MacLochlainns Weblog

Sometimes I give parts of a solution to increase the play time to solve a problem. I didn’t anticipate a problem when showing how to perform a sort operation with a CASE statement. It’s a sweet solution when you need to sort something differently than a traditional ascending or descending sort.

I gave my students this ORDER BY clause as an example:

  CASE
    WHEN filter = 'Debit' THEN 1
    WHEN filter = 'Credit' THEN 2
    WHEN filter = 'Total' THEN 3
  END;

It raises the following error in MySQL for students:

ERROR 1064 (42000): You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'ORDER BY
  CASE
    WHEN filter = 'Debit' THEN 1
    WHEN filter = 'Credit' THEN' at line 6

It raises the following error in Oracle for some students:

  CASE
  *
ERROR AT line 7:
ORA-01785: ORDER BY item must be the NUMBER OF a SELECT-list expression

So, I built a little test case to replicate the problem and error message they encountered:

SQL> SELECT 'Debit' AS filter FROM dual
  2  UNION ALL
  3  SELECT 'Credit' AS filter FROM dual
  4  UNION ALL
  5  SELECT 'Total' AS filter FROM dual
  6  ORDER BY
  7    CASE
  8      WHEN filter = 'Debit' THEN 1
  9      WHEN filter = 'Credit' THEN 2
 10      WHEN filter = 'Total' THEN 3
 11    END;

They said, great but how can you fix it? That’s simple, with a Common Table Expression (CTE) in Oracle or with an inline view in MySQL. The Oracle CTE solution is:

  1  WITH results AS
  2  (SELECT 'Debit' AS filter FROM dual
  3   UNION ALL
  4   SELECT 'Credit' AS filter FROM dual
  5   UNION ALL
  6   SELECT 'Total' AS filter FROM dual)
  7  SELECT filter
  8  FROM   results
  9  ORDER BY
 10    CASE
 11	 WHEN filter = 'Debit'  THEN 1
 12	 WHEN filter = 'Credit' THEN 2
 13	 WHEN filter = 'Total'  THEN 3
 14    END;

There are two MySQL solutions. One simply removes the FROM dual clauses from the query components and the other uses an inline view in the FROM clause. This is the inline view:

SELECT filter
FROM  (SELECT 'Debit' AS filter FROM dual
       UNION ALL
       SELECT 'Credit' AS filter FROM dual
       UNION ALL
       SELECT 'Total' AS filter FROM dual) resultset
ORDER BY
  CASE
    WHEN filter = 'Debit' THEN 1
    WHEN filter = 'Credit' THEN 2
    WHEN filter = 'Total' THEN 3
  END;

This is the solution without the FROM dual clauses:

SELECT 'Debit' AS filter
UNION ALL
SELECT 'Credit' AS filter
UNION ALL
SELECT 'Total' AS filter
ORDER BY
  CASE
    WHEN filter = 'Debit' THEN 1
    WHEN filter = 'Credit' THEN 2
    WHEN filter = 'Total' THEN 3
  END;

Both MySQL solutions yield the following:

+--------+
| filter |
+--------+
| Debit  |
| Credit |
| Total  |
+--------+
3 rows in set (0.00 sec)

It puts the fabricating query inside a result set, and then lets you use the column alias to filter the set. If you have a better approach, please share it.

This how you install SQL Developer on Mac OS Yosemite. The first thing you need to do is download and install Java 8, not Java 7 on your Mac OS Yosemite as suggested on some web sites. You can determine whether or not Java is installed by running the following command:

Mac-Pro-3:~ username$ java -version
No Java runtime present, requesting install.

You must accept the Java license to install Java 8 on the Mac OS X operating system:

You have the option of installing the Java SDK or JDK. I’ve opted to install Netbeans 8 with JDK 8u45, as you can tell from the screen capture after you launched the file:

It is a standard Mac OS installation, which is why I didn’t bother showing any dialog messages. After installing the Java JDK or SDK, you should download SQL Developer 4.1 from Oracle’s web site. Below is a screen shot of the Oracle download web page where I’ve accepted the license agreement:

If you attempt to launch the installation and you’ve set your Mac Security to the “Mac App Store and identified developers” setting, you should raise the following exception:

If you reset the Mac Security to an “Anywhere” setting, you can install Oracle SQL Developer on Yosemite. Just make sure you reset it to the “Mac App Store and identified developers” setting after you install SQL Developer.

If you launch SQL Developer with the Security “Anywhere” setting, it displays the following dialog:

After you launch the program, you will see the following progress dialog:

The last step of the installation launches SQL Developer, as shown below:

Click the Connections icon to create an initial connection, like the following:

After connecting to the database, you can write and execute a query as shown in the next screen capture:

As always, I hope that this helps those who require an example to install SQL Server on a Mac OS.

Last week, I wrote about how to use bash arrays and the MySQL database to create unit and integration test scripts. While the MySQL example was nice for some users, there were some others who wanted me to show how to write bash shell scripts for Oracle unit and integration testing. That’s what this blog post does.

If you don’t know much about bash shell, you should start with the prior post to learn about bash arrays, if-statements, and for-loops. In this blog post I only cover how to implement a bash shell script that runs SQL scripts in silent mode and then queries the database in silent mode and writes the output to an external file.

I’ve copied the basic ERD for the example because of a request from a reader. In their opinion it makes cross referencing the two posts unnecessary.

To run the bash shell script, you’ll need the following SQL files, which you can see by clicking not he title below. There are several differences. For example, Oracle doesn’t support a DROP IF EXISTS syntax and requires you to write anonymous blocks in their PL/SQL language; and you must explicitly issue a QUIT; statement even when running in silent mode unlike MySQL, which implicitly issues an exit.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

-- Drop actor table and actor_s sequence.
BEGIN
  FOR i IN (SELECT   object_name
            ,        object_type
            FROM     user_objects
            WHERE    object_name IN ('ACTOR','ACTOR_S')) LOOP
    IF    i.object_type = 'TABLE' THEN
      EXECUTE IMMEDIATE 'DROP TABLE ' || i.object_name || ' CASCADE CONSTRAINTS';
    ELSIF i.object_type = 'SEQUENCE' THEN
      EXECUTE IMMEDIATE 'DROP SEQUENCE ' || i.object_name;
    END IF;
  END LOOP;
END;
/
 
-- Create an actor table.
CREATE TABLE actor
( actor_id    NUMBER CONSTRAINT actor_pk PRIMARY KEY
, actor_name  VARCHAR(30)  NOT NULL );
 
-- Create an actor_s sequence.
CREATE SEQUENCE actor_s;
 
-- Insert two rows.
INSERT INTO actor VALUES (actor_s.NEXTVAL,'Chris Hemsworth');
INSERT INTO actor VALUES (actor_s.NEXTVAL,'Chris Pine');
INSERT INTO actor VALUES (actor_s.NEXTVAL,'Chris Pratt');
 
-- Quit session.
QUIT;

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

-- Drop film table and film_s sequence.
BEGIN
  FOR i IN (SELECT   object_name
            ,        object_type
            FROM     user_objects
            WHERE    object_name IN ('FILM','FILM_S')) LOOP
    IF    i.object_type = 'TABLE' THEN
      EXECUTE IMMEDIATE 'DROP TABLE ' || i.object_name || ' CASCADE CONSTRAINTS';
    ELSIF i.object_type = 'SEQUENCE' THEN
      EXECUTE IMMEDIATE 'DROP SEQUENCE ' || i.object_name;
    END IF;
  END LOOP;
END;
/
 
-- Create a film table.
CREATE TABLE film
( film_id    NUMBER CONSTRAINT film_pk PRIMARY KEY
, film_name  VARCHAR(30)  NOT NULL );
 
-- Create an actor_s sequence.
CREATE SEQUENCE film_s;
 
-- Insert four rows.
INSERT INTO film VALUES (film_s.NEXTVAL,'Thor');
INSERT INTO film VALUES (film_s.NEXTVAL,'Thor: The Dark World');
INSERT INTO film VALUES (film_s.NEXTVAL,'Star Trek');
INSERT INTO film VALUES (film_s.NEXTVAL,'Star Trek into Darkness');
INSERT INTO film VALUES (film_s.NEXTVAL,'Guardians of the Galaxy');
 
-- Quit session.
QUIT;

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81

-- Drop movie table and movie_s sequence.
BEGIN
  FOR i IN (SELECT   object_name
            ,        object_type
            FROM     user_objects
            WHERE    object_name IN ('MOVIE','MOVIE_S')) LOOP
    IF    i.object_type = 'TABLE' THEN
      EXECUTE IMMEDIATE 'DROP TABLE ' || i.object_name || ' CASCADE CONSTRAINTS';
    ELSIF i.object_type = 'SEQUENCE' THEN
      EXECUTE IMMEDIATE 'DROP SEQUENCE ' || i.object_name;
    END IF;
  END LOOP;
END;
/
 
-- Create an movie table.
CREATE TABLE movie
( movie_id   NUMBER  CONSTRAINT movie_pk   PRIMARY KEY
, actor_id   NUMBER  CONSTRAINT movie_nn1  NOT NULL
, film_id    NUMBER  CONSTRAINT movie_nn2  NOT NULL
, CONSTRAINT actor_fk FOREIGN KEY (actor_id)
  REFERENCES actor (actor_id)
, CONSTRAINT film_fk  FOREIGN KEY (film_id)
  REFERENCES film(film_id));
 
-- Create table constraint.
CREATE SEQUENCE movie_s;
 
-- Insert translation rows.
INSERT INTO movie
VALUES
( movie_s.NEXTVAL
,(SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Hemsworth')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Thor'));
 
INSERT INTO movie
VALUES
( movie_s.NEXTVAL
,(SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Hemsworth')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Thor: The Dark World'));
 
INSERT INTO movie
VALUES
( movie_s.NEXTVAL
,(SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Pine')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Star Trek'));
 
INSERT INTO movie
VALUES
( movie_s.NEXTVAL
,(SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Pine')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Star Trek into Darkness'));
 
INSERT INTO movie
VALUES
( movie_s.NEXTVAL
,(SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Pratt')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Guardians of the Galaxy'));
 
-- Quit session.
QUIT;

1
2
3
4
5
6
7
8
9

-- Set Oracle column width.
COL table_name FORMAT A30 HEADING "Table Name"
 
-- Query the tables.
SELECT   table_name
FROM     user_tables;
 
-- Exit SQL*Plus.
QUIT;

1
2
3
4
5
6
7
8
9
10
11

-- Format query.
COL film_actors FORMAT A40 HEADING "Actors in Films"
 
-- Diagnostic query.
SELECT   a.actor_name || ', ' || f.film_name AS film_actors
FROM     actor a INNER JOIN movie m
ON       a.actor_id = m.actor_id INNER JOIN film f
ON       m.film_id = f.film_id;
 
-- Quit the session.
QUIT;

If you don’t have a sample test schema to use to test this script, you can create a sample schema with the following create_user.sql file. The file depends on the existence of a users and temp tablespace.

Click the link below to see the source code for a script that let’s you create a sample user account as the system user:

1
2

SET LONG 200000
SELECT dbms_metadata.get_ddl('TABLESPACE','USERS') FROM dual;

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

-- Drop the sample user table.
BEGIN
  FOR i IN (SELECT   username
            FROM     dba_users
            WHERE    username = 'SAMPLE') LOOP
    EXECUTE IMMEDIATE 'DROP USER ' || i.username || ' CASCADE';
  END LOOP;
END;
/
 
-- Create the sample user table.
CREATE USER sample IDENTIFIED BY sample
  DEFAULT TABLESPACE users
  TEMPORARY TABLESPACE temp
  QUOTA 50M ON users;
 
-- Grant privileges to sample user.
GRANT CREATE CLUSTER, CREATE INDEXTYPE, CREATE OPERATOR
,     CREATE PROCEDURE, CREATE SEQUENCE, CREATE SESSION
,     CREATE TABLE, CREATE TRIGGER, CREATE TYPE
,     CREATE VIEW TO sample;

The following list_oracle.sh shell script expects to receive the username, password, and fully qualified path in that specific order. The script names are entered manually in the array because this should be a unit test script.

This is an insecure version of the list_oracle.sh script because you provide the password on the command line. It’s better to provide the password as you run the script.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39

#!/usr/bin/bash
 
# Assign user and password
username="${1}"
password="${2}"
directory="${3}"
 
echo "User name:" ${username}
echo "Password: " ${password}
echo "Directory:" ${directory}
 
# Define an array.
declare -a cmd
 
# Assign elements to an array.
cmd[0]="actor.sql"
cmd[1]="film.sql"
cmd[2]="movie.sql"
 
# Call the array elements.
for i in ${cmd[*]}; do
  sqlplus -s ${username}/${password} @${directory}/${i} > /dev/null
done
 
# Connect and pipe the query result minus errors and warnings to the while loop.
sqlplus -s ${username}/${password} @${directory}/tables.sql 2>/dev/null |
 
# Read through the piped result until it's empty.
while IFS='\n' read actor_name; do
  echo $actor_name
done
 
# Connect and pipe the query result minus errors and warnings to the while loop.
sqlplus -s ${username}/${password} @${directory}/result.sql 2>/dev/null |
 
# Read through the piped result until it's empty.
while IFS='\n' read actor_name; do
  echo $actor_name
done

The IFS (Internal Field Separator) works with whitespace by default. The IFS on lines 29 and 37 sets the IFS to a line return ('\n'). That’s the trick to display the data, and you can read more about the IFS in this question and answer post.

You can run the shell script with the following syntax:

./list_oracle.sh sample sample /home/student/Code/bash/oracle > output.txt

You can then display the results from the output.txt file with the following command:

cat output.txt command:

It will display the following output:

User name: sample
Password:  sample
Directory: /home/student/Code/bash/oracle
 
Table Name
------------------------------
MOVIE
FILM
ACTOR
 
Actors in Films
----------------------------------------
Chris Hemsworth, Thor
Chris Hemsworth, Thor: The Dark World
Chris Pine, Star Trek
Chris Pine, Star Trek into Darkness
Chris Pratt, Guardians of the Galaxy

As always, I hope this helps those looking for a solution.

A reader posted A dynamic level limiting hierarchical query about Oracle’s hierarchical queries. They wanted to know how to capture only the hierarchy above the level where the first leaf node occurs. They gave me the following hierarchy map as an example:

               1                                    2
        +-------------+                       +-----------+
        |             |                       |           |      
        3             5                       4           6
    +---------+    +-----------+           +-----+    +------+
    |         |    |           |           |     |    |      |
    7         9    11          13          8     10   12     14
+-----+   +-----+  +--+    +-------+                       +-----+ 
|     |   |     |     |    |       |                       |     |
15    17  19    21    23   27      29                     16     18
                                                                 +---+
                                                                     |
                                                                     20

You can find the node values and hierarchical level with the following query:

SELECT   tt.child_id
,        LEVEL
FROM     test_temp tt
WHERE    CONNECT_BY_ISLEAF  = 1
START
WITH     tt.parent_id IS NULL 
CONNECT
BY PRIOR tt.child_id = tt.parent_id
ORDER BY 2;

We really don’t need the node values to solve the problem. We only need the lowest LEVEL value returned by the query, which is 3. The combination of the MIN and CONNECT_BY_ISLEAF functions let us solve this problem without writing a PL/SQL solution. The subquery returns the lowest level value, which is the first level where a leaf node occurs.

SELECT   LPAD(' ', 2*(LEVEL - 1)) || tt.child_id AS child_id
FROM     test_temp tt
WHERE    LEVEL <= (SELECT   MIN(LEVEL)
                   FROM     test_temp tt
                   WHERE    CONNECT_BY_ISLEAF  = 1
                   START
                   WITH     tt.parent_id IS NULL 
                   CONNECT
                   BY PRIOR tt.child_id = tt.parent_id)
START
WITH     tt.parent_id IS NULL
CONNECT
BY PRIOR tt.child_id = tt.parent_id;

It returns:

               1                                    2
        +-------------+                       +-----------+
        |             |                       |           |      
        3             5                       4           6
    +---------+    +-----------+           +-----+    +------+
    |         |    |           |           |     |    |      |
    7         9    11          13          8     10   12     14

While I answered the question in a comment originally, it seemed an important trick that should be shared in its own post.

This is the continuation of my efforts to stage an awesome Fedora developer’s instance. It shows you how to install Java 1.8 software development kit, which is nice to have. Though you can’t use Java 1.8 officially with Oracle SQL Developer 4.0.3 it is required for Oracle SQL Developer 4.1. Fortunately, the Oracle Product Manager, Jeff Smith has advised us that you can use Java 1.8 JDK with Oracle SQL Developer 4.0.3, and he’s written a comment to the blog post that it runs better with the Java 1.8 SDK.

After you install Oracle SQL Developer 4.0.3 or Oracle SQL Developer 4.1, you can watch Jeff Smith’s YouTube Video on SQL Developer 3.1 to learn how to use the basics of SQL Developer. I couldn’t find an updated version of the video for SQL Developer 4 but I didn’t try too hard.

You use yum as the root user to install Java SDK 1.8, much like my earlier Installing the Java SDK 1.7 and Java-MySQL Sample Program. The following command installs Java 8:

yum install -y java-1.8*

It produces the following output:

Loaded plugins: langpacks, refresh-packagekit
fedora/20/x86_64/metalink                                   |  18 kB  00:00     
mysql-connectors-community                                  | 2.5 kB  00:00     
mysql-tools-community                                       | 2.5 kB  00:00     
mysql56-community                                           | 2.5 kB  00:00     
pgdg93                                                      | 3.6 kB  00:00     
updates/20/x86_64/metalink                                  |  16 kB  00:00     
updates                                                     | 4.9 kB  00:00     
(1/2): mysql-tools-community/20/x86_64/primary_db           |  21 kB  00:00     
(2/2): updates/20/x86_64/primary_db                         |  13 MB  00:09     
updates/20/x86_64/pkgtags
updates
(1/2): updates/20/x86_64/pkgtags                            | 1.4 MB  00:02     
(2/2): updates/20/x86_64/updateinfo                         | 1.9 MB  00:04     
Package 1:java-1.8.0-openjdk-headless-1.8.0.31-1.b13.fc20.x86_64 already installed and latest version
Package 1:java-1.8.0-openjdk-javadoc-1.8.0.31-1.b13.fc20.noarch already installed and latest version
Resolving Dependencies
--> Running transaction check
---> Package java-1.8.0-openjdk.x86_64 1:1.8.0.31-1.b13.fc20 will be installed
---> Package java-1.8.0-openjdk-accessibility.x86_64 1:1.8.0.31-1.b13.fc20 will be installed
---> Package java-1.8.0-openjdk-demo.x86_64 1:1.8.0.31-1.b13.fc20 will be installed
---> Package java-1.8.0-openjdk-devel.x86_64 1:1.8.0.31-1.b13.fc20 will be installed
---> Package java-1.8.0-openjdk-src.x86_64 1:1.8.0.31-1.b13.fc20 will be installed
--> Finished Dependency Resolution
 
Dependencies Resolved
 
================================================================================
 Package                          Arch   Version                  Repository
                                                                           Size
================================================================================
Installing:
 java-1.8.0-openjdk               x86_64 1:1.8.0.31-1.b13.fc20    updates 201 k
 java-1.8.0-openjdk-accessibility x86_64 1:1.8.0.31-1.b13.fc20    updates  12 k
 java-1.8.0-openjdk-demo          x86_64 1:1.8.0.31-1.b13.fc20    updates 1.9 M
 java-1.8.0-openjdk-devel         x86_64 1:1.8.0.31-1.b13.fc20    updates 9.2 M
 java-1.8.0-openjdk-src           x86_64 1:1.8.0.31-1.b13.fc20    updates  45 M
 
Transaction Summary
================================================================================
Install  5 Packages
 
Total download size: 56 M
Installed size: 92 M
Downloading packages:
(1/5): java-1.8.0-openjdk-accessibility-1.8.0.31-1.b13.fc20 |  12 kB  00:00     
(2/5): java-1.8.0-openjdk-1.8.0.31-1.b13.fc20.x86_64.rpm    | 201 kB  00:02     
(3/5): java-1.8.0-openjdk-demo-1.8.0.31-1.b13.fc20.x86_64.r | 1.9 MB  00:03     
(4/5): java-1.8.0-openjdk-devel-1.8.0.31-1.b13.fc20.x86_64. | 9.2 MB  00:07     
(5/5): java-1.8.0-openjdk-src-1.8.0.31-1.b13.fc20.x86_64.rp |  45 MB  05:05     
--------------------------------------------------------------------------------
Total                                              187 kB/s |  56 MB  05:05     
Running transaction check
Running transaction test
Transaction test succeeded
Running transaction (shutdown inhibited)
  Installing : 1:java-1.8.0-openjdk-1.8.0.31-1.b13.fc20.x86_64              1/5 
  Installing : 1:java-1.8.0-openjdk-devel-1.8.0.31-1.b13.fc20.x86_64        2/5 
  Installing : 1:java-1.8.0-openjdk-demo-1.8.0.31-1.b13.fc20.x86_64         3/5 
  Installing : 1:java-1.8.0-openjdk-accessibility-1.8.0.31-1.b13.fc20.x86   4/5 
  Installing : 1:java-1.8.0-openjdk-src-1.8.0.31-1.b13.fc20.x86_64          5/5 
  Verifying  : 1:java-1.8.0-openjdk-devel-1.8.0.31-1.b13.fc20.x86_64        1/5 
  Verifying  : 1:java-1.8.0-openjdk-demo-1.8.0.31-1.b13.fc20.x86_64         2/5 
  Verifying  : 1:java-1.8.0-openjdk-1.8.0.31-1.b13.fc20.x86_64              3/5 
  Verifying  : 1:java-1.8.0-openjdk-accessibility-1.8.0.31-1.b13.fc20.x86   4/5 
  Verifying  : 1:java-1.8.0-openjdk-src-1.8.0.31-1.b13.fc20.x86_64          5/5 
 
Installed:
  java-1.8.0-openjdk.x86_64 1:1.8.0.31-1.b13.fc20                               
  java-1.8.0-openjdk-accessibility.x86_64 1:1.8.0.31-1.b13.fc20                 
  java-1.8.0-openjdk-demo.x86_64 1:1.8.0.31-1.b13.fc20                          
  java-1.8.0-openjdk-devel.x86_64 1:1.8.0.31-1.b13.fc20                         
  java-1.8.0-openjdk-src.x86_64 1:1.8.0.31-1.b13.fc20                           
 
Complete!

Then, you go to Oracle’s SQL Developer 4.0.3 web page or Oracle’s Beta SQL Developer 4.1 web page and download the SQL Developer RPM. At the time of writing, you download the following SQL Developer 4.0.3 RPM:

sqldeveloper-4.0.3.16.84-1.noarch.rpm

Assuming you download the sqldeveloper-4.0.3.16.84-1.noarch.rpm file to the student user’s account. It will download into the /home/student/Downloads directory. You run the SQL Developer RPM file with the following syntax as the root user:

rpm -Uhv /home/student/Downloads/sqldeveloper-4.0.3.16.84-1.noarch.rpm

Running the SQL Developer RPM produces the following output:

Preparing...                          ################################# [100%]
Updating / installing...
   1:sqldeveloper-4.0.3.16.84-1       ################################# [100%]

You can now run the sqldeveloper.sh file as the root user with the following syntax:

/opt/sqldeveloper/sqldeveloper.sh

At this point, it’s important to note that my download from the Oracle SQL Developer 4.1 page turned out to be SQL Developer 4.0.3. It prompts you for the correct Java JDK, as shown below. You may opt to enter the path to the Java JDK 1.8 for SQL Developer 4.1 because until today you downloaded the Oracle SQL Developer 4.0.3 version from the Oracle SQL Developer 4.1 page. Naturally, the Oracle SQL Developer 4.1 instructions say to use the Java 1.8 JDK on the RPM for Linux Installation Notes web page, as shown below:

If you assume from the instructions on the Oracle instruction page above that Oracle SQL Developer 4.0.3 and Oracle SQL Developer 4.1 support Java 1.8 JDK, you may enter the location for the Java JDK 1.8 when prompted. Jeff Smith, the Product Manager wrote this blog post on Oracle SQL Developer 4: Windows and the JDK. Unfortunately, you’ll see the following message if you attempt to run Oracle SQL Developer 4.0.3 with the Java 1.8 SDK at the command-line:

 Oracle SQL Developer
 Copyright (c) 1997, 2014, Oracle and/or its affiliates. All rights reserved.
 
Type the full pathname of a JDK installation (or Ctrl-C to quit), the path will be stored in /root/.sqldeveloper/4.0.0/product.conf
/usr/lib/jvm/java-1.8.0-openjdk-1.8.0.31.x86_64
OpenJDK 64-Bit Server VM warning: ignoring option MaxPermSize=256M; support was removed in 8.0

It also raises the following error message dialog:

Text version of Unsupported JDK Version error message:

The error dialog message tells us that the instructions on the RPM for Linux Installation Notes web page can be misleading. You really need to use the Java JDK 1.7 to be supported officially, but you can safely ignore the error.

If you want a certified component, leave the “Skip This Message Next Time” checkbox unchecked and click the “No” button to continue. At this point, there’s no automatic recovery. You need to open the following file:

/root/.sqldeveloper/4.0.0/product.conf

You need to change the SetJavaHome parameter in the file to the following:

# SetJavaHome /path/jdk
SetJavaHome /usr/lib/jvm/java-1.7.0-openjdk-1.7.0.79-2.5.5.0.fc20.x86_64

After making the change, you can re-run the sqldeveloper.sh shell as follows:

/opt/sqldeveloper/sqldeveloper.sh

It launches the following dialog message:

The installation pauses to ask you if you want to transfer an existing SQL Developer configuration by raising the following dialog. Assuming this is a new installation, the installer won’t find a prior configuration file. You need to click the “No” button to proceed.

The installation continues and launches SQL Developer. The first time launch shows you the following Oracle Usage Tracking dialog. If you don’t want your use monitored, uncheck the “Allow automated usage reporting to Oracle” checkbox. Click the “OK” button to continue.

After dismissing the Oracle Usage Tracking dialog, you see the SQL Developer environment:

After installing SQL Developer in the root account, you can install it as the student user. You use this command as the student user:

/opt/sqldeveloper/sqldeveloper.sh

It returns the following error because it’s the second installation and SQL Developer doesn’t prompt you to configure the user’s product.conf file with the working JDK location:

 Oracle SQL Developer
 Copyright (c) 1997, 2014, Oracle and/or its affiliates. All rights reserved.
 
Type the full pathname of a JDK installation (or Ctrl-C to quit), the path will be stored in /home/student/.sqldeveloper/4.0.0/product.conf
Error:  Unable to get APP_JAVA_HOME input from stdin after 10 tries

You need to edit the /home/student/.sqldeveloper/4.0.0/product.conf file, and add the following line to the file:

# SetJavaHome /path/jdk
SetJavaHome /usr/lib/jvm/java-1.7.0-openjdk-1.7.0.79-2.5.5.0.fc20.x86_64

Now, you can launch SQL Developer with the following command:

/opt/sqldeveloper/sqldeveloper.sh

Alternatively, you can add the following alias to the student user’s .bashrc file:

# Set alias for SQL Developer tool.
alias sqldeveloper="/opt/sqldeveloper/sqldeveloper.sh"

You can now launch the SQL Developer tool, like this as the student user:

sqldeveloper

You see the following when SQL Developer launches:

As always, I hope this helps those trying to sort out installing SQL Developer on a Fedora server.

Back in January 2014, I wrote a script to cleanup an Oracle student schema. It worked well until I started using APEX 4 in my student schema. You create the following 75 objects when you create an APEX 4 schema.

OBJECT TYPE    TOTAL
------------ -------
TABLE		  17
INDEX		  28
SEQUENCE	   5
TRIGGER 	  14
LOB		   9
FUNCTION	   2

Here’s the modified script that ignores the objects created automatically by Oracle APEX when you create a student workspace:

BEGIN
  FOR i IN (SELECT    object_name
            ,         object_type
            ,         last_ddl_time
            FROM      user_objects
            WHERE     object_name NOT IN
                       ('APEX$_WS_WEBPG_SECTION_HISTORY','APEX$_WS_WEBPG_SECTIONS_T1'
                       ,'APEX$_WS_WEBPG_SECTIONS_PK','APEX$_WS_WEBPG_SECTIONS'
                       ,'APEX$_WS_WEBPG_SECHIST_IDX1','APEX$_WS_TAGS_T1'
                       ,'APEX$_WS_TAGS_PK','APEX$_WS_TAGS_IDX2','APEX$_WS_TAGS_IDX1'
                       ,'APEX$_WS_TAGS','APEX$_WS_ROWS_T1','APEX$_WS_ROWS_PK'
                       ,'APEX$_WS_ROWS_IDX','APEX$_WS_ROWS','APEX$_WS_NOTES_T1'
                       ,'APEX$_WS_NOTES_PK','APEX$_WS_NOTES_IDX2','APEX$_WS_NOTES_IDX1'
                       ,'APEX$_WS_NOTES','APEX$_WS_LINKS_T1','APEX$_WS_LINKS_PK'
                       ,'APEX$_WS_LINKS_IDX2','APEX$_WS_LINKS_IDX1','APEX$_WS_LINKS'
                       ,'APEX$_WS_HISTORY_IDX','APEX$_WS_HISTORY','APEX$_WS_FILES_T1'
                       ,'APEX$_WS_FILES_PK','APEX$_WS_FILES_IDX2','APEX$_WS_FILES_IDX1'
                       ,'APEX$_WS_FILES','APEX$_ACL_T1','APEX$_ACL_PK','APEX$_ACL_IDX1'
                       ,'APEX$_ACL','CUSTOM_AUTH','CUSTOM_HASH','DEPT','EMP'
                       ,'UPDATE_ORDER_TOTAL')
            AND NOT ((object_name LIKE 'DEMO%' OR
                      object_name LIKE 'INSERT_DEMO%' OR
                      object_name LIKE 'BI_DEMO%') AND
                      object_type IN ('TABLE','INDEX','SEQUENCE','TRIGGER'))
            AND NOT (object_name LIKE 'SYS_LOB%' AND object_type = 'LOB')
            AND NOT (object_name LIKE 'SYS_C%' AND object_type = 'INDEX')
            ORDER BY object_type DESC) LOOP
 
    /* Drop types in descending order. */
    IF i.object_type = 'TYPE' THEN
 
      /* Drop type and force operation because dependencies may exist. Oracle 12c
         also fails to remove object types with dependents in pluggable databases
         (at least in release 12.1). Type evolution works in container database
         schemas. */
      EXECUTE IMMEDIATE 'DROP '||i.object_type||' '||i.object_name||' FORCE';
 
    /* Drop table tables in descending order. */
    ELSIF i.object_type = 'TABLE' THEN
 
      /* Drop table with cascading constraints to ensure foreign key constraints
         don't prevent the action. */
      EXECUTE IMMEDIATE 'DROP '||i.object_type||' '||i.object_name||' CASCADE CONSTRAINTS PURGE';
 
      /* Oracle 12c ONLY: Purge the recyclebin to dispose of system-generated
         sequence values because dropping the table doesn't automatically 
         remove them from the active session.
         CRITICAL: Remark out the following when working in Oracle Database 11g. */
      EXECUTE IMMEDIATE 'PURGE RECYCLEBIN';
 
    ELSIF i.object_type = 'LOB' OR i.object_type = 'INDEX' THEN
 
      /* A system generated LOB column or INDEX will cause a failure in a
         generic drop of a table because it is listed in the cursor but removed
         by the drop of its table. This NULL block ensures there is no attempt
         to drop an implicit LOB data type or index because the dropping the
         table takes care of it. */
      NULL;
 
    ELSE
 
      dbms_output.put_line('DROP '||i.object_type||' '||i.object_name||';');
      /* Drop any other objects, like sequences, functions, procedures, and packages. */
      EXECUTE IMMEDIATE 'DROP '||i.object_type||' '||i.object_name;
 
    END IF;
  END LOOP;
END;
/

As always, I hope this helps others.

The following walks you through how you sign on to a STUDENT Workspace with Oracle’s APEX product. It shows you how to create a new table with the Object Browser tool.

You can find instructions on how to create your own STUDENT Workspace in this blog post. Overall, Oracle APEX is a valuable tool to learn and master.

1. You start the process by accessing the Oracle Database 11g APEX, which you can access at http://localhost:8080/apex by default on the server. If you’ve got a static IP address for your instance, you can replace localhost with the IP address or hostname for the IP address.
   • Workspace: STUDENT
   • Username:  ADMIN
   • Password:  STUDENT

2. After you login to the STUDENT workspace, you have four options. They are the: Application Builder, SQL Workshop, Team Development, and Administration. You start the process by accessing the Oracle Database 11g APEX, which you can access at http://localhost:8080/apex by default on the server. If you’ve got a static IP address for your instance, you can replace localhost with the IP address or hostname for the IP address. Click on the Object Browser icon to proceed.

3. Clicking the SQL Workshop icon takes you to the second level menu. You click the Object Browser icon to create a database object.

4. After clicking the Object Browser icon, you see the screen at the left. Click the Create button to create a table.

5. After clicking the Create button, you see the screen at the left. Click the type of database object that you want to create. In our case, we click the Table hypertext to start the create table workflow.

6. After clicking the Table hyperlink, you see the Create Table screen at the left. Enter the column names, choose their data types and set the scale and precision. You should also check the Not Null checkbox when you want a column to be mandatory. Click the Next button to continue the create table workflow.

7. After entering the column names, you should choose the data types, enter the scale and precision, and check the NOT NULL checkbox to make appropriate columns mandatory by applying NOT NULL database constraints. If you run out of entry rows, you can click the Add Column button to add new rows. Click the Next button to continue the create table workflow when you’ve defined the columns.

8. After defining the column names, you should choose whether the primary key will use a new sequence or an existing sequence. You also have the ability to not assign a primary key value or simply leave it unpopulated when inserting new rows. The example creates an IMAGE_PK primary key constraint on the IMAGE_ID column, and declares an IMAGE_SEQ sequence value. Click the Next button to continue the create table workflow when you’ve defined the primary key constraint and any new sequence value for the primary key column.

9. After defining the primary key constraint, you can define foreign key column constraints. You enter a foreign key constraint name, choose between a Disallow Delete, Cascade Delete, or Set Null on Delete rule, select the foreign key column, the foreign key’s referenced table and column. Click the Add button to continue the create table workflow.

10. After defining a foreign key constraint, you can see the constraint that you created. Then, you can define another foreign key column constraints. You repeat the steps from the prior steps to add another foreign key constraint. Click the Add button to create a second foreign key constraint and complete the create table workflow.

11. After defining a second foreign key constraint, you see the following two foreign key constraints. Click the Next button to complete the create table workflow.

12. After defining all the foreign key constraints, you can create check and unique constraints. You check a radio button for a check or unique constraint, and then you select the columns for the constraint’s key. Click the /Add button to create any check or unique constraints as part of the create table workflow.

13. After defining all check and unique key constraints, you can see them in the Constraints box. Click the Next button to complete the create table workflow.

14. After defining all items about the table, you can see the SQL to create the IMAGE table and its constraints. You can copy the SQL into a file for later use when writing a re-runnable script. Click the Create button to complete the create table workflow and create the table.

     

    The following are the contents of the script for the actions you’ve defined:

    CREATE table "IMAGE" (
        "IMAGE_ID"         NUMBER NOT NULL,
        "FILE_NAME"        VARCHAR2(60) NOT NULL,
        "MIME_TYPE"        NUMBER NOT NULL,
        "ITEM_IMAGE"       BLOB,
        "CREATED_BY"       NUMBER NOT NULL,
        "CREATION_DATE"    DATE NOT NULL,
        "LAST_UPDATED_BY"  NUMBER NOT NULL,
        "LAST_UPDATE_DATE" DATE NOT NULL,
        constraint  "IMAGE_PK" primary key ("IMAGE_ID")
    )
    /
    
    CREATE sequence "IMAGE_SEQ" 
    /
    
    CREATE trigger "BI_IMAGE"  
      before insert on "IMAGE"              
      for each row 
    begin  
      if :NEW."IMAGE_ID" is null then
        select "IMAGE_SEQ".nextval into :NEW."IMAGE_ID" from dual;
      end if;
    end;
    /   
    
    ALTER TABLE "IMAGE" ADD CONSTRAINT "IMAGE_FK1" 
    FOREIGN KEY ("CREATED_BY")
    REFERENCES "SYSTEM_USER" ("SYSTEM_USER_ID")
    
    /
    ALTER TABLE "IMAGE" ADD CONSTRAINT "IMAGE_FK2" 
    FOREIGN KEY ("LAST_UPDATED_BY")
    REFERENCES "SYSTEM_USER" ("SYSTEM_USER_ID")
    
    /
    alter table "IMAGE" add
    constraint "IMAGE_UK1" 
    unique ("FILE_NAME","MIME_TYPE")
    /   
    

15. After creating the table, trigger, sequence, and constraints, you can see the table definition. You also have the ability to modify the table. At this point, you can create another structure or you can click the Home or SQL Workshop menu choice.

As always, I hope this helps those looking to learn new things and approaches.

The following walks through how you sign on to a STUDENT Workspace with Oracle’s APEX product and write and run free-form SQL statements. You can find instructions on how to create your own STUDENT Workspace.

While this blog introduces several concepts and features of Oracle APEX, it only focuses on how to write and run free-form SQL statements. Overall, Oracle APEX is a valuable tool to learn and master.

1. You start the process by accessing the Oracle Database 11g APEX, which you can access at http://localhost:8080/apex by default on the server. If you’ve got a static IP address for your instance, you can replace localhost with the IP address or hostname for the IP address.
   • Workspace: STUDENT
   • Username:  ADMIN
   • Password:  STUDENT

2. After you login to the STUDENT workspace, you have four options. They are the: Application Builder, SQL Workshop, Team Development, and Administration. You start the process by accessing the Oracle Database 11g APEX, which you can access at http://localhost:8080/apex by default on the server. If you’ve got a static IP address for your instance, you can replace localhost with the IP address or hostname for the IP address. Click on the SQL Workshop icon to proceed.
   • Application Builder: Let’s you build custom APEX applications.
   • SQL Workshop: Let’s you work with custom SQL, and APEX provides you with the following utilities:
     • Object Browser: Lets you create tables, views, and other objects.
     • SQL Commands: Lets you run individual SQL statements inside a browser window and returns results in the bottom pane.
     • SQL Scripts: Lets you create, upload, delete, and run scripts from the browser.
     • Query Builder: Lets you create free form queries that include joins between tables, but limits you to primary to foreign key table relationships. That means you can’t write range joins with a cross join and the BETWEEN operator and you can’t write self-joins.
     • Utilities: Lets you work with the Data Workshop (imports and exports data), Object Reports (a SQL report writer tool), Generate DDL (a tool that creates structures in the database), User Interface Defaults (coordinate data dictionary), Schema Comparison (a tool to compare similarities between schemas, About Database (the ability to connect as the database administrator), and Recycle Bin (dropped and purged structures).
   • Team Development: A project management tool.
   • Administration: Lets you manage database services, users and groups, monitor activities, and dashboards. You should note that the SQL query doesn’t have a semicolon like it would in a SQL*Plus environment. The Run button acts as the execution operator and effectively replaces the role of the semicolon, which traditionally executes a statement.

3. Clicking the SQL Workshop icon takes you to the second level menu. You click the SQL Commands icon to enter a free-form SQL statement. Click on the SQL Commands icon to proceed.

4. The first text panel lets you enter free-form queries. The Autocommit checkbox is enabled, which means the result of INSERT and UPDATE statements are immediate and don’t require a COMMIT statement. The second text panel displays results from a query or acknowledgment of statement completion.

5. This screen shot shows a query in the first panel and the results of the query in the second panel.

As always, I hope this helps those looking to learn new things and approaches.

My students asked if you could embed an OFFSET x ROWS FETCH NEXT y ROWS ONLY clause in a SQL Server T-SQL user-defined function. The answer is no, it isn’t Oracle (yes, you can do that in Oracle Database 12c with an NDS statement). There’s an example in Chapter 2 of my Oracle Database 12c PL/SQL Programming book if you’re interested. I also demonstrate a different approach to SQL Server T-SQL table functions in this older post. However, an attempt to add the clause to a SQL Server T-SQL function, like this:

CREATE FUNCTION studentdb.getBatch
(@rows   AS INT
,@offset AS INT) RETURNS @output TABLE
( marvel_id    INT 
, avenger_name VARCHAR(30)
, first_name   VARCHAR(20)
, last_name    VARCHAR(20)) AS
BEGIN
  /* Insert the results into the table variable. */
  INSERT @output
  SELECT marvel_id
  ,      avenger_name
  ,      first_name
	,      last_name
  FROM   studentdb.marvel
  OFFSET (@offset - 1) ROWS FETCH NEXT @rows ROWS ONLY;
 
  /* Return the table variable from the function. */  
  RETURN;
END;

Throws the following errors trying to compile the function:

Msg 102, Level 15, State 1, Procedure getBatch, Line 16
Incorrect syntax near '@offset'.
Msg 153, Level 15, State 2, Procedure getBatch, Line 16
Invalid usage of the option NEXT in the FETCH statement.

If you have a strong background in Oracle and can sort through the dynamic SQL syntax for T-SQL, you might try re-writing the function to use the EXEC SP_EXECUTESQL @var; command. That rewrite that attempts to use NDS (Native Dynamic SQL) would look like this:

CREATE FUNCTION studentdb.getBatch
(@rows   AS INT
,@offset AS INT) RETURNS @output TABLE
( marvel_id    INT 
, avenger_name VARCHAR(30)
, first_name   VARCHAR(20)
, last_name    VARCHAR(20)) AS
BEGIN
  DECLARE
    /* Declare a variable for a dynamic SQL statement. */
    @stmt          VARCHAR(400);
 
    /* Assign the SQL statement to a variable. */
    SET @stmt = N'SELECT marvel_id ' +
                N',      avenger_name ' +
                N',      first_name ' +
                N',      last_name ' +
                N'FROM   studentdb.marvel ' +
                N'OFFSET ' + (@offset - 1) + N' ' +
                N'ROWS FETCH NEXT ' + @rows + N' ROWS ONLY;';
  BEGIN
    /* Insert the results into the table variable. */
    INSERT @output
    EXEC sp_executesql @stmt; 
  END;
 
  /* Return the table variable from the function. */  
  RETURN;
END;

Throws the following exception because you can’t use dynamic dispatch inside a T-SQL function:

Msg 443, Level 16, State 14, Procedure getBatch, Line 23
Invalid use of a side-effecting operator 'INSERT EXEC' within a function.

On the other hand you can rewrite the statement with a BETWEEN operator and it works somewhat like an OFFSET and FETCH operation. That refactored function would be written as follows:

CREATE FUNCTION studentdb.getBatch
(@rowa  AS INT
,@rowb AS INT) RETURNS @output TABLE
( marvel_id    INT 
, avenger_name VARCHAR(30)
, first_name   VARCHAR(20)
, last_name    VARCHAR(20)) AS
BEGIN
  /* Insert the results into the table variable. */
  INSERT @output
  SELECT marvel_id
  ,      avenger_name
  ,      first_name
  ,      last_name
  FROM   studentdb.marvel
  WHERE  marvel_id BETWEEN @rowa AND @rowb;
 
  /* Return the table variable from the function. */  
  RETURN;
END;

It doesn’t raise an exception. You can call the table function like this:

SELECT * FROM getBatch(2,3);

It returns the two rows for Iron Man and Black Widow. As always, I hope this helps.

If you want to create the test case, here’s the script you need:

SELECT 'Conditionally drop studentdb.marvel table.' AS "Statement";
IF OBJECT_ID('studentdb.marvel','U') IS NOT NULL
  DROP TABLE studentdb.marvel;
 
SELECT 'Create studentdb.marvel table.' AS "Statement";
CREATE TABLE studentdb.marvel
( marvel_id     INT          NOT NULL IDENTITY(1,1) CONSTRAINT marvel_pk PRIMARY KEY
, avenger_name  VARCHAR(30)  NOT NULL
, first_name    VARCHAR(20)  NOT NULL
, last_name     VARCHAR(20)  NOT NULL);
 
/* Insert the rows. */
INSERT INTO studentdb.marvel (avenger_name, first_name, last_name) VALUES ('Hulk','Bruce','Banner');
INSERT INTO studentdb.marvel (avenger_name, first_name, last_name) VALUES ('Iron Man','Tony','Stark');
INSERT INTO studentdb.marvel (avenger_name, first_name, last_name) VALUES ('Black Widow','Natasha','Romanoff');
INSERT INTO studentdb.marvel (avenger_name, first_name, last_name) VALUES ('Thor','Thor','Odinsson');
INSERT INTO studentdb.marvel (avenger_name, first_name, last_name) VALUES ('Captain America','Steve','Rogers');
INSERT INTO studentdb.marvel (avenger_name, first_name, last_name) VALUES ('Hawkeye','Clint','Barton');
INSERT INTO studentdb.marvel (avenger_name, first_name, last_name) VALUES ('Winter Soldier','Bucky','Barnes');
INSERT INTO studentdb.marvel (avenger_name, first_name, last_name) VALUES ('Iron Patriot','James','Rhodey');
 
/* Query the contents of the MARVEL table. */
SELECT * FROM studentdb.marvel;

A question came up about how to verify dates from a string without throwing a casting error because of a non-conforming date. You can throw a number of exceptions, and I wrote a function to filter bad string formats like the DD-MON-RR or DD-MON-YYYY.

The first one is for a day between 1 and the last day of month, which is:

ORA-01847: day of month must be between 1 and last day of month

An incorrect string for a month, raises the following error:

ORA-01843: not a valid month

A date format mask longer than a DD-MON-RR or DD-MON-YYYY raises the following exception:

ORA-01830: date format picture ends before converting entire input string

The verify_date function checks for non-conforming DD-MON-RR and DD-MON-YYYY date masks, and substitutes a SYSDATE value for a bad date entry:

CREATE OR REPLACE
  FUNCTION verify_date
  ( pv_date_in  VARCHAR2) RETURN DATE IS
  /* Local return variable. */
  lv_date  DATE;
BEGIN
  /* Check for a DD-MON-RR or DD-MON-YYYY string. */
  IF REGEXP_LIKE(pv_date_in,'^[0-9]{2,2}-[ADFJMNOS][ACEOPU][BCGLNPRTVY]-([0-9]{2,2}|[0-9]{4,4})$') THEN
    /* Case statement checks for 28 or 29, 30, or 31 day month. */
    CASE
      /* Valid 31 day month date value. */
      WHEN SUBSTR(pv_date_in,4,3) IN ('JAN','MAR','MAY','JUL','AUG','OCT','DEC') AND
           TO_NUMBER(SUBSTR(pv_date_in,1,2)) BETWEEN 1 AND 31 THEN 
        lv_date := pv_date_in;
      /* Valid 30 day month date value. */
      WHEN SUBSTR(pv_date_in,4,3) IN ('APR','JUN','SEP','NOV') AND
           TO_NUMBER(SUBSTR(pv_date_in,1,2)) BETWEEN 1 AND 30 THEN 
        lv_date := pv_date_in;
      /* Valid 28 or 29 day month date value. */
      WHEN SUBSTR(pv_date_in,4,3) = 'FEB' THEN
        /* Verify 2-digit or 4-digit year. */
        IF (LENGTH(pv_date_in) = 9 AND MOD(TO_NUMBER(SUBSTR(pv_date_in,8,2)) + 2000,4) = 0 OR
            LENGTH(pv_date_in) = 11 AND MOD(TO_NUMBER(SUBSTR(pv_date_in,8,4)),4) = 0) AND
            TO_NUMBER(SUBSTR(pv_date_in,1,2)) BETWEEN 1 AND 29 THEN
          lv_date := pv_date_in;
        ELSE /* Not a leap year. */
          IF TO_NUMBER(SUBSTR(pv_date_in,1,2)) BETWEEN 1 AND 28 THEN
            lv_date := pv_date_in;
          ELSE
            lv_date := SYSDATE;
          END IF;
        END IF;
      ELSE
        /* Assign a default date. */
        lv_date := SYSDATE;
    END CASE;
  ELSE
    /* Assign a default date. */
    lv_date := SYSDATE;
  END IF;
  /* Return date. */
  RETURN lv_date;
END;
/

You can check valid dates with a DD-MON-RR format:

SELECT verify_date('28-FEB-10') AS "Non-Leap Year"
,      verify_date('29-FEB-12') AS "Leap Year"
,      verify_date('31-MAR-14') AS "31-Day Year"
,      verify_date('30-APR-14') AS "30-Day Year"
FROM   dual;

You can check valid dates with a DD-MON-YYYY format:

SELECT verify_date('28-FEB-2010') AS "Non-Leap Year"
,      verify_date('29-FEB-2012') AS "Leap Year"
,      verify_date('31-MAR-2014') AS "31-Day Year"
,      verify_date('30-APR-2014') AS "30-Day Year"
FROM   dual;

They both return:

Non-Leap    Leap YEAR 31-DAY YEAR 30-DAY YEAR
----------- --------- ----------- -----------
28-FEB-10   29-FEB-12 31-MAR-14   30-APR-14

You can check badly formatted dates with the following query:

SELECT verify_date('28-FEB-2010') AS "Non-Leap Year"
,      verify_date('29-FEB-2012') AS "Leap Year"
,      verify_date('31-MAR-2014') AS "31-Day Year"
,      verify_date('30-APR-2014') AS "30-Day Year"
FROM   dual;

You can screen for an alphanumeric string with the following expression:

SELECT 'Valid alphanumeric string literal' AS "Statement"
FROM   dual
WHERE  REGEXP_LIKE('Some Mythical String $200','([:alnum:]|[:punct:]|[:space:])*');

You can screen for a numeric literal as a string with the following expression:

SELECT 'Valid numeric literal' AS "Statement"
FROM   dual
WHERE  REGEXP_LIKE('123.00','([:digit:]|[:punct:])');

As always, I hope this helps those who need this type of solution.