SELECT name, GROUP_CONCAT(tag.tag_id) FROM person LEFT JOIN tag ON person.id = tag.person_id
GROUP BY person.id HAVING ( ( 1 IN (GROUP_CONCAT(tag.tag_id) ) ) ) AND ( 2 IN (GROUP_CONCAT(tag.tag_id)) );
Empty set (0.00 sec)
So it seems it's taking a GROUP_CONCAT string, of either 1,2 or 2,3, and is treating it as a single entity rather than an expression list. Is there any way to turn a grouped column into an expression list that IN () or =ANY() will treat as a list?
Essentially, I'm trying to make IN() loop iteratively over something that resembles an array or a dynamic expression list, which contains all the rows of data that come from a join.
Think about what your code is doing logically:
( 1 IN (tag.tag_id) ) AND ( 2 IN (tag.tag_id) )
is equivalent to
( 1 = (tag.tag_id) ) AND (2 = (tag.tag_id) )
There's no way tag.tag_id can satisfy both conditions at the same time, so the AND is never true.
It looks like the OR version you cited in your question is the one you really want:
SELECT DISTINCT name FROM person LEFT JOIN tag ON person.id = tag.person_id
WHERE ( ( 1 IN (tag.tag_id) ) OR ( 2 IN (tag.tag_id) ) );
Using the IN clause more appropriately, you could write that as:
SELECT DISTINCT name FROM person LEFT JOIN tag ON person.id = tag.person_id
WHERE tag.tag_id in (1,2);
One final note, because you're referencing a column from the LEFT JOINed table in your WHERE clause (tag.tag_id), you're really forcing that to behave like an INNER JOIN. To truly get a LEFT JOIN, you'd need to move the criteria out of the WHERE and make it part of the JOIN conditions instead:
SELECT DISTINCT name FROM person LEFT JOIN tag ON person.id = tag.person_id
AND tag.tag_id in (1,2);
WHERE ( ( 1 IN (tag.tag_id) ) AND ( 2 IN (tag.tag_id) ) );
This will never return any results since tag.tag_id cannot be 1 and 2 at the same time.
Additionally is there a reason you're using 1 IN (blah) rather than blah = 1?