Copy of Complex SQL Queries 1664988700
Copy of Complex SQL Queries 1664988700
SELECT DISTINCT SAL FROM EMP A WHERE &N=(SELECT COUNT (DISTINCT B.SAL)
FROM EMP B WHERE A.SAL<=B.SAL);
SELECT DNAME, DEPTNO FROM DEPT WHERE EXISTS (SELECT * FROM EMP WHERE
DEPT.DEPTNO = EMP.DEPTNO)
4. To Find The Not Null Column Alone In A Table.SELECT COLUMN_NAME FROM
USER_TAB_COLUMNS WHERE NULLABLE = 'N' AND TABLE_NAME = 'COUNTRY'
DELETE DEPT WHERE ROWID NOT IN (SELECT MAX (ROWID) FROM DEPT GROUP BY
DEPTNO HAVING COUNT (*) >=1)
2.SELECT SAL FROM EMP WHERE SAL >= ALL (SELECT SAL FROM EMP) 7.
SELECT EMPNO, LPAD (‘ ‘, 6*(LEVEL – 1)) || ENAME “EMPLOYEE NAME” FROM EMP START
WITH ENAME=’KING’ CONNECT BY PRIOR EMPNO = MGR
Choose one of the following queries to identify or remove duplicate rows from a table leaving one record:
Method 1:
DELETE FROM table_name A WHERE ROWID > (SELECT min (rowid) FROM table_name B
-1-
COMPLEX QUERIES
SQL> Delete from my_table where rowid not in (select max (rowid) from my_table group by
my_column_name);
SQL> delete from my_table t1 where exists (select 'x' from my_table t2 where t2.key_value1 =
t1.key_value1
And t2.key_value2 = t1.key_value2and t2.rowid > t1.rowid);
Note: If you create an index on the joined fields in the inner loop, you for all intensive purposes eliminate
N^2 operations (no need to loop through the entire table on each pass by a record).
Create your table with a NOT NULL column (say SEQNO). This column can now be populated with unique
values:
14. How can I get the time difference between two date columns?
Select dept, sum (decode (sex,'M', 1,0)) MALE, sum (decode (sex,'F', 1,0)) FEMALE, count
(decode (sex,'M', 1,'F', 1)) TOTAL from my_emp_table group by dept;
A value x will be between values y and z if GREATEST (x, y) = LEAST (x, z). Look at this example:
Select f2, count (decode (greatest (f1, 59), least (f1, 100), 1, 0)) "Range 60-100", Count
(decode (greatest (f1, 30), least (f1, 59), 1, 0)) "Range 30-59",
Count (decode (greatest (f1, 29), least (f1, 0), 1, 0)) "Range 00-29"
From my_table group by f2;
For equal size ranges it might be easier to calculate it with DECODE (TRUNC (value/range), 0, rate_0, 1,
rate_1,).
-2-
COMPLEX QUERIES
E.g.
Select ename "Name", sal "Salary", decode (trunc (f2/1000, 0), 0, 0.0,1, 0.1, 2, 0.2, 3, 0.31) "Tax
rate"
From my_table;
17. Can one only retrieve the Nth row from a table?
SELECT f1 FROM t1 WHERE rowid = (SELECT rowid FROM t1 WHERE rownum <= 10
MINUS
SELECT rowid FROM t1 WHERE rownum < 10);
SELECT * FROM tableX WHERE rowid in (SELECT rowid FROM tableX WHERE rownum <= 7
MINUS
SELECT rowid FROM tableX WHERE rownum < 5);
19. How does one select EVERY Nth row from a table?
One can easily select all even, odd, or Nth rows from a table using SQL queries like this:
SELECT *FROM EMP WHERE (ROWID, 0) IN (SELECT ROWID, MOD (ROWNUM, 4) FROM EMP);
SELECT * FROM (SELECT rownum rn, empno, ename FROM EMP) temp WHERE MOD (temp.
ROWNUM, 4) = 0;
20. How does one select the TOP N rows from a table?
SELECT * FROM my_table a WHERE 10 >= (SELECT COUNT (DISTINCT maxcol) FROM my_table
b
WHERE b.maxcol >= a.maxcol) ORDER BY maxcol DESC;
This is definitely non-relational (enough to kill Codd and then make him roll in his grave) and is a feature I
have not seen in the competition.
The definitive example is in the example SCOTT/TIGER database, when looking at the EMP table
(EMPNO and MGR columns). The MGR column contains the employee number of the "current"
employee's boss.
You have available an extra pseudo-column, LEVEL, that says how deep in the tree you are. Oracle can
handle queries with a depth up to 255.
Select LEVEL, EMPNO, ENAME, MGR from EMP connect by prior EMPNO = MGR start with MGR is
NULL;
You can get an "indented" report by using the level number to sub-string or lpad a series of spaces and
Concatenate that to the string.
-3-
COMPLEX QUERIES
You use the start with clause to specify the start of the tree(s). More than one record can match the starting
condition. One disadvantage of a "connect by prior" is that you cannot perform a join to other tables. Still, I
have not managed to see anything else like the "connect by prior" in the other vendor offerings and I like
trees. Even trying to doing this programmatic ally in embedded SQL is difficult as you have to do the top
level query, for each of them open a cursor to look for child nodes, for each of these open a cursor....
Pretty soon you blow the cursor limit for your installation.
The way around this is to use PL/SQL, open the driving cursor with the "connect by prior" statement, and
the select matching records from other tables on a row-by-row basis, inserting the results into a temporary
table for later retrieval.
The Oracle decode function acts like a procedural statement inside an SQL statement to return different
values or columns based on the values of other columns in the select statement.
Some examples:
Select decode (sex, 'M', 'Male', 'F', 'Female', 'Unknown') from employees;
Select a, b, decode( abs (a-b), a-b, 'a > b',0, 'a = b','a < b') from tableX;
Select decode (GREATEST (A, B), A, 'A is greater than B', 'B is greater than A')...
Note: The decode function is not ANSI SQL and are rarely implemented in other RDBMS offerings. It is
one of the good things about Oracle, but use it sparingly if portability is required.
23. How can one dump/ examine the exact content of a database column?
DUMP (COL1)
-------------------------------------
Typ=96 Len=4: 65,66,67,32
For this example the type is 96, indicating CHAR, and the last byte in the column is 32, which is the ASCII
code for a space. This tells us that this column is blank-padded.
Oracle does not provide a way to DROP a column (reference: Enhancement Request 51118). However,
Joseph S. Testa wrote a DROP COLUMN package that can be downloaded from
Other workarounds:
-4-
COMPLEX QUERIES
1. Rename t1 to t1_base;
Create view t1 <column list with new name> as select * from t1_base;
2. create table t2 <column list with new name> as select * from t1;
Drop table t1;
Rename t2 to t1;
Declare
a integer;
b integer;
Begin
a := dbms_pipe.create_pipe('kumaran');
dbms_pipe.pack_message('kumaran software is a good company');
b := dbms_pipe.send_message('kumaran');
if b = 0 then
dbms_output.put_line('successfully send');
else
dbms_output.put_line('not send');
end if;
end;
declare
a integer;
b varchar2(30);
begin
a := dbms_pipe.receive_message('kumaran');
dbms_pipe.unpack_message(b);
if a = 0 then
dbms_output.put_line('successfully received');
dbms_output.put_line(b);
else
dbms_output.put_line('not received');
end if;
end;
-5-
COMPLEX QUERIES
Only local or packaged subprograms can be overloaded. Therefore, you cannot overload standalone
subprograms. Also, you cannot overload two subprograms if their formal parameters differ only in name or
parameter mode. For example, you cannot overload the following
Finally, you cannot overload two functions that differ only in return type (the datatype of the result value)
even if the types are in different families. For example, you cannot overload the following functions:
-6-
COMPLEX QUERIES
Sure there is, but it is hard to find them... In Server Manager from Oracle7.3: ORADEBUG HELP
SMON coalesces free space (extents) into larger, contiguous extents every 2 hours and even then only for
a short period of time. SMON will not coalesce free space if a tablespace's default storage parameter
"pctincrease" is set to 0. With Oracle 7.3 one can manually coalesce a tablespace using the ALTER
TABLESPACE ... COALESCE; command, until then use:
-7-
COMPLEX QUERIES
SQL> alter session set events 'immediate trace name coalesce level n';
Where 'n' is the tablespace number you get from SELECT TS#, NAME FROM SYS.TS$;
You can get status information about this process by selecting from the
DBA_FREE_SPACE_COALESCED view.
Always set PCTINCREASE to 0 or 100. Bizarre values for PCTINCREASE will contribute to fragmentation.
For example if you set PCTINCREASE to 1 you will see that your extents are going to have weird and
wacky sizes: 100K, 100K, 101K, 102K, etc. Such extents of bizarre size are rarely re used in their entirety.
PCTINCREASE of 0 or 100 gives you nice round extent sizes that can easily be reused. E.g. 100K, 100K,
200K, 400K, etc.
6. Where can one find the high water mark for a table?
There is no system table which containts the high water mark (HWM) information. You can calculate the
HWM using the results from the following SQL statements:
SELECT BLOCKS FROM DBA_SEGMENTS
WHERE OWNER = UPPER (owner) AND SEGMENT_NAME = UPPER (table);
• You can also use the DBMS_SPACE package and calculate the HWM = TOTAL_BLOCKS -
UNUSED_BLOCKS - 1.
You can prevent ORA-600 space leak messages during database shutdown by telling the kernel not to
check for memory leakage. This undocumented feature :-) was introduced with Oracle 7.1.6 and can be
prevented by setting:
SQL> ALTER SESSION SET EVENTS "10262 trace name context forever, level 1024" 8.
Oracle recommends that your database blocks size matches, or be multiples of your operating system
block size. One can go smaller, but the performance cost is significant. Your choice should depend on the
type of application you are running. If you have lots of small transaction like with OLTP, use a small block
size. With fewer but larger transactions, like with a DSS application, use a large block size. If you are using
a volume manager, consider your "operating system block size" to be 8K. This is because volume manager
products use 8K blocks (and this is not configurable).
-8-
COMPLEX QUERIES
drop user A
You can manually increase or decrease the size of a datafile in Oracle 7.2 using the
Because you can change the sizes of datafiles, you can add more space to your database without adding
more datafiles. This is beneficial if you are concerned about reaching the maximum number of datafiles
allowed in your database. Manually reducing the sizes of datafiles allows you to reclaim unused space in
the database. This is useful for correcting errors in estimates of space requirements. Also, datafiles can be
allowed to automatically extend if more space is required. Look at the following command:
CREATE TABLESPACE pcs_data_ts
DATAFILE 'c:\ora_apps\pcs\pcsdata1.dbf' SIZE 3M
AUTOEXTEND ON NEXT 1M MAXSIZE UNLIMITED
DEFAULT STORAGE ( INITIAL 10240
NEXT 10240
MINEXTENTS 1
MAXEXTENTS UNLIMITED
PCTINCREASE 0)
ONLINE
PERMANENT;
While your production database is running, take an ON-LINE backup and restore it on duplicate hardware.
Note that an export will not work! On your standby database, issue the following commands:
Write a job to copy archived redo log files from your primary database to the standby system, and apply
the redo log files to the standby database (pipe it). Remember the database is recovering and will prompt
you for the next log file to apply.
When you need the standby database stop the recovery process and activate it:
-9-
COMPLEX QUERIES
14. How can I give developers access to trace files (required as input to tkprof)?
The "alter session set SQL_TRACE = true" command generates trace files in USER_DUMP_DEST that is
used by developers as input to tkprof. On Unix the default file mask for these files are "rwx r-- -- -".
There is an undocumented INIT.ORA parameter that will allow everybody to read (rwx r-- r--) this trace
files:
_trace_files_public = true
Include this in your INIT.ORA file and bounce your database for it to take effect. 15.
SELECT
C.INSTANCE,
TO_DATE(JUL.VALUE, 'J')
|| TO_CHAR(FLOOR(SEC.VALUE/3600), '09') || ':'
|| SUBSTR (TO_CHAR(MOD(SEC.VALUE/60, 60), '09'), 2, 2)|| '.'
|| SUBSTR (TO_CHAR(MOD(SEC.VALUE, 60), '09'), 2, 2) STARTED
FROM
V$INSTANCE JUL,
V$INSTANCE SEC,
V$THREAD C
WHERE
JUL.KEY LIKE '%JULIAN%'
AND SEC.KEY LIKE '%SECOND%';
Of course it is not advisable to bridge Oracle's security, but look at this example:
PASSWORD
-----------------------------
F894844C34402B67
- 10 -
COMPLEX QUERIES
17. Is it true that OPS$ accounts are a security risk in a C/S environment?
If you allow people to log in with OPS$ accounts from Windows95, you cannot tell who that really is. With
terminals, you can rely on passwords, with Win95, you cannot. If you set REMOTE_OS_AUTHENT=TRUE
in your init.ora file, Oracle Assumes that the remote OS has authenticated the user.
If REMOTE_OS_AUTHENT is set to FALSE, you have no exposure from remote clients - you also won't
be able to connect from a remote client without a password (recommended). IDENTIFIED EXTERNALLY
will only be in effect from the local host. Also, if you're using OPS$ as your prefix, you'll be able to log on
locally with or without a password, regardless of whether you've identified your ID with a password or
defined it to be IDENTIFIED EXTERNALLY.
For every user using temporary space, there is an entry in SYS.V$_LOCK with type 'TS'. All temporary
segments are named 'ffff.bbbb' where 'ffff' is the file it is in and 'bbbb' is first block of the segment.
If your temporary tablespace is set to TEMPORARY, all sorts are done in one large temporary segment.
For usage status, see SYS.V_$SORT_SEGMENT
QUESTIONS & ANSWERS
1. Explain SCN.
Whenever a transaction is committed, LGWR writes transactions redo entries from the redo log buffer of
SGA to an online redo file and a (System Change Number) SCN is assigned to identify the redo entries for
each committed transaction.
- 11 -
COMPLEX QUERIES
High SCN:- During the prepare phase ( The Global coordinator asks participants to prepare (to promise to
Commit or Rollback the transaction, even if there is a failure). The highest SCN at all node in the
transaction is determined. The transaction is then committed with the high SCN at the commit point site.
The SCN is then sent to all prepared nodes along with the commit decision.
High water mark is the Highest Block Number in which data has been stored in the Segment. 4.
Row Chaining: If an update to a row causes that row to no longer completely fit in a Single data block,
then that row may be moved to another data block or the row may be Chained to another block. If row
length is greater that the Oracle block size, the row will be chained.
Row Migration: If a row in a data block is updated so that overall row length increases and the block's free
space has been completely filled, the data for the entire row is Migrated to a new data block, assuming the
entire row can fit in a new block. Oracle preserves the original row piece of a migrated row to point to the
new block containing the migrated row; the ROWID of a migrated row does not change.
Chained rows data is stored in a chain of data blocks and in Row Migration the entire row is shifted to a
new location.
6. Coalesce Details.
Is the process by which SMON automatically Coalesces neighboring free extents in a single large free
space. It is done automatically if PCT Increase of the Tablespace is non zero or can be done manually by
issuing "ALTER TABLESPACE <table_space_name>COALESCE".
The Header contains general block information, such as Block Address, Segment Type, such as data,
Index or Rollback. Some headers are fixed in size, the total block overhead size is variable. On an Average
fixed & variable size of data block overhead is total 84 to 107 bytes.
When oracle process accesses a buffer the process moves the buffer to the most-recently used (MRU)
end of the LRU list. As most of the buffers moved to the MRU the dirty "Age" towards the LRU end of the
LRU list. This process is called Buffer Aging.
The Control file of a database is a small binary file necessary for the database to start and operate
successfully. A control file is updated continuously by Oracle during database use, so it must be available
for writing whenever the database is open. Each control file is associated with only one Oracle Database.
Among other things, a control file contains information such as
- 12 -
COMPLEX QUERIES
• The names and locations of associated database and online redo log files • The
current log sequence number
• Checkpoint information
Each time a data file or a online redo log file is added to, renamed in, or dropped from the database, the
control file is updated to reflect this physical structure change. These changes are recorded so that
• Oracle can identify the datafiles and online redo log files to open during database startup. • Oracle
can identify files that are required or available in case database recovery is necessary. •
It is highly recommended that we backup up our Control file as soon as we make some change to the
physical structure of the database.
The PCTINCREASE parameter has been replaced by a parameter called OPTIMAL. This specifies the
optimal size of a rollback segment in bytes. It can also be specified in kilobytes or megabytes. The RDBMS
tries to keep the segment at its specified optimal size. The size is rounded up to the extent boundary,
which means that the RDBMS tries to have the fewest number of extents such that the total size is greater
than or equal to the size specified as OPTIMAL. If additional space is needed beyond the optimal size, it
will eventually deallocate extents to shrink back to this size. The process of deallocating extents is
performed when the head moves from one extent to the next. At this time, the segment size is checked and
the RDBMS determines if the next extent should be deallocated. The extent can only be deallocated if
there are no active transaction in it. If necessary, the RDBMS will deallocate multiple extents at one time
until the segment has shrunk back to its optimal size. The RDBMS always deallocates the oldest inactive
extents as they are the least likely to be used for read consistency.
12. Checkpoint 0
Checkpoint process does not hamper the performance of the database but incorrect values for the above
two parameters can cause performance degradation.
13. Where analyzed information stored. The analyze information is stored in views like
• DBA_TABLES
• ALL_TABLES
• USER_TABLES
- 13 -
COMPLEX QUERIES
There is nothing like activating an Index. But I can say "Oracle automatically maintains and uses indexes
once they are created." There is a possibility of forcing a specific index to be used in our query by using
Hints.
Example: SELECT (+INDEX name_idx) emp_id, name FROM EMP WHERE name = "ALAM"; The
example can be considered as an Activation of an index. (If every reader agrees). But we can
deactivate or disable the indexes or make the optimizer not to use the indexes.
When executing the above statement Oracle optimizer will use the index available on the table to
resolve the query. But if we want oracle not to use the index we can rewrite the query as follows.
Case 2: SELECT emp_id, name FROM EMPWHERE name ||' ' ="ALAM";
Two space management parameters, control the use of free space for inserts of and updates to the row in
data blocks.
PCTFREE: - The PCTFREE parameter is used to set the percentage of a block to be reserved (kept free)
for possible updates to rows that already are contained in the blocks.
PCTUSED: - After a data block becomes full, as determined by PCTFREE, oracle does not consider the
block is for the insertion of new rows until the percentage of the block being used falls below the parameter
PCTUSED.
When processing the queries, Oracles often requires TEMPORARY workspace for intermediate stages of
SQL statement processing. Oracle automatically allocates this disk space called a TEMPORARY
SEGMENT. Typically, oracle requires a temporary segment as a work area for sorting. Oracle does not
create a segment if the sorting operation can be done in memory or if oracle finds some other way to
perform the operation using indexes.
CREATE INDEX
SELECT …ORDER BY
SELECT DISTINCT…
SELECT … GROUP BY
SELECT … UNION
SELECT … INTERSECT
SELECT … MINUS
Unindexed joins
Certain correlated subqueries.
17. Use of SQL* plus trace utility.
- 14 -
COMPLEX QUERIES
SQL* plus trace utility provides information on tuning that can be used in improving the performance of the
system.
A profile is a named set of resource limits. If resources limits are turned on, oracle limits user 's use of
database and instance resources to that given in his profile. We can assign a profile to a user, and a
default profile to all users who do not have specific profiles.
Oracle stores data in DATA BLOCKS also called as oracle blocks. One data blocks correspond to a
specific number of bytes of physical database space on disk. It is set using the parameter
DB_BLOCK_SIZE usually 2K or 4K. No of blocks for an extents depends on the size of the Extent itself.
Extent is a logical unit of database storage space allocation made up of a number of contiguous data
blocks. The extents are allocated based on the storage parameters specified, while creating the objects.
No matter what type, each segment in a database is created with at least one extent to hold its data. This
extent is called the segment's Initial extent. Exception to this rule is the Rollback Segments; they always
have at least two extents.
ORACLE database is comprised of three types of files: one or more Data files, two or more Redo
log file, and one or more Control files.
3. What is a Tablespace?
- 15 -
COMPLEX QUERIES
Every ORACLE database contains a Tablespace named SYSTEM, which is automatically created when
the database is created. The SYSTEM Tablespace always contains the data dictionary tables for the entire
database.
Each database is logically divided into one or more tablespaces. One or more data files are explicitly
created for each Tablespace.
6. What is a Schema?
Schema objects are logical structures that directly refer to the database's data. Schema objects include
tables, views, sequences, synonyms, indexes, clusters, database triggers, procedures, functions,
packages and database links.
Yes.
Yes.
- 16 -
COMPLEX QUERIES
Ans : D
6. SET TRANSACTION USE ROLLBACK SEGMENT <rbs_name> is used to create user objects in a
particular Tablespace
a] True, b] False
Ans : False
10. It is very difficult to grant and manage common privileges needed by different groups of
database users using the roles
a] True, b] False
Ans : False
11. What is difference between a DIALOG WINDOW and a DOCUMENT WINDOW regarding
moving the window with respect to the application window
a] Both windows behave the same way as far as moving the window is concerned. b] A
document window can be moved outside the application window while a dialog window
cannot be moved
c] A dialog window can be moved outside the application window while a document
window cannot be moved
Ans : C
- 17 -
COMPLEX QUERIES
d] An message box requires an response from the userwhile a alert just flashes a
message an only requires an acknowledment from the user
Ans : C
13. Which of the following is not an reason for the fact that most of the processing is done at the
server ?
a] To reduce network traffic. b] For application sharing, c] To implement business rules
centrally, d] None of the above
Ans : D
16. What is the difference between a LIST BOX and a COMBO BOX ?
a] In the list box, the user is restricted to selecting a value from a list but in a combo box the
user can type in a value which is not in the list
b] A list box is a data entry area while a combo box can be used only for control purposes c]
In a combo box, the user is restricted to selecting a value from a list but in a list box the user
can type in a value which is not in the list
d] None of the above
Ans : A
17. In a CLIENT/SERVER environment , which of the following would not be done at the client ? a]
User interface part, b] Data validation at entry line, c] Responding to user events, d] None of
the above
Ans : D
18. Why is it better to use an INTEGRITY CONSTRAINT to validate data in a table than to use a
STORED PROCEDURE ?
a] Because an integrity constraint is automatically checked while data is inserted into or
updated in a table while a stored procedure has to be specifically invoked
b] Because the stored procedure occupies more space in the database than a integrity
constraint definition
c] Because a stored procedure creates more network traffic than a integrity constraint
definition
Ans : A
- 18 -
COMPLEX QUERIES
a] Navigational trigger
b] Key trigger
c] Transactional trigger
d] None of the above
Ans : A
22. The system variable that records the select statement that SQL * FORMS most recently used to
populate a block is
a] SYSTEM.LAST_RECORD
b] SYSTEM.CURSOR_RECORD
c] SYSTEM.CURSOR_FIELD
d] SYSTEM.LAST_QUERY
Ans: D
23. Which of the following is TRUE for the ENFORCE KEY field
a] ENFORCE KEY field characterstic indicates the source of the value that SQL*FORMS
uses to populate the field
b] A field with the ENFORCE KEY characterstic should have the INPUT ALLOWED
charaterstic turned off
a] Only 1 is TRUE
b] Only 2 is TRUE
c] Both 1 and 2 are TRUE
d] Both 1 and 2 are FALSE
Ans : A
27. The packaged procedure that makes data in form permanent in the Database is a]
Post
b] Post form
c] Commit form
d] None of the above
Ans : C
28. Which of the following is TRUE for the SYSTEM VARIABLE $$date$$ a]
Can be assigned to a global variable
b] Can be assigned to any field only during design time
c] Can be assigned to any variable or field during run time
d] None of the above
- 19 -
COMPLEX QUERIES
Ans : B
32. Name the two files that are created when you generate a form using Forms 3.0 a]
FMB & FMX, b] FMR & FDX, c] INP & FRM, d] None of the above
Ans : C
35. All datafiles related to a Tablespace are removed when the Tablespace is dropped a]
TRUE
b] FALSE
Ans : B
- 20 -
COMPLEX QUERIES
Ans : A
45. A database trigger doesnot apply to data loaded before the definition of the trigger a]
TRUE
b] FALSE
Ans : A
47. Which of the following does not affect the size of the SGA
a] Database buffer
b] Redolog buffer
c] Stored procedure
d] Shared pool
Ans : C
- 21 -
COMPLEX QUERIES
51. Declare
fvar number := null; svar number := 5
Begin
goto << fproc>>
if fvar is null then
<< fproc>>
svar := svar + 5
end if;
End;
Ans : A
Ans : C
Ans : C
- 22 -
COMPLEX QUERIES
Ans : D
Ans : C
Ans : A
Ans : A
Ans C
Ans : B
61. Which of the following is not correct about the “TABLE” datatype ?
a] Can contain any no of columns
b] Simulates a One-dimensional array of unlimited size
c] Column datatype of any Scalar type
d] None of the above
Ans : A
- 23 -
COMPLEX QUERIES
Ans C
Ans : B
Ans : C
66. Declare
a number := 5; b number := null; c number := 10;
Begin
if a > b AND a < c then
a := c * a;
end if;
End;
What will be the value of ‘a’ after execution ?
a] 50
b] NULL
c] 5
d] None of the above
Ans : C
67. Does the Database trigger will fire when the table is TRUNCATED ? a]
Yes
b] No
Ans : B
- 24 -
COMPLEX QUERIES
Ans : C
Ans : A
Select SAL from EMP E1 where 3 > ( Select count(*) from Emp E2 where
E1.SAL > E2.SAL ) will retrieve
a] 3500,5000,2500
b] 5000,2850
c] 2850,5750
d] 5000,5750
Ans : A
72. Is it possible to modify a Datatype of a column when column contains data ? a]
Yes
b] No
Ans B
Ans : B
Ans : C
Ans : D
- 25 -
COMPLEX QUERIES
Ans : C
77. If an UNIQUE KEY constraint on DATE column is created, will it accept the rows that are
inserted with SYSDATE ?
a] Will
b] Won’t
Ans : B
Ans : C
Ans : C
Ans : A
Ans : B
Ans : A
- 26 -
COMPLEX QUERIES
Ans : C
Ans : C
Ans : A
Ans : D
Ans : D
Ans : B
Ans : C - Please check the Correcness of this Answer ( The correct answer is 2 )
Ans : C
- 27 -
COMPLEX QUERIES
Ans : C
Ans : B
Ans : B
94. What type of file is used for porting Forms 4.5 applications to various platforms ?
a] .FMB file
b] .FMX file
c] .FMT file
d] .EXE file
Ans : C
Ans A
96. When a form is invoked with CALL_FORM does Oracle forms issues SAVEPOINT ? a]
Yes
b] No
Ans : A
97. Can we attach the same LOV to different fields in Design time ? a]
Yes
b] No
Ans : A
98. How do you pass values from one form to another form ?
a] LOV
b] Parameters
c] Local variables
d] None of the above
Ans : B
99. Can you copy the PROGRAM UNIT into an Object group ?
a] Yes
b] No
Ans : B
- 28 -
COMPLEX QUERIES
Ans : A
Ans : C
Ans : D
Ans : A
Ans : B
Ans : A
Ans : B
- 29 -
COMPLEX QUERIES
b] 20
c] 50
d] None of the above
Ans D
Ans : D
Ans : D
111. What is the appropriate destination type to send the output to a printer ? a]
Screen
b] Previewer
c] Either of the above
d] None of the above
Ans : D
113. If the maximum records retrieved property of a query is set to 10, then a summary value will be
calculated
a] Only for 10 records
b] For all the records retrieved
c] For all therecords in the referenced table
d] None of the above
Ans : A
114. With which function of a summary item in the COMPUTE AT option required ? a]
Sum
b] Standard deviation
c] Variance
d] % of Total function
Ans : D
115. For a field in a repeating frame, can the source come from a column which does not exist in the
datagroup which forms the base of the frame ?
a] Yes
- 30 -
COMPLEX QUERIES
b] No
Ans : A
116. What are the different file extensions that are created by Oracle Reports ?
a] .RDF file & .RPX file
b] .RDX file & .RDF file
c] .REP file & .RDF file
d] None of the above
Ans : C
117. Is it possible to Disable the Parameter form while running the report ? a]
Yes
b] No
Ans : A
118.What are the SQL clauses supported in the link property sheet ?
a] WHERE & START WITH
b] WHERE & HAVING
c} START WITH & HAVING
d] WHERE, START WITH & HAVING
Ans : D
Ans : D
120. If two groups are not linked in the data model editor, what is the hierarchy between them ? a]
There is no hierarchy betweeen unlimked groups
b] The group that is right ranks higher than the group that is to the left
c] The group that is above or leftmost ranks higher than the group that is to right or below it
d] None of the above
Ans : C
Ans : B
Ans : B
- 31 -
COMPLEX QUERIES
c] DB_BLOCK_SIZE
d] DB_BLOCK_BUFFERS
Ans : A
Ans : B
Ans : A
Ans : B
Ans : A
128. The following parameters are optional in init.ora parameter file DB_BLOCK_SIZE,
PROCESS
a} TRUE
b] FALSE
Ans : B
Ans : B
130. Constraints cannot be exported through Export command ?
a] TRUE
b] FALSE
Ans : B
131. It is very difficult to grant and manage common priveleges needed by different groups of
database users using roles
a] TRUE
b] FALSE
Ans : B
- 32 -
COMPLEX QUERIES
d] DBA_ROLLBACK_SEG
Ans : D
Ans : A
Ans : A
Yes. For example, if we had a table DEPT_SUMMARY, we could update the number of employees’ field
as follows:
update DEPT_SUMMARY s
set NUM_EMPS = (
select count(1)
from EMP E
where E.DEPTNO = S.DEPTNO
);
Can I remove duplicate rows?
Yes, using the ROWID field. The ROWID is guaranteed unique. There are many variations on this theme,
but the logic is to delete all but one record for each key value.
- 33 -
COMPLEX QUERIES
Yes! Those migrating from non-RDBMS application commonly ask this. This is definitely non relational
(enough to kill Codd and then make him roll in his grave) and is a feature I have not seen in the
competition.
The definitive example is in the example SCOTT/TIGER database, when looking at the EMP table
(EMPNO and MGR columns). The MGR column contains the employee number of the "current"
employee's boss.
You have available an extra pseudo-column, LEVEL, that says how deep in the tree you are. Oracle can
handle queries with a depth up to 255.
select LEVEL, EMPNO, ENAME, MGR from EMP connect by prior EMPNO = MGR start with MGR is
NULL;
You can get an "indented" report by using the level number to sub-string or lpad a series of spaces and
concatenate that to the string.
You use the start with clause to specify the start of the tree(s). More than one record can match the starting
condition.
One disadvantage of a "connect by prior" is that you cannot perform a join to other tables. Still, I have not
managed to see anything else like the "connect by prior" in the other vendor offerings and I like trees. Even
trying to do this programmatically in embedded SQL is difficult, as you have to do the top level query, for
each of them open a cursor to look for lower levelrows, for each of these...
The way around this is to use PL/SQL, open the driving cursor with the "connect by prior" statement, and
the select matching records from other tables on a row-by-row basis, inserting the results into a temporary
table for later retrieval.
Note that you can't trick Oracle by using CONNECT BY PRIOR on a view that does the join.
Imagine we have the EMP table and want details on the employee who has the highest salary. You need
to use a sub query.
You could get similar info on employees with the highest salary in their departments as follows
select e.ENAME, e.DEPTNO, e.SAL
from EMP e
where e.SAL = (
select max (e2.SAL)
from EMP e2
where e2.DEPTNO = e.DEPTNO
);
How can I get a name for a temporary table that will not clash?
- 34 -
COMPLEX QUERIES
Use a sequence, and use the number to help you build the temporary table name. Note that SQL-92 is
developing specific constructs for using temporary tables.
Oracle maintains a live set of views that you can query to tell you what you have available. In V6, the first
two to look at are DICT and DICT_COLUMNS, which act as a directory of the other dictionary views. It is a
good idea to be familiar with these. Not all of these views are accessible by all users. If you are a DBA you
should also create private DBA synonyms by running
There is no way a column can be renamed using normal SQL. It can be done carefully by the DBA playing
around with internal SYS dictionary tables and bouncing the database, but this is not supported. (I have
successfully done it in V4 through V7). Do backup the database first unless you feel brave. I've written a
quick and dirty script rncol.sql to do this. If you can't figure out how to use it from the source you definitely
should not run it. You can use a similar dirty trick for changing ownership of tables if storage space is
limited.
There are a number of "beautifiers" for various program languages. The CB and indent programs for the C
language spring to mind (although they have slightly different conventions). As far as I know there is no PD
formatter for SQL available.
Given that there are PD general SQL parsers and that the SQL standards are drafted in something close to
BNF, maybe someone could base a reformatted based on the grammar.
Note that you CANNOT use CB and indent with Pro *C as both these programs will screw up the
embedded SQL code.
I have recently heard that Kumaran Systems (see Vendor list) have a Forms PL/SQL and SQL formatter,
but I do not now if they have unbundled it.
You *know* there are records for that day - but none of them are coming back to you.
What has happened is that your records are not set to midnight (which is the default value if time of day not
specified)?
You can either use to_char and to_date functions, which can be a bad move regarding SQL performance,
or you can say
WHERE date_field >= '18-jun-60' AND date_field < '19-jun-60'
When converting to dates from characters when you only have two characters for the year, the picture
format "RR" will be interpreted as the year based on a guess that that date is between 1950 and 2049.
- 35 -
COMPLEX QUERIES
There are a number of tables/views beginnings with V$ that holds gory details for performance monitoring.
These are not guaranteed to be stable from minor release to minor release and are for DBAs only.
There are usually no real underlying tables (unlike SYS.OBJ$) and are dummied up by the RDBMS kernel
software in much the same way that UNIX System V.4 dummies up the files in the /proc or /dev/proc
directories.
If you have any code depending on these (and the widely used tools supplied by Oracle but unsupported
are in this category) then you need to verify that everything works each time you upgrade your database.
And when a major revision changes, all bets are off.
This question often gets the response WHERE ROWNUM <= 10 but this will not work (except accidentally)
because the ROWNUM pseudo column is generated before the ORDER or WHERE clauses come into
effect.
[email protected] (although it will be a bitch on a large table) suggested one elegant SQL-only approach
I do not believe that straight SQL is the way to go for such problems when you have PL/SQL available.
My approach is to use PL/SQL instead (in SQL*Plus):
Late news: index-descending hint to SQL works if you use a dummy restriction to force use of the index.
Needs V7, etc.
In SQL, you may need to control the rollback segment used as the default rollback segment may be too
small for the required transaction, or you may want to ensure that your transaction runs in a special
rollback segment, unaffected by others. The statement is as follows:
On a related note, if all you are doing are SELECTS, it is worth telling the database of this using the
following:
Say we are getting a list of names and codes and want it ordered by the name, using both EMP and DEPT
tables:
These three users are common in many databases. See the glossary entries under SCOTT, SCOTT and
SYS. Another common user/password is PLSQL/SUPERSECRET used for PL/SQL demo Stuff.
The simple answer is make sure you have them big enough and keep your transactions small, but that is
being a smartness. More recent versions of Oracle have an option for the session that you can set that
commits every so many DML statements. This is OK except for where you are doing your work in a single
statement rather than using PL/SQL and a loop construct.
Imagine you have a HUGE table and need to update it, possibly updating the key. You cannot update it in
one go because your rollback segments are too small. You cannot open a cursor and commit every n
records, because usually the cursor will close. You cannot have a number of updates of a few records
each because the keys may change - causing you to visit records more than once.
The solution I have used was to have one process select ROWID from the appropriate rows and pump
these (via standard I/O) to another process that looped around reading ROWIDs from standard input,
updating the appropriate record and committing every 10 records or so. This was very easy to program
and also was quite fast in execution. The number of locks and size of rollback segments required was
minimal.
If you are writing in Pro *C and use MODE=ORACLE, there are ways around it too, but not if you are using
MODE=ANSI.
OK, so this is really a DBA question, but it is worth putting in here because it involves SQL regardless of
interface.
First, look at the PASSWORD column in DBA_USERS. It looks like gobbledygook because it is an
encrypted password.However you can use this if you have saved it somewhere else. Say you want to
impersonate a user in a batch run overnight.First stash the gobbledygook password away
- 37 -
COMPLEX QUERIES
somewhere, grant connect to the user identified by some password you know and then run your batches
using the new known password.
To restore the password to what it was use the following syntax (which I think is undocumented).
How you organize your SQL and indices controls what access methods will be used. The following ranking
is valid for V6. I do not know about V7. QUERY PATH RANKING (lowest rank is the best) Rank Path
1 ROWID = constant
2 Unique index column(s) = constant(s)
3 Entire unique concatenated index = constant
4 Entire cluster key = corresponding key in another table in same cluster 5 Entire
cluster key = constant
6 Entire non-unique concatenated index = constant
7 Non-unique single column index merge
8 Most leading concatenated index = constant
9 Index column BETWEEN low AND hi or LIKE 'C%'
10 Sort/merge (joins only)
11 MAX/MIN of single indexed column
12 ORDER BY entire index
13 Full table scans
14 Unindexed column = constant or column IS NULL or column LIKE '%C%'
Views
---------
The typical example is the view on EMP limited to a department and not including salary. Also see CHECK
OPTION discussion.
Imagine we have created a view of EMP limited to a department, (where DEPTNO = 10). Now, that is fine
for querying, but you can still write records through this view (either by update or insert) with a value of 20
for DEPTNO. (Next time you query the view, such records will be invisible.)
Now if you want to stop someone doing this (and consider whether you want them to be able to do this or
not very carefully) use the "check option" when creating the view:
- 38 -
COMPLEX QUERIES
Because view queries that involve sorting, grouping, etc can lead to a high performance overhead, it might
be better to write some reports with a procedural component that fills up a temporary table and then does a
number of queries from it.
While this is non-relational, it can be justified for some cases. Nevertheless, it is useful to have the view
definition in the database. You can then test the output from the view against the output from your
procedural manipulations. The view definition can also be used as the unambiguous gospel.
There are a number of dictionary views that include the text of views. You can select these quite happily,
but remember, if using SQL*Plus to use the SET command to fiddle with ARRAYSIZE, MAXDATA and
LONG parameters
- 39 -
COMPLEX QUERIES
ALL_SYNONYMS
All synonyms accessible to the user
ALL_TABLES
Description of tables accessible to the user
ALL_TAB_COLUMNS
Columns of all tables, views and clusters
ALL_TAB_COMMENTS
Comments on tables and views accessible to the user
ALL_TAB_PRIVS
Grants on objects for which the user is the grantor, grantee, owner, or an enabled role or PUBLIC is the
grantee
ALL_TAB_PRIVS_MADE
User's grants and grants on user's objects
ALL_TAB_PRIVS_RECD
Grants on objects for which the user, PUBLIC or enabled role is the grantee
ALL_TRIGGERS
Triggers accessible to the current user
ALL_TRIGGER_COLS
Column usage in user's triggers or in triggers on user's tables
ALL_USERS
Information about all users of the database
ALL_VIEWS
Text of views accessible to the user
AUDIT_ACTIONS
Description table for audit trail action type codes. Maps action type numbers to action type names
CAT Synonym for USER_CATALOG
CLU Synonym for USER_CLUSTERS
COLS
Synonym for USER_TAB_COLUMNS
COLUMN_PRIVILEGES
Grants on columns for which the user is the grantor, grantee, owner, or an enabled role or PUBLIC is
the grantee
DBA_2PC_NEIGHBORS
information about incoming and outgoing connections for pending transactions
DBA_2PC_PENDING
info about distributed transactions awaiting recovery
DBA_ANALYZE_OBJECTS
DBA_AUDIT_EXISTS
Lists audit trail entries produced by AUDIT NOT EXISTS and AUDIT EXISTS
DBA_AUDIT_OBJECT
Audit trail records for statements concerning objects, specifically: table, cluster, view, index, sequence,
[public]
database link, [public] synonym, procedure, trigger, rollback segment, table space, role, user
DBA_AUDIT_SESSION
DBA_AUDIT_STATEMENT
Audit trail records concerning grant, revoke, audit, noaudit and alter system
DBA_AUDIT_TRAIL
All audit trail entries
DBA_CATALOG
All database Tables, Views, Synonyms, Sequences
DBA_CLUSTERS
Description of all clusters in the database
DBA_CLU_COLUMNS
Mapping of table columns to cluster columns
DBA_COL_COMMENTS
Comments on columns of all tables and views
DBA_COL_PRIVS
All grants on columns in the database
DBA_CONSTRAINTS
Constraint definitions on all tables
- 40 -
COMPLEX QUERIES
DBA_CONS_COLUMNS
Information about accessible columns in constraint definitions
DBA_DATA_FILES
Information about database files
DBA_DB_LINKS
All database links in the database
DBA_DEPENDENCIES
Dependencies to and from objects
DBA_ERRORS
Current errors on all stored objects in the database
DBA_EXP_FILES
Description of export files
DBA_EXP_OBJECTS
Objects that have been incrementally exported
DBA_EXP_VERSION
Version number of the last export session
DBA_EXTENTS
Extents comprising all segments in the database
DBA_FREE_SPACE
Free extents in all tablespaces
DBA_INDEXES
Description for all indexes in the database
DBA_IND_COLUMNS
COLUMNs comprising INDEXES on all TABLES and CLUSTERS
DBA_OBJECTS
All objects in the database
DBA_OBJECT_SIZE
Sizes, in bytes, of various PL/SQL objects
DBA_OBJ_AUDIT_OPTS
Auditing options for all tables and views
DBA_PRIV_AUDIT_OPTS
Describes current system privileges being audited across the system and by use r
DBA_PROFILES
Display all profiles and their limits
DBA_ROLES
All Roles which exist in the database
DBA_ROLE_PRIVS
Roles granted to users and roles
DBA_ROLLBACK_SEGS
Description of rollback segments
DBA_SEGMENTS
Storage allocated for all database segments
DBA_SEQUENCES
Description of all SEQUENCES in the database
DBA_SNAPSHOTS
All snapshots in the database
DBA_SNAPSHOT_LOGS
All snapshot logs in the database
DBA_SOURCE
Source of all stored objects in the database
DBA_STMT_AUDIT_OPTS
Describes current system auditing options across the system and by user
DBA_SYNONYMS
All synonyms in the database
DBA_SYS_PRIVS
System privileges granted to users and roles
DBA_TABLES
Description of all tables in the database
DBA_TABLESPACES
Description of all tablespaces
- 41 -
COMPLEX QUERIES
DBA_TAB_COLUMNS
Columns of all tables, views and clusters
DBA_TAB_COMMENTS
Comments on all tables and views in the database
DBA_TAB_PRIVS
All grants on objects in the database
DBA_TRIGGERS
All triggers in the database
DBA_TRIGGER_COLS
Column usage in all triggers
DBA_TS_QUOTAS
Tablespace quotas for all users
DBA_USERS
Information about all users of the database
DBA_VIEWS
Text of all views in the database
DICT Synonym for DICTIONARY
DICTIONARY
Description of data dictionary tables and views
DICT_COLUMNS
Description of columns in data dictionary tables and views
DUAL
GLOBAL_NAME
global database name
IND Synonym for USER_INDEXES
INDEX_HISTOGRAM
statistics on keys with repeat count
INDEX_STATS
statistics on the b-tree
OBJ Synonym for USER_OBJECTS
RESOURCE_COST
Cost for each resource
ROLE_ROLE_PRIVS
Roles which are granted to roles
ROLE_SYS_PRIVS
System privileges granted to roles
ROLE_TAB_PRIVS
Table privileges granted to roles
SEQ Synonym for USER_SEQUENCES
SESSION_PRIVS
Privileges which the user currently has set
SESSION_ROLES
Roles which the user currently has enabled.
SYN Synonym for USER_SYNONYMS
TABLE_PRIVILEGES
Grants on objects for which the user is the grantor, grantee, owner, or an enabled role or PUBLIC is the
grantee
TABS Synonym for USER_TABLES
USER_AUDIT_OBJECT
Audit trail records for statements concerning objects, specifically: table, cluster, view, index, sequence,
[public] database link, [public] synonym, procedure, trigger, rollback segment, tablespace, role, user
USER_AUDIT_SESSION
USER_AUDIT_STATEMENT
Audit trail records concerning grant, revoke, audit, noaudit and alter system
USER_AUDIT_TRAIL
Audit trail entries relevant to the user
USER_CATALOG
Tables, Views, Synonyms and Sequences owned by the user
USER_CLUSTERS
- 42 -
COMPLEX QUERIES
- 43 -
COMPLEX QUERIES
- 44 -
COMPLEX QUERIES
- 45 -
COMPLEX QUERIES
Structured Query Language (SQL) is a language that provides an interface to relational database systems.
SQL was developed by IBM in the 1970s for use in System R. SQL is a defacto standard, as well as an
ISO and ANSI standard. SQL is often pronounced SEQUEL.
- 46 -
COMPLEX QUERIES
In common usage SQL also encompasses DML (Data Manipulation Language), for INSERTs, UPDATEs,
DELETEs and DDL (Data Definition Language), used for creating and modifying tables and other database
structures.
The development of SQL is governed by standards. A major revision to the SQL standard was completed
in 1992 called SQL2. SQL3 supports object extensions and will be (partially?) implemented in Oracle8.
Choose one of the following queries to identify or remove duplicate rows from a table leaving one record:
Method 1:
DELETE FROM TABLE_NAME A WHERE ROWID > (SELECT MIN (ROWID) FROM TABLE_NAME B
WHERE A.KEY_VALUES = B.KEY_VALUES);
Method 2:
Method 3:
DELETE FROM MY_TABLE WHERE ROWID NOT IN (SELECT MAX(ROWID) FROM MY_TABLE
GROUP BY MY_COLUMN_NAME );
Method 4:
DELETE FROM MY_TABLE T1 WHERE EXISTS (SELECT 'X' FROM MY_TABLE T2 WHERE
T2.KEY_VALUE1 = T1.KEY_VALUE1 AND T2.KEY_VALUE2 = T1.KEY_VALUE2 AND T2.ROWID >
T1.ROWID);
Note: If you create an index on the joined fields in the inner loop, you for all intensive purposes eliminate
N^2 operations (no need to loop through the entire table on each pass by a record).
Create your table with a NOT NULL column (say SEQNO). This column can now be populated with unique
values:
How can I get the time difference between two date columns?
SELECT DEPT, SUM (DECODE (SEX,'M', 1,0)) MALE, SUM (DECODE (SEX,'F', 1,0)) FEMALE, COUNT
(DECODE (SEX,'M', 1,'F', 1)) TOTAL FROM MY_EMP_TABLE GROUP BY DEPT;
A value x will be between values y and z if GREATEST (x, y) = LEAST(x, z). Look at this example:
SELECT F2, COUNT (DECODE (GREATEST (F1, 59), LEAST (F1, 100), 1, 0)) "RANGE 60-100", COUNT
(DECODE (GREATEST (F1, 30), LEAST (F1, 59), 1, 0)) "RANGE 30-59", COUNT (DECODE (GREATEST
(F1, 29), LEAST (F1, 0), 1, 0)) "RANGE 00-29" FROM MY_TABLE GROUP BY F2;
For equal size ranges it might be easier to calculate it with DECODE (TRUNC (value/range), 0, rate_0, 1,
rate_1...).
Example:
SELECT ENAME "NAME", SAL "SALARY", DECODE (TRUNC (F2/1000, 0), 0, 0.0,1, 0.1, 2, 0.2, 3, 0.31)
"TAX RATE" FROM MY_TABLE;
SELECT * FROM TABLEX WHERE ROWID IN (SELECT ROWID FROM TABLEX WHERE ROWNUM <=
7
MINUS
SELECT ROWID FROM TABLEX WHERE ROWNUM < 5);
One can easily select all even, odd, or Nth rows from a table using SQL queries like this:
SELECT * FROM EMP WHERE (ROWID, 0) IN (SELECT ROWID, MOD (ROWNUM,4) FROM EMP);
SELECT * FROM (SELECT ROWNUM RN, EMPNO, ENAME FROM EMP) TEMP WHERE MOD
(TEMP.ROWNUM, 4) = 0;
SELECT * FROM MY_TABLE A WHERE 10 >= (SELECT COUNT (DISTINCT MAXCOL) FROM
MY_TABLE B WHERE B.MAXCOL >= A.MAXCOL) ORDER BY MAXCOL DESC;
This is definitely non-relational (enough to kill Codd and then make him roll in his grave) and is a feature I
have not seen in the competition.
- 48 -
COMPLEX QUERIES
The definitive example is in the example SCOTT/TIGER database, when looking at the EMP table
(EMPNO and MGR columns). The MGR column contains the employee number of the "current"
employee's boss.
You have available an extra pseudo-column, LEVEL, that says how deep in the tree you are. Oracle can
handle queries with a depth up to 255.
SELECT LEVEL, EMPNO, ENAME, MGR from EMP connect by prior EMPNO = MGR start with MGR is
NULL;
You can get an "indented" report by using the level number to sub-string or lpad a series of spaces and
concatenate that to the string.
You use the start with clause to specify the start of the tree(s). More than one record can match the starting
condition. One disadvantage of a "connect by prior" is that you cannot perform a join to other tables. Still, I
have not managed to see anything else like the "connect by prior" in the other vendor offerings and I like
trees. Even trying to doing this programmatic ally in embedded SQL is difficult as you have to do the top
level query, for each of them open a cursor to look for child nodes, for each of these open a cursor. Pretty
soon you blow the cursor limit for your installation.
The way around this is to use PL/SQL, open the driving cursor with the "connect by prior" statement, and
the select matching records from other tables on a row-by-row basis, inserting the results into a temporary
table for later retrieval.
The Oracle decode function acts like a procedural statement inside an SQL statement to return different
values or columns based on the values of other columns in the select statement.
Some examples:
SELECT DECODE (SEX, 'M', 'MALE', 'F', 'FEMALE', 'UNKNOWN') FROM EMPLOYEES;
SELECT A, B, DECODE (ABS (A-B), A-B, 'A > B', 0, 'A = B','A < B') FROM TABLEX;
SELECT DECODE (GREATEST (A, B), A, 'A IS GREATER THAN B', 'B IS GREATER THAN A')...
Note: The DECODE function is not ANSI SQL and are rarely implemented in other RDBMS offerings. It is
one of the good things about Oracle, but use it sparingly if portability is required.
How can one dump/ examine the exact content of a database column?
DUMP (COL1)
-------------------------------------
Typ=96 Len=4: 65,66,67,32
For this example the type is 96, indicating CHAR, and the last byte in the column is 32, which is the ASCII
code for a space. This tells us that this column is blank-padded.
Oracle does not provide a way to DROP a column (reference: Enhancement Request 51118). However,
Joseph S. Testa wrote a DROP COLUMN package that can be downloaded from
- 49 -
COMPLEX QUERIES
1. Rename t1 to t1_base;
Create view t1 <column list with new name> as select * from t1_base;
2. Create table t2 <column list with new name> as select * from t1;
Drop table t1;
Rename t2 to t1;
From Oracle8 you can just type "password" from SQL* Plus, or if you need to change another user's
password, type "password username".
How does one find the next value of a sequence?
Perform an "ALTER SEQUENCE ... NOCACHE" to unload the unused cached sequence numbers from
the Oracle library cache. This way, no cached numbers will be lost. If you then select from the
USER_SEQUENCES dictionary view, you will see the correct high water mark value that would be
returned for the next NEXTVALL call. Afterwards, perform an "ALTER SEQUENCE ... CACHE" to restore
caching.
You can use the above technique to prevent sequence number loss before a SHUTDOWN ABORT, or any
other operation that would cause gaps in sequence values.
You can use the SQL* Plus COPY command instead of snapshots if you need to copy LONG and LONG
RAW variables from one location to another. E.g.
COPY TO SCOTT/TIGER@REMOTE -
CREATE IMAGE_TABLE USING -
SELECT IMAGE_NO, IMAGE - FROM IMAGES;
Note: If you run Oracle8, convert your LONGs to LOBs, as it can be replicated.
PL/SQL is Oracle's Procedural Language extension to SQL. PL/SQL's language syntax, structure and data
types are similar to that of ADA. The language includes object oriented programming techniques
- 50 -
COMPLEX QUERIES
such as encapsulation, function overloading, information hiding (all but inheritance) and is commonly used
to write data-centric programs to manipulate Oracle data.
This is done via a standalone utility that transforms the PL/SQL source code into portable binary object
code (somewhat larger than the original). This way you can distribute software without having to worry
about exposing your proprietary algorithms and methods. SQL* Plus and SQL*DBA will still understand
and know how to execute such scripts. Just be careful, there is no "decode" command available.
One can use the DBMS_OUTPUT package to write information to an output buffer. This buffer can be
displayed on the screen from SQL* Plus if you issue the SET SERVEROUTPUT ON; command. For
example:
begin
dbms_output.put_line ('Look Ma, I can print from PL/SQL!!!');
end;
/
But what if you forget to set server output on? No problem, just type SET SERVEROUTPUT ON once you
remember, and then EXEC NULL; If you haven't cleared the DBMS_OUTPUT buffer with the disable or
enable procedure, SQL* Plus will display the entire contents of the buffer when it executes this dummy
PL/SQL block.
Included in Oracle 7.3 is an UTL_FILE package that can read and write operating system files. The
directory you intend writing to has to be in your INIT.ORA file (see UTL_FILE_DIR=...parameter).
Before Oracle 7.3 the only means of writing a file was to use DBMS_OUTPUT with the SQL* Plus SPOOL
command.
DECLARE
fileHandler UTL_FILE.FILE_TYPE;
BEGIN
fileHandler := UTL_FILE.FOPEN('/tmp', 'myfile', 'w');
UTL_FILE.PUTF(fileHandler, 'Look ma, I''m writing to a file!!!\n');
UTL_FILE.FCLOSE(fileHandler);
EXCEPTION
WHEN utl_file.invalid_path THEN
raise_application_error(-20000, 'ERROR: Invalid path for file or
path not in INIT.ORA.');
END;
From PL/SQL V2.1 one can use the DBMS_SQL package to execute dynamic SQL statements. E.g.:
- 51 -
COMPLEX QUERIES
cur integer;
rc integer;
BEGIN
cur := DBMS_SQL.OPEN_CURSOR;
DBMS_SQL.PARSE(cur, 'CREATE TABLE X (Y DATE)', DBMS_SQL.NATIVE);
rc := DBMS_SQL.EXECUTE(cur);
DBMS_SQL.CLOSE_CURSOR(cur);
END;
/
Another example:
How does one get the value of a sequence into a PL/SQL variable?
i := sq_sequence.NEXTVAL;
In Oracle7 there is no direct way to execute an operating system command from PL/SQL. However, one
can write an external program (using one of the pre-compiler languages, OCI or Perl with Oracle access
modules) to act as a listener on a DBMS_PIPE. Your PL/SQL then places requests to run commands in
the pipe, the listener picks it up and runs them. Results are passes back on a different pipe. For a Pro *C
example, see chapter 8 of the Oracle Application Developers Guide. This example is also on the Support
Notes CD-ROM that all supported customers should be getting quarterly. Just search on "DBMS_PIPE".
- 52 -
COMPLEX QUERIES
In Oracle8 you can call external 3GL code in a dynamically linked library (DLL or shared object). So you
just write a library in C doing what you want, ie in your case a host function taking the command line as
input argument. And that function will be callable from PL/SQL.
Let's say I had a PL/SQL block and I wanted to do a "ls -l" to get a directory listing. Is there any way to do
that?
In C Language, I can do
{
x = system("ls -l");
}
The way most people do that is to use the pipe facility of the DBMS kernel. Set up a pipe for use by a
sender (the PL/SQL program that needs to invoke a command) and a receiver, written in Pro *C. This is
responsible for executing the command piped by the sender. Maybe there are some tools around so one
doesn't need to code this, as it involves a listener and multiple server processes on your machine.
Alternatively I have a more complex solution, which uses a table to pass arguments and the command to
execute - just as the DBMS_PIPE package does internally. You would insert a row into the table and the
listener would execute a command, passing back succession status and indicating "in progress" on long-
running commands. This tool allows for non-blocking execution of commands.
DECLARE
CURSOR dept_cur IS
SELECT deptno
FROM dept
ORDER BY deptno;
-- Employee cursor all employees for a dept number
CURSOR emp_cur (v_dept_no DEPT.DEPTNO%TYPE) IS
SELECT ename
FROM emp
WHERE deptno = v_dept_no;
BEGIN
FOR dept_rec IN dept_cur LOOP
dbms_output.put_line('Employees in Department
'||TO_CHAR(dept_rec.deptno));
FOR emp_rec in emp_cur(dept_rec.deptno) LOOP
dbms_output.put_line('...Employee is '||emp_rec.ename);
END LOOP;
END LOOP;
END;
- 53 -
COMPLEX QUERIES
No. Unlike Oracle Forms, SQL* Plus does not have a PL/SQL engine. Thus, all your PL/SQL are send
directly to the database engine for execution. This makes it much more efficient as SQL statements are not
stripped off and send to the database individually.
Yes, the max size is not an explicit byte limit, but related to the parse tree that is created when you compile
the code.
You can run the following select statement to query the size of an existing package or procedure:
I switched the page size to 11x8.5, but the printer still prints in portrait.
Even though you set the page size in the report properties, there is a another variable in the system
parameters section under the data model in the object navigator called orientation. This sets the printer
orientation. Oracle starts by setting it to "default" which means that no matter how you set the page size,
the user's default printer setup will be used. You can also set it to either "Landscape" or "Portrait" to force
the printer orientation no matter what the user has set as default. These sorts of picky, minor details are
the ones, which are invariably forgotten when you are designing your report and are the reason I created
our two report templates, reptmp_p and reptmp_l (portrait and landscape). For anyone who wants a
consistent look in their reports I strongly recommend building a similar pair to save yourself an ulcer,
unless you actually like starting from scratch every time.
I moved this field into that repeating frame, but I'm still getting a "frequency below it's group" error.
Moving fields around does not change what enclosing object is considered it's parent group. Oracle
carefully remembers what repeating frame a field was originally placed in and assigns that as it's parent. If
you then reference a column further down the line of the query structure it will return that error. If you are
not exactly sure which repeating frame a field belongs to, try dragging it out of all of them. Whichever frame
will not allow it to escape is it's parent. To change a field's parent, first click on the lock button on the speed
button bar. It should now look like an unlocked padlock. Now all of the fields on the layout can be
repositioned regardless of their original parent items. When you are satisfied with the repositioning click the
lock button again to lock the layout. Oracle will parse the layout and assumes that any item fully enclosed
in a repeating frame is a child object of that frame. This can be confirmed again by trying to drag an object
out of it's parent. (CTRL - Z or EDIT…. UNDO will put it back where it came from)
Sometimes, for unknown and mysterious reasons, this method does not work. The alternative in this case
is to highlight the field (or fields), cut it (CTRL-X), and then paste it into the desired frame. The paste does
not initially set it into the right frame, but if you drag and drop it there before clicking on any other objects,
and then click on something else, Oracle will usually figure what your intent was and assign the object(s)
as a child of that frame. This is my preferred method of changing a field's parent as it works much more
consistently then the unlock/lock method. One note though, if you are reassigning a group of fields, make
sure the frame you are going to move them into is large enough to accept the whole group at once before
you do the cut/paste. If you do the paste and then try to grow the frame to fit, you will have to cut and paste
again. Once you de-select an object that has just been pasted, Oracle will assign it as a child of whatever it
is in at the time.
If this technique also fails, you are probably going to have to delete and then recreate the objects within
the desired frame. If the object has triggers attached, save yourself some typing by creating the new object
in the right frame, copying over the trigger code, and then deleting the old object.
I must put a repeating frame around these fields. How do I do this easily?
Well congratulations, you have just discovered one of the main reasons why good planning goes a long
way. Oracle looks at the layout as a sort of layered inheritance model such that anything created on top of
and completely inside another object is by definition a child of that object. Creation order is therefor critical
to the layout process. This means that placing a repeating frame on top of a field but
- 54 -
COMPLEX QUERIES
larger than that field fails the ownership criteria. At best, if the new frame is fully enclosed within the same
higher level frame as the field thenthe two will be considered sibling children of the higher level frame.
From this point you have two options. First, you can place the new repeating frame in the correct place and
then, use the techniques shown above in the "I moved this field but am still getting a frequency error" to
reassign the fields into the new frame. There is also a second choice (which can also be used as a solution
to the above). Go ahead and draw the new frame around the fields you want to have placed in it. Now if
you try to click on one of the fields you will not be able to as they are fully covered by the new frame. Now
go to the "Arrange" menu. You will find the options send to back, bring to front, move forwards and move
backwards. These are used to alter an object position in the Reports layer ordering. You use the "send
backwards" option to move the frame backwards until all of the fields have popped to the front and are now
enclosed in it. Oracle reassigns the new repeating frame as each object's parent as they pop to the front.
Note that you can only move an object back and forth amongst it's siblings. You cannot set it back below
its parent, or in front of its children. This means that once an object has popped to the front and had a
reassignment of parent, you cannot move it back using these tools.
Why does part of a row sometimes get shifted to the next page, but not all of it?
This is due to the way the scan works when Oracle is parsing the layout. If the tops of all the fields in a row
are aligned and the fields are all of the same height and font, they should all stay together. I suspect,
however, that Reports bases it's decision on the printed size rather than the field size you define to
determine which objects are too large and must be shifted to the next page. This means that even if you
set two fields top-aligned with the same height and font but one of them is bolded, the bolded field could
get shifted to the next page due to it's bigger footprint. The solution is to put the whole row into a regular
frame, which is page protected.
The print condition type First, All, All but first, Last, All but last refer to the frequency with which you want to
appear based upon the setting of the print condition object. A print condition object of Enclosing Object is
whichever object encloses the current object (could be the parent or a frame within the parent), while
Anchoring Object is the parent object (unless you have explicitly anchored the object in which case it is the
object to which it is anchored). The key here is that this is about the pages on which the Print Condition
Object appears, not the current object. Oracle views First as the first page on which any part of the Print
Condition Object is printed, likewise Last is the last page on which any part of the Print Condition Object is
printed. For objects inside a repeating frame, this condition is re-evaluated for each instance of the frame.
As an example, assume we have created a field inside a repeating frame with Print Condition Object set to
'anchoring object', and Print Condition Typeset to 'All But First'. On every instance of that repeating frame
which is printed entirely within a single page, our object will not print. However, if an instance of that frame
spans more than one page then our object will print on the second and every subsequent page that this
instance of the repeating frame spans.
For most objects you will not have to play with this print condition setting as the default setting is pretty
good at determining what pages to print on, even though it only chooses between 'first' and 'last'. Only
such things as heading objects you want reprinted on multiple pages are normally candidates for fooling
around with this setting.
How do I create a truly dynamic 'where' condition which the user can input on the parameter form for my
select statement
While setting a simple parameter for use in defining the select statement, such as a date, bill_period_id etc.
Is simple, there are times when you may wish to allow a user to add any "where" statement they wish.
However, if you create a VARCHAR user variable and try to reference it as an SQL condition ( e.g. Select *
from account where :user condition) you will get an error. The secret is that the variable must be initialized
to a valid SQL condition before the Data Model will accept it. This
- 55 -
COMPLEX QUERIES
is done in the "Initial Value" spot on the variable's property form. The usual default is "1 = 1" which simply
means all rows meeting whatever other conditions are included in the select statement will pass this
condition if the user does not change it in the parameter form.
Quite simply, you can't. Once the BeforeReport trigger has fired, reports locks down the user parameters
until the report is finished. Oh, I know you can put a statement into a layout trigger at design time and the
compiler will accept it, but the moment you run the report you will get a nasty error and the report will die.
Why they couldn't catch those problems at compile time I have no idea, except that it probably uses the
same PL/SQL compiler as Forms which uses that same syntax for the perfectly acceptable function of
changing field values.
That being said, there is valid technique to mimic having a user variable, which can be changed over the
course of the report execution. What you have to do is create a PL/SQL package that contains a variable
as well as the functions to read and write to that variable. Since variables inside a package are both local
to that package and persistent over the duration of the run, you use this to save and change your variable
value. I know that this seems like overkill, but it is the most efficient way of handling an issue that is very
rarely encountered. As you can probably guess, this technique is a last resort to finding an SQL work
around if one exists.
How do I set the initial values of parameters for the parameter form at runtime?
This is what the BeforeForm trigger is primarily used for. Even if you have used a select statement to
create a lookup list for the parameter, this statement is fully parsed before the parameter form is opened.
Simply setting the parameter to a given value in the BeforeForm trigger will select that option as the default
value displayed to the user. For example, assume you have a parameter called p_input_date, which is
intended to hold an invoice date. The following example will select the most recent invoice date as the
default, and note that it properly handles exceptions to ensure that the report does not arbitrarily die if this
default setting fails. Note also that like all report triggers, it must return a true or false value.
function BeforePForm return boolean is
begin
select max(bill_period_end_date + 1)
into :p_input_date
from billing_period
where bill_period_end_date <= (select trunc(sysdate) from dual);
return (TRUE);
exception
when others then
:p_input_date := null;
return true;
end;
Why can't I highlight a bunch of fields and change their entire format masks or prints conditions at once?
You can. If you highlight a bunch of objects and then right click and select "properties..", Oracle gives you a
stacked set of the individual properties forms for each of the selected objects. While this may be useful for
some things, it requires changing values individually for each object. However, instead you can select the
group of fields and then select "Common properties" from the "Tools" menu which will allow you to set the
format mask , print conditions etc. for the whole set of objects at once.
Triggers are intended to simply provide a true or false return value to determine whether an object should
be, printed. It is generally not allowed to change any values held in the cursor, make changes to the
database, or change the value of it's objects value. That being said, there is a highly unpublicized method
of doing just that using the SRW. Set_Field_Char procedure. The syntax is
- 56 -
COMPLEX QUERIES
SRW.Set_Field_char (0,) and the output of the object that the current trigger is attached to will be replaced
by. There are also RW.set_fileld_num and SRW.set_field_date for numeric or date fields.
While these options do work, they should only be used if a suitable NVL or DECODE statement in the
original query is not possible as they are much, much slower to run. Also, note that this change of value
only applies to the formatted output. It does not change the value held in the cursor and so can not be
used for evaluating summary totals.
SQL* Loader is a utility used for moving data from external files into the Oracle database. Its syntax is
similar to that of the DB2 Load utility, but comes with more options. SQL* Loader supports various load
formats, selective filters, and multi-table loads.
One load data into the Oracle database by using the sqlldr (sqlload on some platforms) utility. Look at the
following example:
load data
infile *
replace
into table departments
( dept position (02:05) char(4),
deptname position (08:27) char(20)
)
begindata
COSC COMPUTER SCIENCE
ENGL ENGLISH LITERATURE
MATH MATHEMATICS
POLY POLITICAL SCIENCE
Can I load variable and fix length data records?
Yes, look at the following control file examples. In the first we will load delimited data (variable length):
LOAD DATA
INFILE *
INTO TABLE load_delimited_data
FIELDS TERMINATED BY "," OPTIONALLY ENCLOSED BY '"'
TRAILING NULLCOLS
( data1,
data2
)
BEGINDATA
11111,AAAAAAAAAA
22222,"A,B,C,D,"
If you need to load positional data (fixed length), look at the following control file example:
LOAD DATA
INFILE *
INTO TABLE load_positional_data
( data1 POSITION(1:5),
data2 POSITION(6:15)
- 57 -
COMPLEX QUERIES
)
BEGINDATA
11111AAAAAAAAAA
22222BBBBBBBBBB
LOAD DATA
INFILE *
INTO TABLE modified_data
( rec_no "my_db_sequence.nextval",
region CONSTANT '31',
time_loaded "to_char(SYSDATE, 'HH24:MI')",
data1 POSITION(1:5) ":data1/100",
data2 POSITION(6:15) "upper(:data2)"
)
BEGINDATA
11111AAAAAAAAAA
22222BBBBBBBBBB
LOAD DATA
INFILE 'mail_orders.txt'
BADFILE 'bad_orders.txt'
APPEND
INTO TABLE mailing_list
FIELDS TERMINATED BY ","
( addr,
city,
state,
zipcode,
mailing_addr "decode(:mailing_addr, null, :addr,:mailing_addr)",
mailing_city "decode(:mailing_city, null, :city,:mailing_city)",
mailing_state
)
LOAD DATA
INFILE *
REPLACE INTO TABLE EMP
WHEN empno != ' '
( empno POSITION(1:4) INTEGER EXTERNAL,
ename POSITION(6:15) CHAR,
deptno POSITION(17:18) CHAR,
mgr POSITION(20:23) INTEGER EXTERNAL
)
INTO TABLE proj
WHEN projno != ' '
( projno POSITION(25:27) INTEGER EXTERNAL,
empno POSITION(1:4) INTEGER EXTERNAL
)
Can one selectively load only the data that you need?
Look at this example, (01) is the first character, (30:37) are characters 30 to 37:
- 58 -
COMPLEX QUERIES
LOAD DATA
APPEND INTO TABLE db_trace_19980517
WHEN (01) <> 'H' and (01) <> 'T' and (30:37) = '19980517'
(
region CONSTANT '31',
service_key POSITION(01:11) INTEGER EXTERNAL,
call_b_no POSITION(12:29) CHAR
)
One can create one logical record from multiple physical records using one of the following two clauses:
CONCATENATE: - use when SQL* Loader should add the same number of physical records together to
form one logical record.
CONTINUEIF - use if a condition indicates that multiple records should be treated as one, e.g. '#' in line 1.
How can get SQL* Loader to commit only at the end of the load file?
You can not, but by setting the ROWS= parameter to a large value, committing can be reduced. Make sure
you have big rollback segments ready when you use a high value for ROWS=.
1. A very simple but easily overlooked hint, do not have any indexes and/or constraints (primary key) on
your load tables during the load process. This will significantly slowdown load times even with ROWS=
set to a high value.
2. Add the following option in the command line: DIRECT=TRUE. This will effectively bypass most of the
RDBMS processing. However, there are cases when you can't use direct load. Refer to chapter 8 on
Oracle server Utilities manual.
3. Turn off database logging by specifying the UNRECOVERABLE option. This option can only be used
with direct data loads.
Oracle doesn't supply any data unload tools. However, you can use SQL* Plus to select and format your
data and then spool it to a file:
set echo off newpage 0 space 0 pagesize 0 feed off head off
trimspool on
spool oradata.txt
select col1 || ',' || col2 || ',' || col3
from tab1
where col2 = 'XYZ';
spool off
- 59 -
COMPLEX QUERIES
2.The user is granted the SYSDBA or SYSOPER previleges and database user the password files to
authenticate database administration.
3.The database has a passsword for the internal login.
oracle server
/\
Oracle database oracle instance
/\
Physical database structure Logical database structure
(An oracle database's physical structure is datermined by the O.S files that constitute database)
Oracle Instance
System global area(SGA) is an area of memory used for database admin. shared by the database
servers. The combination of the background process and memory buffers is called an instance.
Oracle Instance
/\
User process Oracle process
/\
server process background process
1. Which package construct must be declared and defined within the package body? •
- 60 -
COMPLEX QUERIES
• public procedure
• boolean variable
• exception
After creating this trigger, you test it by inserting a row into the PLAYER table. You receive this
Error message
The storage of the PLAYER table is so fragmented that the trigger cannot run successfully. Triggers
cannot change data in the primary key, foreign key, or unique key of a constraining table. (Answer)
References to sequences are not allowed in triggers.
Triggers, unless otherwise specified, are read only.
3. You have been granted the necessary privilege to invoke a function created by another
developer. You are having problems calling this function from your procedure & want to
contact the developer who created the function. Which table would you query to determine the
owner of this function?
• USER_OBJECTS
• ALL_FUNCTIONS
• ALL_OBJECTS (Answer)
• USER_FUNCTION_CODE
PROCEDURE UPD_PLAYER_STAT
(V_ID IN NUMBER, V_AB IN NUMBER DEFAULT 4, V_HITS IN NUMBER) IS
BEGIN
UPDATE PLAYER_BAT_STAT SET AT_BATS = AT_BATS + V_AB, HITS = HITS + V_HITS
WHERE PLAYER_ID = V_ID;
COMMIT;
END UPD_PLAYER_STAT;
PROCEDURE ADD_PLAYER
- 61 -
COMPLEX QUERIES
END BB_PACK;
/
• The outside procedure and the package body are invalidated. (Answer)
Answer:
CREATE OR REPLACE FUNCTION CALC_PLAYER_AVG
(V_ID IN NUMBER)
RETURN NUMBER
IS
V_AVG NUMBER;
BEGIN
SELECT HITS/AT_BATS
INTO V_AVG
FROM PLAYER_BAT_STAT
WHERE PLAYER_ID = V_ID;
RETURN (V_AVG);
END;
- 62 -
COMPLEX QUERIES
RETURN NUMBER
IS
V_AVG NUMBER;
BEGIN
SELECT HITS/AT_BATS
INTO V_AVG
FROM PLAYER_BAT_STAT
WHERE PLAYER_ID = V_ID;
RETURN (V_AVG);
END;
• PACKAGE BODY
• DATABASE TRIGGER
• USER_PROCEDURES
• USER_PROCS
• USER_OBJECTS (Answer)
• USER_PLSQL_UNITS
Which statement should be added if you want this trigger to execute just once for each INSERT
operation on the PLAYER table?
• WHEN (NEW.%ROWCOUNT = 1)
9. Which statement best describes the difference between auditing objects with triggers and auditing
objects within the server?
• Triggers can audit data manipulation, data retrieval, and data definition statements. •
Triggers can only audit data manipulation statements (Answer)
• Triggers can only audit data definition statements.
- 63 -
COMPLEX QUERIES
RETURN BOOLEAN
IS
V_AVG NUMBER;
BEGIN
SELECT HITS / AT_BATS
INTO V_AVG
FROM PLAYER_BAT_STAT
WHERE PLAYER_ID = V_ID;
IF V_AVG > .333 THEN
RETURN (TRUE);
ELSE
RETURN (FALSE);
END IF;
END;
• The 'RETURN' statement can only return VARCHAR2 and NUMBER datatypes. • The 'RETURN'
statement cannot return BOOLEAN datatypes if invoked from SQL* Plus. (Answer)
• The function contains two 'RETURN' statements. Only one is allowed.
PROCEDURE UPD_PLAYER_STAT
(V_ID IN NUMBER, V_AB IN NUMBER DEFAULT 4, V_HITS IN NUMBER) IS
BEGIN
UPDATE PLAYER_BAT_STAT SET AT_BATS = AT_BATS + V_AB, HITS = HITS + V_HITS
WHERE PLAYER_ID = V_ID;
COMMIT;
END UPD_PLAYER_STAT;
PROCEDURE ADD_PLAYER
(V_ID IN NUMBER, V_LAST_NAME VARCHAR2, V_SALARY NUMBER)
IS
BEGIN
INSERT INTO PLAYER(ID,LAST_NAME,SALARY) VALUES (V_ID, V_LAST_NAME,
V_SALARY);
UPD_PLAYER_STAT(V_ID,0,0);
END ADD PLAYER;
END BB PACK;
/
- 64 -
COMPLEX QUERIES
• The outside procedure and the package body are invalidated. (Answer)
12. One of your procedures must create a table during execution. Which Oracle supplied package
must you use to accomplish this task?
• DBMS_CREATE_OBJECT
• DBMS_SQL (Answer)
• DBMS_PIPE
• DBMS_TRANSACTION
Which two statements will successfully invoke this function in SQL* Plus? (Choose two.)
• EXECUTE TEAM_TAX(1)
PROCEDURE UPD_PLAYER_STAT
(V_ID IN NUMBER, V_AB IN NUMBER DEFAULT 4, V_HITS IN NUMBER) IS
BEGIN
UPDATE PLAYER_BAT_STAT
SET AT_BATS = AT_BATS + V_AB,
- 65 -
COMPLEX QUERIES
PROCEDURE ADD_PLAYER
(V_ID IN NUMBER, V_LAST_NAME VARCHAR2, V_SALARY NUMBER)
IS
BEGIN
INSERT INTO PLAYER(ID,LAST_NAME,SALARY)
VALUES (V_ID, V_LAST_NAME, V_SALARY);
UPD_PLAYER_STAT(V_ID,0,0);
END ADD_PLAYER;
END BB_PACK;
The V_PLAYER_AVG variable needs to be assigned a value determined from a calculation. You
want this value to be assigned only when the package is first invoked.
• PACKAGE SPECIFICATION
• PACKAGE EXCEPTION
PROCEDURE UPD_PLAYER_STAT
(V_ID IN NUMBER, V_AB IN NUMBER DEFAULT 4, V_HITS IN NUMBER) IS
BEGIN
UPDATE PLAYER_BAT_STAT
SET AT_BATS = AT_BATS + V_AB,
HITS = HITS + V_HITS
WHERE PLAYER_ID = V_ID;
COMMIT;
VALIDATE_PLAYER_STAT(V_ID);
END UPD_PLAYER_STAT;
PROCEDURE ADD_PLAYER
(V_ID IN NUMBER, V_LAST_NAME VARCHAR2, V_SALARY NUMBER)
IS
BEGIN
INSERT INTO PLAYER(ID,LAST_NAME,SALARY) VALUES (V_ID, V_LAST_NAME,
V_SALARY);
UPD_PLAYER_STAT(V_ID,0,0);
END ADD_PLAYER;
- 66 -
COMPLEX QUERIES
END BB_PACK;
SCOTT is responsible for updating the statistics of players. He cannot add players. Which
command must you issue to allow SCOTT the use of this package to accomplish his job?
• UPD_PLAYER_STAT cannot be executed from outside this package. Therefore, SCOTT cannot use
this package to accomplish his job. (Answer)
• GRANT EXECUTE ON BB_PACK.BODY.UPD_PLAYER_STAT TO SCOTT;
16. Your can enter new ballplayers to the PLAYER table from different Oracle Forms applications
and from an application written in C. For each new ballplayer, a record must be inserted into
the PLAYER_BAT_STAT table. Which action should you perform to accomplish this
requirement?
17. Which table and column can you query to see all procedures and functions that have been
marked invalid?
18. Your procedure references a function created by another application developer. You need to
contact this person to discuss the code contained in the function. Which table can you query to
determine the owner of this function?
• USER_DEPENDENCIES (Answer)
• USER_REFERENCES
• USER_CONSTRAINTS
• USER_LINKS
• Both the ADD_PLAYER and UPD_PLAYER_STAT procedures are marked invalid. (Answer) • The
ADD_PLAYER procedure is marked invalid.
• The UPD_PLAYER_STAT procedure is marked invalid.
- 67 -
COMPLEX QUERIES
PROCEDURE ADD_PLAYER
(V_ID IN NUMBER, V_LAST_NAME VARCHAR2, V_SALARY NUMBER)
IS
BEGIN
INSERT INTO PLAYER(ID,LAST_NAME,SALARY)
VALUES (V_ID, V_LAST_NAME, V_SALARY);
UPD_PLAYER_STAT(V_ID,0,0);
END ADD_PLAYER;
END BB_PACK;
SCOTT is responsible for adding new ball players to the system. Which command must you
issue to allow SCOTT the use of this package to accomplish his job?
Which set of commands must be added to handle the non-predefined error: ORA-02292?
• STATS_EXIST_EXCEPTION NUMBER;
PRAGMA EXCEPTION_INIT(STATS_EXISTS_EXCEPTION, -2292);
- 68 -
COMPLEX QUERIES
• STATS_EXIST_EXCEPTION EXCEPTION;
PRAGMA EXCEPTION_INIT(-2292, STATS_EXISTS_EXCEPTION);
• STATS_EXIST_EXCEPTION EXCEPTION;
PRAGMA EXCEPTION_INIT(STATS_EXISTS_EXCEPTION, -2292);
• STATS_EXIST_EXCEPTION EXCEPTION;
PRAGMA EXCEPTION_INIT(STATS_EXISTS_EXCEPTION, 2292);
23. Which statement will successfully create the procedure ADD_PLAYER in SQL* Plus?
(Answer)
CREATE OR REPLACE PROCEDURE ADD_PLAYER
(V_ID IN NUMBER, V_LAST_NAME VARCHAR2)
IS
BEGIN
INSERT INTO PLAYER (ID,LAST_NAME) VALUES (V_ID, V_LAST_NAME);
COMMIT;
END;
PROCEDURE ADD_PLAYER
(V_ID IN NUMBER, V_LAST_NAME VARCHAR2)
IS
BEGIN
INSERT INTO PLAYER (ID,LAST_NAME) VALUES (V_ID, V_LAST_NAME);
COMMIT;
END;
PROCEDURE UPD_PLAYER_STAT
(V_ID IN NUMBER, V_AB IN NUMBER DEFAULT 4, V_HITS IN NUMBER) IS
BEGIN
- 69 -
COMPLEX QUERIES
UPDATE PLAYER_BAT_STAT
SET AT_BATS = AT_BATS + V_AB,
HITS = HITS + V_HITS
WHERE PLAYER_ID = V_ID;
COMMIT;
END UPD_PLAYER_STAT;
PROCEDURE ADD_PLAYER
(V_ID IN NUMBER, V_LAST_NAME VARCHAR2, V_SALARY NUMBER)
IS
BEGIN
INSERT INTO PLAYER(ID,LAST_NAME,SALARY) VALUES (V_ID, V_LAST_NAME,
V_SALARY);
UPD_PLAYER_STAT(V_ID,0,0);
END ADD_PLAYER;
END BB_PACK;
Which statement will successfully execute the ADD_PLAYER procedure from within SQL*
Plus?
• EXECUTE ADD_PLAYER(37,'Nettles',500000);
• RUN BB_PACK.ADD_PLAYER(37,'Nettles',500000);
PROCEDURE UPD_PLAYER_STAT
(V_ID IN NUMBER, V_AB IN NUMBER DEFAULT 4, V_HITS IN NUMBER) IS
BEGIN
UPDATE PLAYER_BAT_STAT
SET AT_BATS = AT_BATS + V_AB,
HITS = HITS + V_HITS
WHERE PLAYER_ID = V_ID;
COMMIT;
END UPD_PLAYER_STAT;
PROCEDURE ADD_PLAYER
(V_ID IN NUMBER, V_LAST_NAME VARCHAR2, V_SALARY NUMBER)
IS
BEGIN
INSERT INTO PLAYER(ID,LAST_NAME,SALARY) VALUES (V_ID, V_LAST_NAME,
V_SALARY);
UPD_PLAYER_STAT(V_ID,0,0);
END ADD_PLAYER;
END BB_PACK;
- 70 -
COMPLEX QUERIES
Which statement will successfully execute the UPD_PLAYER_STAT procedure from within
SQL* Plus?
• UPD_PLAYER_STAT(31, 4, 2);
26. Adding ball players to the PLAYER table is limited to the baseball season. You decide to create
a database trigger.
Which trigger timing should you use to prevent users from inserting into this table during the off
season?
on-change
after
before
on-insert
• Testing of procedures and functions requires the database to be restarted to clear out shared SQL
areas for future access.
28. Which statement must be added to make this trigger execute after updating the SALARY
column of the PLAYER table?
30. Which code is stored in the database when a procedure or function is created in SQL* Plus?
- 71 -
COMPLEX QUERIES
• only P-CODE
31. The ADD_PLAYER procedure residing in your local database in Austin executes the
UPD_PLAYER_STAT procedure residing in Chicago. The procedure in Chicago has just been
changed and recompiled.
• An error will occur next time the ADD_PLAYER procedure is executed. (Answer) • There will
A user informs you that she is getting an error whenever she inserts a row into the PLAYER
table. After researching the problem, you realize that the database trigger on the PLAYER table
is inserting a row into the PLAYER_BAT_STAT table. The PLAYER_BAT_STAT table is offline.
Which action should you take to allow users to insert into the PLAYER table until the
PLAYER_BAT_STAT table is online?
• Alter the trigger, changing it from a row level to a statement level trigger.
Which two statements will successfully invoke this function in SQL* Plus? (Choose two.)
• EXECUTE CALC_PLAYER_AVG(31);
• SELECT CALC_PLAYER_AVG(31)
FROM PLAYER_BAT_STAT;
• VARIABLE G_AVG NUMBER
EXECUTE :G_AVG := CALC_PLAYER_AVG(31);
- 72 -
COMPLEX QUERIES
• CALC_PLAYER_AVG(31);
• START CALC_PLAYER_AVG(31);
34. For every new ballplayer added to the PLAYER table, a record must be inserted into the
PLAYER_BAT_STAT table. You have written a trigger to accomplish this task.
• PRE-INSERT
• AFTER (Answer)
• BEFORE
• Create a SQL* Plus script file that contains each procedure's CREATE statement • Create a package
specification. Place the name of each procedure in the new specification. • Create a package
specification and body with the source code of each procedure. Keep the stand-alone procedures for
the specification to reference.
• Create a package specification and body with the source code of each procedure. Drop the old
stand-alone procedures. (Answer)
36. The ADD_PLAYER procedure calls the UPD_PLAYER_STAT procedure. Both procedures are
INVALID.
Which command can you issue to recompile both procedures?
SCOTT needs the privilege to invoke this procedure. Which set of privileges must you issue?
- 73 -
COMPLEX QUERIES
UPDATE PLAYER_BAT_STAT
SET AT_BATS = AT_BATS + V_AB,
HITS = HITS + V_HITS
WHERE PLAYER_ID = V_ID;
COMMIT;
END;
Which two statements will successfully invoke this procedure from SQL* Plus? (Choose two.)
• EXECUTE UPD_BAT_STAT
• EXECUTE UPD_BAT_STAT(31)
• EXECUTE UPD_BAT_STAT(31,'FOUR','TWO')
• EXECUTE UPD_BAT_STAT(31,4,2)
39. The ADD_PLAYER procedure must calculate the amount of social security tax that the team needs to
pay. This algorithm is already written in a C program. You decide to call this program from the
ADD_PLAYER procedure.
Which Oracle Procedure Builder built-in package must you use to accomplish this task?
• ORA_PROF
• ORA_FFI (Answer)
• TOOL_ENV
• STPROC
Why?
• This is a statement level trigger and therefore cannot reference :new. (Answer) • This
is a row level trigger and therefore cannot reference :new.
• This is a statement level trigger and therefore cannot perform data manipulation statements. • This
is a row level trigger and therefore cannot perform data manipulation statements.
41. Using the debugger in Procedure Builder, which action must you take to temporarily halt the
execution of a procedure?
SQL*PLUS QUESTIONS
- 74 -
COMPLEX QUERIES
3. User can have 500 many number of lines (assuming 80 characters per line) per SQL command. 4.
5. Start command is used to run the contents of the specified command file.
6. The intersect operator is used to get only those rows that returned by both the query. 7.
The Grand command is used to set the System privileges, Object privileges.
8. The Savepoint command is used to identify the point in a transaction to which you can later Rollback.
10. The to-char function is used to convert the number datatype to a value of varchar2 datatype.
11. The Truncate command is used to remove all rows in a Table or Cluster instantly.
Note : We can not truncate rows from a table, which is part of a cluster.
We cannot truncate rows from a table, which has a referenced integrity constraint.
12. The Cluster is a schema object that contains one or more tables that have one or more columns in
common.
13. The Create Role command is used to set a set of privileges that can be granted to users or to other
roles.
* To collect statistics about the object used by the optimizer and store them in the data dictionary. * To
delete statistics about the object from the data dictionary.
* To validate the structure of the object.
* To identify migrated and chained rows of the table or cluster.
a] 255
b] 21
c] 16
d] None of the above
Ans: 16
a] 255
b] 254
c] 030
d] None of the above
- 75 -
COMPLEX QUERIES
Ans: 254
3. The maximum number of components in the Decode expression, including searches, results and default
is
a] No limitation
b] 255
Ans: 255
4. ___ Is an alternative name for a table, view, sequence, procedure, stored function, package, snapshot
or another synonym.
a] Synonym
b] Data block
c] View
d] None of the above
Ans: Synonym
5. The _________ operator is used in character string comparisons with pattern matching
a] Between.. And
b] Equal operator
c] Set operator
d] Like
Ans: Like
6. _________ returns only one copy of each set of duplicate rows selected.
a] Unique
b] Distinct
c] Group By
d] None of the above
a] Lock table
b] For update of
c] Object privileges
d] Row share
8. _____ Clause restricts the groups of rows returned to those groups for the specified condition id True
a] Where clause
b] Having Clause
c] Distinct
d] Exists
Ans: Having
- 76 -
a] Connect by start with Ans: Connect by start with
b] Order by COMPLEX QUERIES
10. The ________ function is used to return the number of bytes in the internal representation of
expression
a] Length
b] Vsize
c] LengthLB
d] None of the above
Ans: Vsize
A Database can be one of the two definitions: • A set of dictionary tables and user
What is a Database ?
• One or more operating system files in which ORACLE stores the tables, views, and
other objects, also, the set of database objects used by a given application.
What is a Database system? A Database system is a combination of an Instance and a Database. If the instance
is started and connected to an open database, then the database is available for
access by the users.
Note:-
A DBMS must be able to reliably manage a large amount of data in a multi-
user environment so that many users can concurrently access the same data.
What is an RDBMS ? A relational database Management System (RDBMS) is a computer program for
general purpose data storage and retrieval that organizes data into tables consisting
of one or more units of information (rows), each containing the same set of data
items (columns). ORACLE is a relational database management system.
• Relational.
- 77 -
What is SQL ? a relational database and used in ORACLE and IBM DB2
relational database management systems. SQL is formally
COMPLEX QUERIES pronounced “sequel”, although common usage also
pronounces it “SQL”
What is SQL*PLUS ? SQL*PLUS is the ORACLE database language which includes ANSI standard SQL
commands plus additional commands for accessing data in ORACLE database.
What is PL/SQL ? It is a Procedural Language extension of SQL. It can contain any no of SQL
statements integrated with flow of control statements. Thus it combine the Data
Manipulating power of SQL with data processing power of Procedural language.
What is A DDL statements? DDL statements are one category of SQL statements. DDL statements define
(create) or delete (drop) database objects. Examples are create view, create table,
create index, drop table and rename table. The other categories are DML statements
and DCL statements.
- 78 -
What is a DML statements ? DML statements are one category of SQL statements. DML
statements, such as select, insert, delete and update, query
COMPLEX QUERIES and update the actual data. The other categories are DDL
statements and DCL statements.
What are DCL statements ? DML statements are one category of SQL statements. DCL statements such as,
connect, grant select, grant update and revoke DBA, control access to the data and
to the database. The other categories are DDL and DML statements.
What is a Commit ?
• COMMIT commits any changes made to the database since the last COMMIT was
executed implicitly or explicitly. WORK is optional and has no effect on usage.
What is a Rollback ?
• A ROLLBACK discards part or all of the work you have done in the current
transaction, since the last COMMIT or SAVEPOINT.
What is locking ? To lock is to temporarily restrict other user’s access to data. The restriction is placed
on such data is called “a lock”. The modes are SHARE, SHARE
UPDATE,EXCLUSIVE,ROW SHARE AND ROW EXCLUSIVE. Not all locks can be
acquired in all modes.
EXCLUSIVE locks permit users to query the locked table but not to do anything else.
No other user may lock the table.
SHARED locks permit concurrent queries but no updates to the locked table. With a
ROW SHARE or SHARE UPDATE lock, no users can lock the whole table for
exclusive access, allowing concurrent access for all users to the table. The two types
of locks are synonymous, and SHARE UPDATE exists for compatibility with previous
versions of ORACLE.
ROW EXCLUSIVE locks are similar to ROW SHARE but they prohibit shared
locking, so only one user may access the table at the same time.
- 79 -
What is SHARE UPDATE A SHARE UPDATE lock is one that permits other users to both query and lock data.
LOCK ?
What is EXCLUSIVE LOCK ? An EXCLUSIVE LOCK is one that permits other users to query data, but not to
change it. It differs from the SHARE lock because it does not permit another user to
place any type of lock on the same data; several users may place SHARE locks on
the same data at the same time.
What is a ROW SHARE With a ROW SHARE or SHARE UPDATE lock, no users can lock the whole table for
LOCK ? exclusive access, allowing concurrent access for all users to the table.
What is a ROW EXCLUSIVE ROW EXCLUSIVE locks are similar to ROW SHARE but they prohibit shared
LOCK ? locking, so only one user may access the table at the same time.
What is a DEAD LOCK ? A DEAD lock is a rare situation in which two or more user processes of a database
cannot complete their transactions. This occurs because each process is holding a
resource that the other process requires (such as a row in a table) in order to
complete. Although these situations occur rarely, ORACLE detects and resolves
deadlocks by rolling back the work of one of the processes.
What are INTEGRITY INTEGRITY CONSTRAINT is a rule that restricts the range of valid values for a
CONSTRAINTS ? column, it is placed on a column when the table is created.
What is REFERENTIAL REFERENTIAL INTEGRITY is the property that guarantees that values from one
INTEGRITY ? column depend on values from another column. This property is enforced through
integrity constraints.
What is a PRIMARY KEY ? The PRIMARY KEY is the column(s) used to uniquely identify each row of a table.
What is a FOREIGN KEY ? A FOREIGN KEY is one or more columns whose values are based on the PRIMARY
or CANDITATE KEY values from the database.
What is a UNIQUE KEY ? A UNIQUE KEY is one or more columns that must be unique for each row of the
table.
What is the difference The UNIQUE KEY column restricts entry of duplicate values but entry of NULL value
between UNIQUE and is allowed.
PRIMARY KEY ? In case of PRIMARY KEY columns entry of duplicate as well as NULL value is
restricted.
- 80 -