Database

Re: Update Advice - Getting most recent value

21-Feb-07 2:25

Hi Brad

Normally you would use a trigger to keep the two tables synchronised. If the tables are currently out of sync then the following should fix the master table:

update DocMaster set
   DocStatus = A.DocStatus,
   UpdateTime = A.UpdateTime
  from DocUpdate A
  where A.DocId = DocMaster.DocId
  and A.UpdateTime = (select max(UpdateTime) from DocUpdate
     where DocUpdate.DocId = DocMaster.DocId)

This assumes that the primary key on DocMaster is named "DocId", and that the "UpdateTime" is unique for each document (if you have an identity column on DocUpdate then you should use that instead).

Hope that helps.
Andy

Brady Kelly21-Feb-07 2:58

Brady Kelly

21-Feb-07 2:58

That's perfect, thanks Andy. I do have triggers doing the normal sync work, but whenever I create redundant or derived fields like these I always start with an update procedure that makes everything all good again, if something goes wrong.

some question

Imran Khan Pathan21-Feb-07 1:04

Imran Khan Pathan

21-Feb-07 1:04

Hi friends

What is a clustered index? How many records can take clustered index in sql server?
what is a cursor?
what is a sql injuction?

thanks in advance

khan

Re: some question

andyharman21-Feb-07 2:38

Re: Average of all columns

21-Feb-07 2:38

imrankhanpathan wrote:
What is a clustered index?

You will find the answer in SQL-Server's "Books Online". A table can only have one clustered index defined (unless told otherwise SQL-Server will use a clustered index for primary keys). There is no practical limit on the number of records that can be held in a clustered index.

imrankhanpathan wrote:
what is a cursor?

Again the answer is in "Books Online".

imrankhanpathan wrote:
what is a sql injuction?

You can find out about "SQL Injection" here[^]

Average of all columns

Sam Heller20-Feb-07 11:11

Sam Heller

20-Feb-07 11:11

I have a single query that selects a product id. Everytime is finds a product id it also selects several other queries to build up a row of statistics. Therefore my query retrieves rows as follows:

Product ID | Stat Number One | Stat Number Two | More Stats
125 | 76% | 45% |

What I want is a final overall score at the end of the query in a final column. Built on the fly from the other columns. How can I do this?

Eg
(StatNumberOne + Stat Number Two + Rest of Stat Columns / Amount of Columns) As Overall Score

I don't know how to reference these columns on the fly like this ??

Krish - KP20-Feb-07 17:01

Krish - KP

20-Feb-07 17:01

one way doing this is using union
Eg.
SELECT prodid, no1, no2, 0 as overall from ......
union
SELECT prodid, 0 as no1, 0 as no2, (no1 + no2) / x as overall from ......

other way of doing is use procedure

Problem in Using ADO in C++ to Get Oracle Ref Cursor Data

Kelvin X Li20-Feb-07 11:02

Kelvin X Li

20-Feb-07 11:02

I am trying to call a Oracle stored procedure using ADO in C++.

The store procedure has 12 input parameters, and 2 output
parameters with one of them being a Ref Cursor. I followed the
example in a post by Koushik Biswas on the subject
(http://www.codeproject.com/script/Articles/list_articles.asp?userid=811562).

But when I called
pRecordset = pCommand->Execute(NULL, NULL,
adCmdStoredProc | adCmdUnspecified );

I got an error message of "Unspecified error".

I don't have visibility to the stored procedure, but using describe
from sqlplus, I got the following:

SQL> describe myproc
Parameter Type Mode Default?
----------------------- ---------- ---- --------
P_1 VARCHAR2 IN
P_2 NUMBER IN
P_3 NUMBER IN
P_4 NUMBER IN
P_5 NUMBER IN
P_6 NUMBER IN
P_7 NUMBER IN
P_8 NUMBER IN
P_9 NUMBER IN
P_10 NUMBER IN
P_11 DATE IN
P_12 VARCHAR2 IN
P_OUT_ERROR_TXT VARCHAR2 OUT
P_OUT_CURSOR REF CURSOR OUT

Most of the parameters can be set to NULL.

Here is part of my code:

spCmdOracle->CommandType = adCmdStoredProc;
spCmdOracle->CommandText = "{CALL myproc(?,?,?,?,?,?,?,?,?,?,?,?,?)}";

_bstr_t strdata = "MYTEST";
_ParameterPtr p_1 = spCmdOracle->CreateParameter("@p_1",
adVarChar, adParamInput,
strdata.length(), strdata);
spCmdOracle->Parameters->Append(p_1);

_variant_t d((float)25);
_ParameterPtr p_2 = spCmdOracle->CreateParameter("@p_2",
adSingle, adParamInput,
sizeof(float), d);
spCmdOracle->Parameters->Append(p_2);

_variant_t a((float)40);
_ParameterPtr p_3 = spCmdOracle->CreateParameter("@p_3",
adSingle, adParamInput,
sizeof(float), a);
spCmdOracle->Parameters->Append(p_3);

_variant_t null_id;
_ParameterPtr p_4 = spCmdOracle->CreateParameter("@p_4",
adSingle, adParamInput,
sizeof(float), null_id);
spCmdOracle->Parameters->Append(p_4);

_ParameterPtr p_5 = spCmdOracle->CreateParameter("@p_5",
adSingle, adParamInput,
sizeof(float), null_id);
spCmdOracle->Parameters->Append(p_5);

_variant_t o((float)10000);
_ParameterPtr p_6 = spCmdOracle->CreateParameter("@p_6",
adSingle, adParamInput,
sizeof(float), o);
spCmdOracle->Parameters->Append(p_6);

_variant_t p((float)20000);
_ParameterPtr p_7 = spCmdOracle->CreateParameter("@p_7",
adSingle, adParamInput,
sizeof(float), p);
spCmdOracle->Parameters->Append(p_7);

_ParameterPtr p_8 = spCmdOracle->CreateParameter("@p_8",
adSingle, adParamInput,
sizeof(float), null_id);
spCmdOracle->Parameters->Append(p_8);

_ParameterPtr p_9 = spCmdOracle->CreateParameter("@p_9",
adSingle, adParamInput,
sizeof(float), null_id);
spCmdOracle->Parameters->Append(p_9);

_ParameterPtr p_10 = spCmdOracle->CreateParameter("@p_10",
adSingle, adParamInput,
sizeof(float), null_id);
spCmdOracle->Parameters->Append(p_10);

_ParameterPtr p_11 = spCmdOracle->CreateParameter("@p_11",
adSingle, adParamInput,
sizeof(float), null_id);
spCmdOracle->Parameters->Append(p_11);

_ParameterPtr p_12 = spCmdOracle->CreateParameter("@p_12",
adBSTR, adParamInput,
sizeof(float), null_id);
spCmdOracle->Parameters->Append(p_12);

char out_error[1000];
_ParameterPtr p_13 = spCmdOracle->CreateParameter("@P_OUT_ERROR_TXT",
adBSTR, adParamOutput,
sizeof(out_error), out_error);
spCmdOracle->Parameters->Append(p_out_error_txt);

spRsOracle = spCmdOracle->Execute(NULL, NULL, adCmdStoredProc | adCmdUnspecified );

The last line caused an exception with "Unspecified error".

If I change the setting of spCmdOracle->CommandText to use the stored procedure name
instead of the excape sequence, I got an error message PLS-00306:

Code meaning = IDispatch error #3092,
Description = ORA-06550: line 1, column 7:
PLS-00306: wrong number or types of arguments in call to 'MYPROC'
ORA-06550: line 1, column 7:
PL/SQL: Statement ignored

I am new on this subject and cannot find any help nearby. I would greatly appreciate any
here can give me some hint on what could be wrong.

Thanks.

Count Problem

ronivars20-Feb-07 9:14

20-Feb-07 9:14

hi there
i have 2 table with no relationship between them
and i need 2 extract from table1 the col1
and the phase of col1 in table2 col2
for example:

Table1
Col1
1000
1001
1002
=============================================
Table2
Col2
A1000A
bbb1000bcd
cdf1000frg
A1001a
bvc1001dcfe
=============================================
Result:
colRes count_in_table_2
1000 3
1001 2
1002 0

thanks in advance

roni vars

Mark J. Miller20-Feb-07 9:33

Mark J. Miller

20-Feb-07 9:33

The following will result in a table scan, but it will do the job:

SELECT T1.Col1, COUNT(*) FROM Table1 T1
INNER JOIN Table2 T2 ON T1.Col1 LIKE '%' + T2.Col2 + '%'

Another alternative would be to create a full-text index on Table2.Col2. It all depends on how much data we're talking about and how often and how busy your server is when you run the query.

ronivars20-Feb-07 10:15

20-Feb-07 10:15

hi there thaks for your quick answer
it's not working performance it's not issue
i work with small table
thanks again
any suggestion

roni vars

ronivars20-Feb-07 10:24

20-Feb-07 10:24

hi again

i think u forget the group by T1.Col1

SELECT T1.Col1, COUNT(*) FROM Table1 T1
INNER JOIN Table2 T2 ON T1.Col1 LIKE '%' + T2.Col2 + '%'

group by T1.Col1

it's still not working

roni vars

Mark J. Miller20-Feb-07 11:52

Mark J. Miller

20-Feb-07 11:52

Sorry,

I wrote that a little quickly. Yes, I forgot the GROUP BY, but the main problem was I reversed the references on the join D'Oh! | :doh:

. Try this (I tested it this time Blush | :O

):

SELECT T1.Col1, COUNT(*) FROM Table1 T1
INNER JOIN Table2 T2 ON T2.Col2 LIKE '%' + T1.Col1 + '%'
group by T1.Col1

ronivars20-Feb-07 12:16

20-Feb-07 12:16

hi Mark

i'm sorry but is still doesn't work i try both sql
and return me 0 rows

SELECT T1.Col, COUNT(*) FROM tbl1 T1
INNER JOIN tbl2 T2 ON T2.col LIKE '%' + T1.col + '%'
group by T1.Col

SELECT T1.Col, COUNT(*) FROM tbl1 T1
INNER JOIN tbl2 T2 ON T1.Col LIKE '%' + T2.Col + '%'
group by T1.Col

i also try this 2 sql sentences and is stil doesn't work
its return me
1000 0
1001 0
etc....

1)
select t1.col ,(SELECT count(*) FROM tbl2 t2
WHERE t1.col in (select t2.col from tbl2 t2))
From tbl1 t1
2)
select tbl1.col ,count(case when CHARINDEX(tbl1.col ,tbl2.col) > 0 then 1 else null end)
from tbl1,tbl2
group by tbl1.col

by the way i work with sql server 2005
thanks again
waiting for rescue

Mark J. Miller20-Feb-07 12:23

Mark J. Miller

20-Feb-07 12:23

I'm running 2005 also. This script:

IF OBJECT_ID('Table1') IS NOT NULL
DROP TABLE Table1
IF OBJECT_ID('Table2') IS NOT NULL
DROP TABLE Table2
go

CREATE TABLE Table1(Col1 Varchar(15))
CREATE TABLE Table2(Col2 VARCHAR(100))

INSERT INTO Table1 VALUES('1000')
INSERT INTO Table1 VALUES('1001')
INSERT INTO Table1 VALUES('1002')

INSERT INTO Table2 VALUES('A1000A')
INSERT INTO Table2 VALUES('bbb1000bcd')
INSERT INTO Table2 VALUES('cdf1000frg')
INSERT INTO Table2 VALUES('A1001a')
INSERT INTO Table2 VALUES('bvc1001dcfe')

SELECT T1.Col1, COUNT(*) FROM Table1 T1
INNER JOIN Table2 T2 ON T2.Col2 LIKE '%' + T1.Col1 + '%'
GROUP BY T1.Col1

Gives me these results:

1000 3
1001 2

If this isn't working for you, you'll need to give me your schema and what the results are (or the error messages).

ronivars20-Feb-07 12:50

20-Feb-07 12:50

thanks mark is work fine when the column type is varchar
but when the type is nchar it's doesnt work becuse it's fix length

roni vars

ronivars20-Feb-07 13:21

20-Feb-07 13:21

hi mark
You realy help in this Query
You are like Books Online
thanks again for your important help

and if it is nchar column type need to use ltrim rtrim function

SELECT T1.Col1, COUNT(*) FROM Table1 T1
INNER JOIN Table2 T2 ON ltrim(rtrim(T2.Col2)) LIKE '%' + ltrim(rtrim(T1.Col1)) + '%'
GROUP BY T1.Col1

regards

roni vars

ronivars20-Feb-07 12:22

20-Feb-07 12:22

sory my mistake i try this Sentence instead of first sentence i wrote before

select t1.col ,(SELECT count(*) FROM tbl2 t2 WHERE t1.col like '%'+ t2.col +'%') as count1
From tbl1 t1

roni vars

Count Query Prob

tadhg8820-Feb-07 6:39

20-Feb-07 6:39

Hi,
I am having some trouble with an sql statement and im just wondering if anyone could help me I have a ScoutingReport Table that stores player scouting info such as the matched, coached and playerId and I need a count of the amount of times a player has been scouted but if the same scout reports on the same match twice then I must only be counted once hers what I have started with any help would be appreciated.

Select Count(Distinct(MatchId))
From ScoutingReport
where ScoutedPersonId = 10002
group by MatchId
order by MatchId

Thanks in advance
Tim

andyharman20-Feb-07 7:05

20-Feb-07 7:05

Try:

select ScotedPersonId, count(distinct MatchId)
from ScotingReport
group by ScoutedPersonId
order by ScoutedPersonId

tadhg8820-Feb-07 7:32

20-Feb-07 7:32

thats what i need so far but i should also have said that if a different scout reports on the same match then it should be counted as 2 reports any ideas
thanks for your reply

andyharman20-Feb-07 8:17

20-Feb-07 8:17

Which database are you using? You may be able to get away with:

select ScoutedPersonId,
   count(distinct convert(varchar, MatchId) + '-' + convert(varchar, ScoutId))
from ScotingReport
group by ScoutedPersonId
order by ScoutedPersonId

(which cheats a bit and turns the MatchId and ScoutId into a single string value that can be counted) or

select ScoutedPersonId, Sum(ScoutCount)
from (
  select ScoutedPersonId, MatchId, count(distinct ScoutId) AS ScoutCount
  from ScoutingReport
  group by ScoutedPersonId, MatchId
) A
group by ScoutedPersonId
order by ScoutedPersonId

(which uses an inline view).

Regards
Andy

tadhg8821-Feb-07 0:37

21-Feb-07 0:37

both worked perfectly
ur an absolute gentleman thanks alot man much appreciated

tadhg8822-Feb-07 1:53