Writting SQL statement

dear all,
is it possible to write sql statement in ECC without abap? e.g. test a sql statement query with any abap required?

check this
https://www.sdn.sap.com/irj/scn/go/portal/prtroot/docs/library/uuid/40bec8da-4cd8-2910-27a9-81f5ce10676c
a®

Similar Messages

How to write Sql statement for M:M relationship set?

If theres two Entity sets "Department" and "Employee".....
Department Entity contains this attribute dno,dname,address etc and
"Employee" entity contain attribute like Empno,Empname,date,time etc
and theres Many to Many relationship between Employee and Deapartment Entity set,i.e.each employee can work in any department, and each department has at least one employee (Many to Many relationship)
What should be the SQL statement to create table for Many to Many relationship sets for Employee and Department set..
Any helps will be appreciated.
Femi

create 3 tables: employee, department and empdept
in the empdept table you reference foreign keys from the 2 other tables. so if my employee id was 42 and I work for departments 1,2 and 3, this table would look like this:
emp_fk | dept_fk
42           1
42           2
42           3

Need help for sql statement

Hi,
Need help to write sql statement.
create table t_dt ( dt_start date, dt_end date, amount number);
insert into t_dt values('1-Jan-10','10-Feb-10',12);
insert into t_dt values('11-Feb-10','10-Mar-10',10);
insert into t_dt values('11-Mar-10','20-Apr-10',8);
insert into t_dt values('21-Apr-10','28-Jun-10',10);
insert into t_dt values('29-Jun-10','20-Sep-10',10);
insert into t_dt values('21-Sep-10','10-Oct-10',10);
insert into t_dt values('11-Oct-10','31-Dec-10',8);
insert into t_dt values('1-Jan-11','10-Feb-11',8);
insert into t_dt values('11-Feb-11','10-Mar-11',7);
insert into t_dt values('11-Mar-11','20-Apr-11',6);
insert into t_dt values('21-Apr-11','28-Jun-11',6);
insert into t_dt values('29-Jun-11','20-Sep-11',6);
insert into t_dt values('21-Sep-11','10-Oct-11',4);
insert into t_dt values('11-Oct-11','31-Dec-11',8);
Result should be like below..
dt_start     dt_end     Amount
1-Jan-10     10-Feb-10     12
11-Feb-10     10-Mar-10     10
11-Mar-10     20-Apr-10     8
21-Apr-10     10-Oct-10     10
11-Oct-10     10-Feb-11     8
11-Feb-11     10-Mar-11     7
11-Mar-11     20-Sep-11     6
21-Sep-11     10-Oct-11     4
11-Oct-11     31-Dec-11     8
Just to explain the example, take a row with start date as 21-Apr-10 in the above insert statements, since it has the same amount for next two rows (i.e. with start date '29-Jun-10' and '21-Sep-10') these 3 rows should be converted to represent only 1 row in the result and the start date and end date should be changed per the result shown above.
Thanks.

Hello
I think this gives yuo what you need....
SELECT
    MIN(dt_start),
    MAX(dt_end),
    amount
FROM
    (   SELECT
            dt_start,
            dt_end,
            MAX(marker) OVER(ORDER BY dt_start) marker,
            amount
        FROM
                Select
                    dt_start,
                    dt_end,
                    amount,
                    CASE
                        WHEN LAG(amount) OVER(ORDER BY dt_start) <> amount THEN
                            ROW_NUMBER() OVER(ORDER BY dt_start)
                    END marker
                from t_dt
GROUP BY
     amount,
     marker
order by
     MIN(dt_start)
MIN(DT_START)        MAX(DT_END)              AMOUNT
01-JAN-2010 00:00:00 10-FEB-2010 00:00:00         12
11-FEB-2010 00:00:00 10-MAR-2010 00:00:00         10
11-MAR-2010 00:00:00 20-APR-2010 00:00:00          8
21-APR-2010 00:00:00 10-OCT-2010 00:00:00         10
11-OCT-2010 00:00:00 10-FEB-2011 00:00:00          8
11-FEB-2011 00:00:00 10-MAR-2011 00:00:00          7
11-MAR-2011 00:00:00 20-SEP-2011 00:00:00          6
21-SEP-2011 00:00:00 10-OCT-2011 00:00:00          4
11-OCT-2011 00:00:00 31-DEC-2011 00:00:00          8
9 rows selected.HTH
David
Edited by: Bravid on Feb 23, 2012 12:08 PM
Beaten to it by Frank! :-)

How to write the sql statement of my finder function in cmp?

hi,
I create a cmp ejb from table INFOCOLUMN,and I create a my finder function ,which sql statement is :
select * from INFOCOLUMN WHERE employee_id=id
employee_id is a column of the table,and id is the finder function parameter.
The error is : invalid column name
So,how to write the sql statement.
Thanks .

Mole-
Bind variables are of the form $1, $2, etc., so your query stmt should look like:
select * from INFOCOLUMN WHERE employee_id=$1
-Jon

Write to screen SQL statement

The below command updates a table through an SP. All the variables are being filled, but it's not updating the table. I hard code the variables on the SQL Server side and it works fine. How can I write to screen whatever's being sent to the below so I can see the actual SQL statement?
thanks!
<%
set rsRMAInsert = Server.CreateObject("ADODB.Command")
rsRMAInsert.ActiveConnection = MM_STRING
rsRMAInsert.CommandText = "jps.stp_Update_CustomerService"
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@Status", 200, 1,50,rsRMAInsert_Status)
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@Notes", 200, 1,6000,rsRMAInsert_Notes)
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@ReqDate", 200, 1,50,rsRMAInsert_ReqDate)
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@RecDate", 200, 1,50,rsRMAInsert_RecDate)
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@AppBy", 200, 1,50,rsRMAInsert_AppBy)
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@SendRep", 200, 1,3,rsRMAInsert_SendRep)
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@MoneyExp", 200, 1,3,rsRMAInsert_MoneyExp)
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@MoneySent", 200, 1,3,rsRMAInsert_MoneySent)
rsRMAInsert.Parameters.Append rsRMAInsert.CreateParameter("@ID", 3, 1,9,rsRMAInsert_ID)
rsRMAInsert.CommandType = 4
rsRMAInsert.CommandTimeout = 0
rsRMAInsert.Prepared = true
rsRMAInsert.Execute()
%>

You would want to post this question to the Dreamweaver Application Development forum -
http://forums.adobe.com/community/dreamweaver/dreamweaver_development

How can I write a SQL statement which checks if a table exists?

How can I write a SQL statement which tells me whether a table exists?

execute an sql query: select * from <tablename>
catch the exception n check whether the erroe code
matches the one that occurs for table doesn't exist
that's itHow is your answer any different from the one given in the first reply?
It isn't.
As WorkForFood says DatabaseMetaData has a bunch of methods for getting information about tables but this is more useful when you don't know the names of any of the tables.. it sounds like you do so I would concur SELECT from table is probably the quickest way to go. If it helps the Xopen error should be either S1000 or 42S01 (I think) but I would try and see if there is a specific vendor code for table not found/not exists error and check for that.

Can I write merge SQL statement having count(*)?

Hi this is a followup question of my previous post. I tried to use a merge SQL statement to solve my problem but now I bump into another problem.
Now I have a table where the field I need to update is a partial PK
when using a merge SQL statement, I have to put a where clause that check if count(*) = 1 because of the PK constraint
Where can I put the count(*) = 1 clause?
Here are the details:
I have two tables TA and TB, where TA contains the fields ID, FULLNAME, TYPE and TB contains the fields ID, FIRSTNAME
I want to update the firstnames in TB to be the firstnames from TA where TB.ID = TA.ID and TA.TYPE = 'ABC'
{ID, FIRSTNAME} are PKs but for the same ID, there can be more than 1 firstname.
e.g.
TA
ID | FULLNAME | TYPE
1 Caroline T ABC
2 Mary C DEF
3 Peter J ABC
TB
ID | FIRSTNAME
1 Caroline
1 Carol
1 C,
3 Peter
I need to update TB with the new firstnames from TA where type is 'ABC' but only for those fields that have count(TB.ID) = 1
when I try to run this SQL statement
merge into TB B using TA A
on (A.ID = B.ID and A.TYPE = 'ABC')
when matched then update set B.FIRSTNAME = substr(A.FULLNAME, 1, instr(A.FULLNAME, ',') - 1)
I got this error SQL Error: ORA-00001: unique constraint (TEST.PK_TB) violated
which I believe is because I updated those fields say ID = 1, all with 'Caroline'
that means I will have to add a clause having count(TB.ID) = 1
How would you do it?
Server is Oracle 11g
Thank you!

Hi,
One way is to join ta and tb in the USING clause, and eliminate the duplicates there.
MERGE INTO     tb     dst
USING   (
         SELECT    ta.id
         ,           REGEXP_SUBSTR ( MIN (ta.fullname)
                                , '^[^,]*'
                        )     AS firstname
         FROM      ta
         JOIN      tb ON ta.id     = tb.id
         WHERE     ta.type      = 'ABC'
         GROUP BY ta.id
         HAVING    COUNT (*)     = 1
     )                   src
ON     (scr.id      = dst.id)
WHEN MATCHED THEN UPDATE
SET     dst.firstname     = src.firstname
;If you'd care to post CREATE TABLE and INSERT statements for your sample data, then I could test this.
I used REGEXP_SUBST instead of SUBSTR and INSTR to find the firstname, because I find it a little cleaner, and becuase it returns something even when there is no ',' in fullname, which I'm guessing is what you really want. (None of the fullnames in your sample data have ','s.) You could use SUBSTR and INSTR instead.

Can we use if/else in sql statement?

Hi,
Is it possible to write a sql statement like the one below (not plsql)?
if (select 'True' from dual) = 'True' -- 1
then
select 'True' from dual; -- 2
else
select 'False' from dual; -- 3
endBasically, I want to execute either sql 2 or 3 based on the result of sql 1. I'm not sure if this is possible using only sql and not having to write a stored proc.
What are the different ways I can implement this?
Thanks for the help.

Billy Verreynne wrote:
The case syntax is a bit funny though as there's not a single condition evaluation (like a DECODE or case structs from some other languages).
This is what I would expect a typical case struct to look like - evaluating a single condition:
case <condition>
when <value-1> then return <result-1>
when <value-n> then return <result-n>
else
return <return-z>
end
?:| You mean like this...?
SQL> ed
Wrote file afiedt.buf
1 select empno, ename, deptno
2        ,case deptno
3           when 10 then 'This is Department 10'
4           when 20 then 'And department 20'
5           when 30 then 'And of course department 30'
6         else
7           'Blimey it is something else!'
8         end as dept_desc
9* from emp
SQL> /
     EMPNO ENAME          DEPTNO DEPT_DESC
      7369 SMITH              20 And department 20
      7499 ALLEN              30 And of course department 30
      7521 WARD               30 And of course department 30
      7566 JONES              20 And department 20
      7654 MARTIN             30 And of course department 30
      7698 BLAKE              30 And of course department 30
      7782 CLARK              10 This is Department 10
      7788 SCOTT              20 And department 20
      7839 KING               10 This is Department 10
      7844 TURNER             30 And of course department 30
      7876 ADAMS              20 And department 20
      7900 JAMES              30 And of course department 30
      7902 FORD               20 And department 20
      7934 MILLER             10 This is Department 10
14 rows selected.

Produce report NOT based on a single sql statement

I want to produce a tabular report based on a series of sql statments. Specifically, a report of managers that wil include counts of employees that are in other tables using differing criterias.
Name Count Count
using using
criteria 1 criteria 2
Manager1 35 242
I would expect to write an anonymous pl/sql block with a driving cursor determining the managers to report on. Within that cursor, I would execute a number of other queries to derive the count for each of the two columns.
I have tried creating a report region based on a sql statement, but that requires a single sql statement. I also tried creating a report region based on plsql, but it required an into statement of defined items. This option looks like it can provide multiple rows, but since it selected 'INTO' named fields, it only creates a report with the last row of data.
I must be missing something. Any suggestions are greatly appreciated!!!

If you want a wizard to create the form and report for you then yes you need to have a table. One thing that you can do is define a view that contains the data you need and define an Instead Of trigger on that view so the automatic fetch and dml will work but you can have the data stored into the different objects. basically the view and the trigger work as a router/dispatcher for the data.
*edit*
I should also add that you can write a pl/sql package which does the fetch and the dml operations with the form items as input. This is the solution I would typically use for any form that was not a simple CRUD form for a table. One thing to note is for the fetch I prefer to use out parameters for the form items so it requires the developer to map the item to the param in the app so it will show up when you are searching through the app. I highly discourage hiding item references inside of packaged code.
Good Luck!
Tyson
Message was edited by: TysonJouglet

How to store data into database by reading sql statements from text file

how to write java program for storing data into database by reading sql statements from text file

Step 1: Create a property file to add various queries.
Step 2: Read the properties file using ResourceBundle
Step 3: Use the jdbc to execute the query read from the property file.
So in future if you need to change query no need do any modifications in java program. But depends on how you use the property file.

Possible to do "grant" sql statement in Native SQL?

We have a need to do a grant of access from one of our systems out for various applications. In order for this to work we need to run a grant access command on the table and are trying to put a wrapper around this so we can use an abap. Below is the code I am unit testing. Two questions. First, can a grant be done via native SQL in abap? Second, if it can be done, what is the error with the logic where I am trying to put in the table name via a parameter.
REPORT ZLJTEST2.
tables dd02l.
DATA scarr_carrid TYPE dd02l-tabname.
SELECT-OPTIONS s_carrid for dd02l-tabname no intervals.
DATA s_carrid_wa LIKE LINE OF s_carrid.
DATA name TYPE c LENGTH 20.
TRY.
    EXEC SQL.
      CREATE FUNCTION selfunc( input CHAR(20) )
        RETURNING char(20);
        DEFINE output char(20);
        set schema sapr3;
        grant select on table input to group infouser;
        RETURN output;
        END FUNCTION;
    ENDEXEC.
    LOOP AT s_carrid INTO s_carrid_wa
                     WHERE sign = 'I' AND option = 'EQ'.
      TRY.
         EXEC SQL.
            EXECUTE PROCEDURE selfunc( IN :s_carrid_wa-low,
                                       OUT :name )
          ENDEXEC.
          WRITE: / s_carrid_wa-low, name.
        CATCH cx_sy_native_sql_error.
          MESSAGE `Error in procedure execution` TYPE 'I'.
      ENDTRY.
    ENDLOOP.
    EXEC SQL.
      DROP FUNCTION selfunc;
    ENDEXEC.
CATCH cx_sy_native_sql_error.
    MESSAGE `Error in procedure handling` TYPE 'I'.
ENDTRY.

Hi,
Yes it is posible.
I made one program like you want. But it need very long code.
Here I explain the idea:
1. Create Screen with input TEXT EDIT CONTROL.
    This is for input SQL Statement.
2. Get SQL Statement from Text Edit Control using method <b>get_text_as_r3table</b>.
3. Now we need to separate SQL Statement into different table.
    We Separate SELECT, FROM, WHERE, GROUP, HAVING, ORDER, etc.
4. We need dynamic internal table to store the data.
5. Select the data according SQL statement.
      SELECT (IT_SELECT)
        into corresponding fields of table <dyn_table>
      FROM (IT_FROM)
      WHERE (IT_WHERE)
      GROUP BY (IT_GROUP)
      HAVING (IT_HAVING)
      ORDER BY (IT_ORDER).
6. Display our data using ALV GRID
Hopefully it will help you.
Regards,

Using Date objects in SQL statements

Hi, I am wondering if it is possible to use Date objects while trying to retrieve entries from an access database with fields set as "Date/Time" fields, as opposed to "Text"? If so, do I need to reformat it before inserting it into the SQL statement, or can I just leave it as a Date object, and call up the variable name?
For example - "SELECT * FROM database WHERE startDate = ' " + dateObject + ' ". And do i need quotes round this value?
Thanks in advance.

I had some problems by just fetching a date and pushing it back to a MS SQL Database.
I used the following reformating step to fix it:
if (o instanceof Timestamp)
String time = o.toString().substring(8,10)+"."+o.toString().substring(5,7)+"."+o.toString().substring(0,4)+" "+o.toString().substring(11,19);
ht.put(columnName, time);
o is the object i receve from the select statement and ht is a Hashtable where i put my data.
To write the data back i now can use:
UPDATE table SET columnname = 'valueOfColumn'
This means i use the value i created above in Quotes to write it back.
Format of Timestamp.toString() is something like yyyy-mm-dd hh:mm:ss.xx
Format used by MS SQL (and i think by Access to) 'dd.mm.yyyy hh:mm:ss'

Update sql statement in sender/receiver jdbc adapter

Hi
What is the update sql statement we give in the sender/ receiver adapter when
1) SELECT statement was written in query sql statement
2) EXECUTE statement was written for a stored procedure in query sql statement
thanks
Anitha

> 1) SELECT statement was written in query sql statement
your table should have some value "already_processed" which prevents a second processing of that table.
the update statement sets that value.
> 2) EXECUTE statement was written for a stored procedure in query sql statement
the same as above.
when your stored procedure already does the update, then write a second stored procedure which does nothing.

Find start and end execution time of a sql statement?

I am have databases with 10.2.0.3 and 9.2.0.8 on HP UNIX 11i and Windows 200x.
I am not in a position to turn on sql tracing in production environment. Yet, I want to find when a sql statement started executing and when it ended. When I look at v$sql, it has information such FIRST_LOAD_TIME, LAST_LOAD_TIME etc. No where it has information last time statement began execution and when it ended execution.. It shows no of executions, elapsed time etc, but they are cumulative. Is there a way to find individual times (time information each time a sql statement was executed. – its start time, its end time ….)? If I were to write my own program how will I do it?
Along the same line, when an AWR snapshot is shown, does it only include statements executed during that snapshot or it can have statements from the past if they have not been flushed from shared memory. If it only has statements executed in the snapshot period, how does it know when statement began execution?

Hi,
For oracle 10g you can use below query to find start and end time, you can see data for last seven days.
select min(to_char(a.sample_time,'DD-MON-YY HH24:MI:SS')) "Start time", max(to_char(a.sample_time,'DD-MON-YY HH24:MI:SS')) "End Time", b.sql_text
from dba_HIST_ACTIVE_SESS_HISTORY a,DBA_HIST_SQLTEXT b where
a.sql_id=b.sql_id
order by 1;
Regards
Jafar

Nested SQL statements for complex, detailed queries.

Is it possible to write nested SQL statements for complex, detailed queries. A nested query
has a WHERE clause that includes a SELECT statement ? Is it true or false ?

Hi wahid,
Here are pretty good examples:
http://www.databasejournal.com/features/mssql/article.php/3464481/Using-a-Subquery-in-a-T-SQL-Statement.htm
http://technet.microsoft.com/en-us/library/aa213252(v=sql.80).aspx
Regards Harsh

Writting SQL statement

Similar Messages

Maybe you are looking for