MacLochlainns Weblog

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.

Student questions are always interesting! They get me to think and to write. The question this time is: “How do I write a Bash Shell script to process multiple MySQL script files?” This post builds the following model (courtesy of MySQL Workbench) by using a bash shell script and MySQL script files, but there’s a disclaimer on this post. It shows both insecure and secure approaches and you should avoid the insecure ones.

It seems a quick refresher on how to use arrays in bash shell may be helpful. While it’s essential in a Linux environment, it’s seems not everyone masters the bash shell.

Especially, since I checked my Learning the Bash Shell (2^nd Edition) and found a typo on how you handle arrays in the bash shell, and it’s a mistake that could hang newbies up (on page 161). Perhaps I should update my copy because I bought it in 1998. 😉 It was good then, and the new edition is probably better. The error is probably corrected in the current Learning the Bash Shell, but if not, the following examples show you how to use arrays in loops.

Naturally, these do presume some knowledge of working with bash shell, like the first line always is the same in any bash shell script. That you open an if-statement with an if and close it with a fi, and that you else-if is elif; and that a semicolon between a for-statement and the do statement is required when they’re on the same line because they’re two statements.

If you’re new to bash shell arrays, click on the link below to expand a brief tutorial. It takes you through three progressive examples of working with bash arrays.

1
2
3
4
5
6
7
8
9
10
11
12
13

#!/usr/bin/bash
 
# Print script name.
echo $0
#!/usr/bin/bash
 
# Define an array.
declare -a cmd=("one" "two" "three")
 
# Call the array elements.
for i in ${cmd[*]}; do
  echo ${i}
done

chmod 755 list1.sh

./list1.sh

one
two
three

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

#!/usr/bin/bash
 
# Define an array.
declare -a cmd
 
# Assign elements to an array.
cmd[0]="one"
cmd[1]="two"
cmd[2]="three"
 
# Call the array elements.
for i in ${cmd[*]}; do
  echo ${i}
done

1
2
3
4

#!/usr/bin/bash
 
# Print script name.
echo $0

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

#!/usr/bin/bash
<code>
# Define an array.
declare -a cmd
 
# Assign elements to an array.
cmd[0]="one.sh"
cmd[1]="two.sh"
cmd[2]="three.sh"
 
# Call the array elements.
for i in ${cmd[*]}; do
  `pwd`/${i}
done

/home/student/Code/bash/one.sh
/home/student/Code/bash/two.sh
/home/student/Code/bash/three.sh

./one.sh

./one.sh

Only one more trick needs to be qualified before our main MySQL examples. That trick is how you pass parameters to a bash shell script. For reference, this is the part that’s insecure because user command histories are available inside the Linux OS.

Here’s a hello_whom.sh script to demonstrates the concept of parameter passing:

1
2
3
4
5
6
7
8
9
10

#!/usr/bin/bash
 
# This says hello to the argument while managing no argument.
if [[ ${#} = 1 ]]; then
  echo 'The '${0}' program says: "Hello '${1}'!"'
elif [[ ${#} > 1 ]]; then
  echo 'The '${0}' program wants to know if you have more than one name?'
else
  echo 'The '${0}' program wants to know if you have a name?'
fi

If you need more on how parameters are passed and managed, you can check a prior blob post on Handling bash Parameters, or check the bash help pages. The following leverages bash arrays to run scripts and query the MySQL database from the command line.

You will need the three batch SQL files first, so here they are:

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

-- Use the sampledb database.
USE sampledb;
 
-- Disable foreign key checking.
SET foreign_key_checks = 0;
 
-- Drop an actor table. 
DROP TABLE IF EXISTS actor;
 
-- Create an actor table.
CREATE TABLE actor
( actor_id    INT UNSIGNED PRIMARY KEY AUTO_INCREMENT
, actor_name  VARCHAR(30)  NOT NULL );
 
-- Insert two rows.
INSERT INTO actor (actor_name) VALUES ('Chris Hemsworth');
INSERT INTO actor (actor_name) VALUES ('Chris Pine');
INSERT INTO actor (actor_name) VALUES ('Chris Pratt');

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

-- Use the sampledb database.
USE sampledb;
 
-- Disable foreign key checking.
SET foreign_key_checks = 0;
 
-- Drop a film table. 
DROP TABLE IF EXISTS film;
 
-- Create a film table.
CREATE TABLE film
( film_id    INT UNSIGNED PRIMARY KEY AUTO_INCREMENT
, film_name  VARCHAR(30)  NOT NULL );
 
-- Insert rows.
INSERT INTO film (film_name) VALUES ('Thor');
INSERT INTO film (film_name) VALUES ('Thor: The Dark World');
INSERT INTO film (film_name) VALUES ('Star Trek');
INSERT INTO film (film_name) VALUES ('Star Trek into Darkness');
INSERT INTO film (film_name) VALUES ('Guardians of the Galaxy');

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

-- Use the sampledb database.
USE sampledb;
 
-- Disable foreign key checking.
SET foreign_key_checks = 0;
 
-- Drop an movie table. 
DROP TABLE IF EXISTS movie;
 
-- Create an movie table.
CREATE TABLE movie
( movie_id   int unsigned PRIMARY KEY AUTO_INCREMENT
, actor_id   int unsigned NOT NULL
, film_id    int unsigned NOT NULL
, CONSTRAINT actor_fk FOREIGN KEY (actor_id)
  REFERENCES actor (actor_id)
, CONSTRAINT film_fk  FOREIGN KEY (film_id)
  REFERENCES film(film_id));
 
-- Use scalar subqueries to discover surrogate keys by using the faux natural key.
INSERT INTO movie
( actor_id
, film_id )
VALUES
((SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Hemsworth')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Thor'));
 
-- Use scalar subqueries to discover surrogate keys by using the faux natural key.
INSERT INTO movie
( actor_id
, film_id )
VALUES
((SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Hemsworth')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Thor: The Dark World'));
 
-- Use scalar subqueries to discover surrogate keys by using the faux natural key.
INSERT INTO movie
( actor_id
, film_id )
VALUES
((SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Pine')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Star Trek'));
 
-- Use scalar subqueries to discover surrogate keys by using the faux natural key.
INSERT INTO movie
( actor_id
, film_id )
VALUES
((SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Pine')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Star Trek into Darkness'));
 
-- Use scalar subqueries to discover surrogate keys by using the faux natural key.
INSERT INTO movie
( actor_id
, film_id )
VALUES
((SELECT   actor_id
  FROM     actor
  WHERE    actor_name = 'Chris Pratt')
,(SELECT   film_id
  FROM     film
  WHERE    film_name = 'Guardians of the Galaxy'));

The following list_mysql.sh shell script expects to receive the username, password, database and fully qualified path in that specific order. The script names are entered manually because this should be a unit test script. Naturally, you can extend the script to manage those parameters but as mentioned I see this type of solution as a developer machine only script to simplify unit testing. Anything beyond that is risky!

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

#!/usr/bin/bash
 
# Assign user and password
username="${1}"
password="${2}"
database="${3}"
directory="${4}"
 
# List the parameter values passed.
echo "Username:  " ${username}
echo "Password:  " ${password}
echo "Database:  " ${database}
echo "Directory: " ${directory}
echo ""
 
# 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 ${cmd[*]}; do
  mysql -s -u${username} -p${password} -D${database} < ${directory}/${i} > /dev/null 2>/dev/null
done
 
# Connect and pipe the query result minus errors and warnings to the while loop.
mysql -u${username} -p${password} -D${database} <<<'show tables' 2>/dev/null |
 
# Read through the piped result until it's empty but format the title.
while IFS='\n' read list; do
  if [[ ${list} = "Tables_in_sampledb" ]]; then
    echo $list
    echo "----------------------------------------"
  else
    echo $list
  fi
done
echo ""
 
# Connect and pipe the query result minus errors and warnings to the while loop.
mysql -u${username} -p${password} -D${database} <<<'SELECT CONCAT(a.actor_name," in ",f.film_name) AS "Actors in Films" 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' 2>/dev/null |
 
# Read through the piped result until it's empty but format the title.
while IFS='\n' read actor_name; do
  if [[ ${actor_name} = "Actors in Films" ]]; then
    echo $actor_name
    echo "----------------------------------------"
  else
    echo $actor_name
  fi
done

The IFS (Internal Field Separator) works with whitespace by default. The IFS on lines 33 and 47 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 this script with the following input parameters from the local directory where you deploy it. The a parameters are: (1) username, (2) password, (3) database, and (4) a fully qualified path to the SQL setup files.

./list_mysql.sh student student sampledb "/home/student/Code/bash/mysql"

With valid input values, the list_mysql.sh bash script generates the following output, which confirms inputs and verifies actions taken by the scripts with queries:

Username:   student
Password:   student
Database:   sampledb
Directory:  /home/student/Code/bash/mysql
 
Tables_in_sampledb
----------------------------------------
actor
film
movie
 
Actors in Films
----------------------------------------
Chris Hemsworth in Thor
Chris Hemsworth in Thor: The Dark World
Chris Pine in Star Trek
Chris Pine in Star Trek into Darkness
Chris Pine in Guardians of the Galaxy

If you forgot to provide the required inputs to the list_mysql.sh bash script, it alternatively returns the following output:

Username:  
Password:  
Database:  
Directory: 
 
./list_mysql.sh: line 25: /actor.sql: No such file or directory
./list_mysql.sh: line 25: /film.sql: No such file or directory
./list_mysql.sh: line 25: /movie.sql: No such file or directory

The secure way removes the password at a minimum! The refactored program will require you to manually enter the password for all elements of the array (three in this sample), and twice for the two queries. Here’s the refactored code:

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

#!/usr/bin/bash
 
# Assign user and password
username="${1}"
database="${2}"
directory="${3}"
 
# List the parameter values passed.
echo "Username:  " ${username}
echo "Database:  " ${database}
echo "Directory: " ${directory}
echo ""
 
# 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 ${cmd[*]}; do
  mysql -s -u${username} -p -D${database} < ${directory}/${i} > /dev/null 2>/dev/null
done
 
# Connect and pipe the query result minus errors and warnings to the while loop.
mysql -u${username} -p -D${database} <<<'show tables' 2>/dev/null |
 
# Read through the piped result until it's empty.
while IFS='\n' read list; do
  if [[ ${list} = "Tables_in_sampledb" ]]; then
    echo $list
    echo "----------------------------------------"
  else
    echo $list
  fi
done
echo ""
 
# Connect and pipe the query result minus errors and warnings to the while loop.
mysql -u${username} -p -D${database} <<<'SELECT CONCAT(a.actor_name," in ",f.film_name) AS "Actors in Films" 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' 2>/dev/null |
 
# Read through the piped result until it's empty.
while IFS='\n' read actor_name; do
  if [[ ${actor_name} = "Actors in Films" ]]; then
    echo $actor_name
    echo "----------------------------------------"
  else
    echo $actor_name
  fi
done

Please let me know if you think there should be any more scaffolding for newbies in this post. As always, I hope this helps those looking for this type of solution.

Fedora comes with vim and gedit installed but the gedit installation is bare bones. You can update gedit to include supplemental Plug-ins with the following yum command as the root user:

yum install -y gedit-plugins

It generates the following log file:

Loaded plugins: langpacks, refresh-packagekit
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                                  |  14 kB  00:00     
updates                                                     | 4.9 kB  00:00     
(1/2): pgdg93/20/x86_64/primary_db                          |  86 kB  00:00     
(2/2): updates/20/x86_64/primary_db                         |  11 MB  00:03     
(1/2): updates/20/x86_64/pkgtags                            | 1.5 MB  00:00     
(2/2): updates/20/x86_64/updateinfo                         | 2.0 MB  00:01     
Resolving Dependencies
--> Running transaction check
---> Package gedit-plugins.x86_64 0:3.10.1-1.fc20 will be installed
--> Processing Dependency: libgit2-glib for package: gedit-plugins-3.10.1-1.fc20.x86_64
--> Running transaction check
---> Package libgit2-glib.x86_64 0:0.0.6-2.fc20 will be installed
--> Processing Dependency: libgit2.so.0()(64bit) for package: libgit2-glib-0.0.6-2.fc20.x86_64
--> Running transaction check
---> Package libgit2.x86_64 0:0.19.0-2.fc20 will be installed
--> Processing Dependency: libxdiff.so.1()(64bit) for package: libgit2-0.19.0-2.fc20.x86_64
--> Processing Dependency: libhttp_parser.so.2()(64bit) for package: libgit2-0.19.0-2.fc20.x86_64
--> Running transaction check
---> Package http-parser.x86_64 0:2.0-5.20121128gitcd01361.fc20 will be installed
---> Package libxdiff.x86_64 0:1.0-3.fc20 will be installed
--> Finished Dependency Resolution
 
Dependencies Resolved
 
================================================================================
 Package          Arch      Version                            Repository  Size
================================================================================
Installing:
 gedit-plugins    x86_64    3.10.1-1.fc20                      updates    830 k
Installing for dependencies:
 http-parser      x86_64    2.0-5.20121128gitcd01361.fc20      fedora      23 k
 libgit2          x86_64    0.19.0-2.fc20                      fedora     281 k
 libgit2-glib     x86_64    0.0.6-2.fc20                       fedora      82 k
 libxdiff         x86_64    1.0-3.fc20                         fedora      33 k
 
Transaction Summary
================================================================================
Install  1 Package (+4 Dependent packages)
 
Total download size: 1.2 M
Installed size: 5.2 M
Downloading packages:
(1/5): http-parser-2.0-5.20121128gitcd01361.fc20.x86_64.rpm |  23 kB  00:00     
(2/5): libgit2-0.19.0-2.fc20.x86_64.rpm                     | 281 kB  00:00     
(3/5): libgit2-glib-0.0.6-2.fc20.x86_64.rpm                 |  82 kB  00:00     
(4/5): libxdiff-1.0-3.fc20.x86_64.rpm                       |  33 kB  00:00     
(5/5): gedit-plugins-3.10.1-1.fc20.x86_64.rpm               | 830 kB  00:01     
--------------------------------------------------------------------------------
Total                                              899 kB/s | 1.2 MB  00:01     
Running transaction check
Running transaction test
Transaction test succeeded
Running transaction (shutdown inhibited)
  Installing : libxdiff-1.0-3.fc20.x86_64                                   1/5 
  Installing : http-parser-2.0-5.20121128gitcd01361.fc20.x86_64             2/5 
  Installing : libgit2-0.19.0-2.fc20.x86_64                                 3/5 
  Installing : libgit2-glib-0.0.6-2.fc20.x86_64                             4/5 
  Installing : gedit-plugins-3.10.1-1.fc20.x86_64                           5/5 
  Verifying  : libgit2-0.19.0-2.fc20.x86_64                                 1/5 
  Verifying  : libgit2-glib-0.0.6-2.fc20.x86_64                             2/5 
  Verifying  : gedit-plugins-3.10.1-1.fc20.x86_64                           3/5 
  Verifying  : http-parser-2.0-5.20121128gitcd01361.fc20.x86_64             4/5 
  Verifying  : libxdiff-1.0-3.fc20.x86_64                                   5/5 
 
Installed:
  gedit-plugins.x86_64 0:3.10.1-1.fc20                                          
 
Dependency Installed:
  http-parser.x86_64 0:2.0-5.20121128gitcd01361.fc20                            
  libgit2.x86_64 0:0.19.0-2.fc20                                                
  libgit2-glib.x86_64 0:0.0.6-2.fc20                                            
  libxdiff.x86_64 0:1.0-3.fc20                                                  
 
Complete!

When you launch the gedit utility, you click on the

Gedit Plug-in Installation

1. After you install the Gedit Plug-ins, you can configure the plug-ins by launching Gedit and then click on the gedit menu option. Then, click on the Preferences menu option to enable the new plugins, like the Embedded Terminal plug-in.

2. You have four tab options when working with the Preferences menu. The first tab is the View tab, as shown to the left.

3. The second tab is the Editor tab, as shown to the left.

4. The third tab is the Font & Colors tab, as shown to the left.

5. The fourth tab is the Plugins tab, as shown to the left. Scroll down the list and check the Embedded Terminal and Python Console plug-ins’ checkbox. The Embedded Terminal lets you edit a file and have command line access to a Terminal session from the gedit menu; and the Python Console session from the gedit menu.

6. Click on the View menu, and then choose the Bottom Panel menu option.

7. After enabling the Bottom Panel in the Gedit menu, you can edit a file and click on the Terminal by simply clicking on the subpanel. You can see the split image on the left. There’s also a set of bottom tabs that lets you switch from a Linux Terminal session to the Python console.

As always, I hope this helps those working with gedit on the Fedora operating system.

While I previously blogged about installing Netbeans 8, some of my students would prefer to use the Eclipse IDE. This post shows how to install and configure Eclipse IDE, include the mysql-connector-java.jar, and write Java to access the MySQL.

You can download Eclipse IDE and then open it in Fedora’s Archive Manager. You can use the Archive Manager to Extract the Eclipse IDE to a directory of your choice. I opted to extract it into my student user’s home directory, which is /home/student.

After extracting the Eclipse IDE, you can check the contents of the eclipse directory with the following command:

ls -al eclipse

You should see the following:

drwxrwxr-x.  8 student student   4096 May  8 22:16 .
drwx------. 33 student student   4096 May  8 21:57 ..
-rw-rw-r--.  1 student student 119194 Mar 20 07:10 artifacts.xml
drwxrwxr-x. 11 student student   4096 May  8 22:16 configuration
drwxrwxr-x.  2 student student   4096 Mar 20 07:10 dropins
-rwxr-xr-x.  1 student student  78782 Mar 20 07:08 eclipse
-rw-rw-r--.  1 student student    315 Mar 20 07:10 eclipse.ini
-rw-rw-r--.  1 student student     60 Mar 17 15:11 .eclipseproduct
drwxrwxr-x. 41 student student   4096 Mar 20 07:10 features
-rwxr-xr-x.  1 student student 140566 Mar 20 07:08 icon.xpm
drwxrwxr-x.  4 student student   4096 Mar 20 07:09 p2
drwxrwxr-x. 12 student student  40960 Mar 20 07:10 plugins
drwxrwxr-x.  2 student student   4096 Mar 20 07:10 readme

You can launch the Eclipse IDE with the following command-line from the eclipse directory:

./eclipse &

While you can run this from the /home/student/eclipse directory, it’s best to create an alias for the Eclipse IDE in the student user’s .bashrc file:

# Set alias for Eclipse IDE tool.
alias eclipse="/home/student/eclipse/eclipse"

The next time you start the student user account, you can launch the Eclipse IDE by entering eclipse in the search box opened by clicking on the Activities menu.

The following steps take you through installing Eclipse on Fedora Linux, which is more or less the same as any Linux distribution. It’s very similar on Windows platforms too.

Eclipse Installation

1. Navigate to eclipse.org/downloads web page to download the current version of the Eclipse software. Click the Linux 32 Bit or Linux 64 Bit link, as required for your operating system.

2. Click the Green Arrow to download the Eclipse software.

3. The next dialog gives you an option to open or save the software. Click the Open with radio button to open the archive file.

4. This the Linux Archive Manager. Click the Extract button from the menu tab to open the archive file.

5. This extract button on file chooser dialog to install Eclipse into the /home/student/eclipse directory. Click the Extract button to let the Archive Manager create a copy of those files.

6. The Archive Manager presents a completion dialog. Click the Close button to close the Archive Manager.

After installing the Eclipse software, you can configure Eclipse. There are sixteen steps to setup the Eclipse product. You can launch the product with the

Eclipse Setup

You need to launch the Eclipse application to perform the following steps. The syntax is the following when you did create the alias mentioned earlier in the blog post:

eclipse &

The following steps cover setting up your workspace, project, and adding the MySQL JDBC Java archive.

1. The branding dialog may display for 30 or more seconds before the Eclipse software application launches.

2. The Workspace Launcher opens first on a new installation. You need to designate a starting folder. I’m using /home/student/workspace as my Workspace. Click the OK button when you enter a confirmed workspace.

3. After setting the Workspace Launcher, you open to the Eclipse Welcome page. Click second of the two icons on the left to open a working Eclipse environment. Alternatively, you can connect to Tutorials on the same page.

4. From the developer view, click on the File menu option, the New option on the list, and the Java Project option on the floating menu. Eclipse will now create a new Java project.

5. The New Java Project dialog lets you enter a project name and it also gives you the ability to set some basic configuration details. As a rule, you simply enter the Project Name and accept the defaults before clicking the Finish button.

6. After creating the new Java project, Eclipse returns you to the Welcome page. Click second of the two icons on the left to open a working Eclipse environment.

7. Now you should see the working environment. Sometimes it takes the full screen but initially it doesn’t. Navigate to the lower right hand side, and expand the window to full size.

8. Now you should see the full screen view of the Eclipse working environment.

9. Now you create a new Java class by navigating to the File menu options, then the New menu option, and finally choosing the Class floating menu.

10. The New Java Class dialog requires you to provide some information about the Java object you’re creating. The most important thing is the Java class name.

11. The only difference in this copy of the New Java Class dialog is that I’ve entered HelloWorld as the Java Class’s name. Click the Finish button when you’re done.

12. Eclipse should show you the following HelloWorld.java file. It’s missing a main() method. Add a static main() method to the HelloWorld.java class source file.

13. This form shows the changes to the HelloWorld.java file. Specifically, it adds the It’s missing a main() method. Add a static main() method to the HelloWorld.java class source file.

14. You can click the green arrow from the tool panel or you can click the Run menu option and Run submenu choice to test your program.

    1 2 3 4

    // Class definition. public class HelloWorld { public static void main(String args[]) { System.out.println("Hello World."); }}

    // Class definition. public class HelloWorld { public static void main(String args[]) { System.out.println("Hello World."); }}

15. The Save and Launch dialog tells you that you’re ready to test creating a copy of the Java class file. Click the OK button to continue.

16. The results from your program are written to the Console portion of the Eclipse IDE. This concludes the setup of a workspace, project, and deployment of actual Java classes.

    Hello World.

    Hello World.

Add MySQL JDBC Library

The following instructions add the MySQL Library and demonstrate how to write Java programs that connect to the MySQL database. They also use the mysql project.

1. Navigate to the Project menu and choose the Properties menu option.

2. The Properties menu option opens the Properties for the mysql project on the Order and Export tab. Click the Libraries tab to add an external library.

3. In the Libraries tab click the Add Library… button on the right to add an external library.

4. In the JAR Selection dialog, click on Computer in the Places list, then click on usr, click on share, and click on java. The Name list should now include mysql-connector-java.jar file, and you should click on it before clicking on the OK button.

5. You create new Java class file by clicking on the File menu. Then, you choose the New menu option and the Class menu option from the floating menu.

6. Enter MysqlConnector as the name of the new Java class file and click the Finish button to continue.

7. Eclipse generates the shell of the MysqlConnector class as shown in the illustration to the left.

8. You should replace the MysqlConnector class shell with the code below. Then, click the green arrow or the Run menu and Run menu option to compile and run the new MysqlConnector Java class file.

   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

   import java.sql.Connection; import java.sql.DriverManager; import java.sql.SQLException;   public class MysqlConnector extends Object { public static void main(String[] args) { try { /* The newInstance() call is a work around for some broken Java implementations. */ Class.forName("com.mysql.jdbc.Driver").newInstance();   /* Verify the Java class path. */ System.out.println("===================="); System.out.println("CLASSPATH [" + System.getProperty("java.class.path") + "]"); System.out.println("====================");   } catch (Exception e) {} finally { /* Verify the Java class path. */ System.out.println("===================="); System.out.println("CLASSPATH [" + System.getProperty("java.class.path") + "]"); System.out.println("===================="); } } }

   import java.sql.Connection; import java.sql.DriverManager; import java.sql.SQLException; public class MysqlConnector extends Object { public static void main(String[] args) { try { /* The newInstance() call is a work around for some broken Java implementations. */ Class.forName("com.mysql.jdbc.Driver").newInstance(); /* Verify the Java class path. */ System.out.println("===================="); System.out.println("CLASSPATH [" + System.getProperty("java.class.path") + "]"); System.out.println("===================="); } catch (Exception e) {} finally { /* Verify the Java class path. */ System.out.println("===================="); System.out.println("CLASSPATH [" + System.getProperty("java.class.path") + "]"); System.out.println("===================="); } } }

9. The Save and Launch dialog informs you are saving a MysqlConnector.java file to your mysql project. Click the OK button to continue.

10. The next screen shows that the program successfully connected to the MySQL database by printing the following information to the Console output tab.

    ==================== CLASSPATH [/home/student/Code/workspace/MySQL/bin:/usr/share/java/mysql-connector-java.jar] ==================== ==================== CLASSPATH [/home/student/Code/workspace/MySQL/bin:/usr/share/java/mysql-connector-java.jar] ====================

    ==================== CLASSPATH [/home/student/Code/workspace/MySQL/bin:/usr/share/java/mysql-connector-java.jar] ==================== ==================== CLASSPATH [/home/student/Code/workspace/MySQL/bin:/usr/share/java/mysql-connector-java.jar] ====================

11. Instead of repeating steps #5 through #10, the image displays the testing of the MysqlResults class file. The code follows below:

    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

    /* Import the java.sql.* package. */ import java.sql.*;   /* You can't include the following on Linux without raising an exception. */ // import com.mysql.jdbc.Driver;   public class MySQLResult { public MySQLResult() { /* Declare variables that require explicit assignments because they're addressed in the finally block. */ Connection conn = null; Statement stmt = null; ResultSet rset = null;   /* Declare other variables. */ String url; String username = "student"; String password = "student"; String database = "studentdb"; String hostname = "localhost"; String port = "3306"; String sql;   /* Attempt a connection. */ try { // Set URL. url = "jdbc:mysql://" + hostname + ":" + port + "/" + database;   // Create instance of MySQL. Class.forName ("com.mysql.jdbc.Driver").newInstance(); conn = DriverManager.getConnection (url, username, password);   // Query the version of the database, relies on *_ri2.sql scripts. sql = "SELECT i.item_title, ra.rating FROM item i INNER JOIN rating_agency ra ON i.item_rating_id = ra.rating_agency_id"; stmt = conn.createStatement(); rset = stmt.executeQuery(sql);   System.out.println ("Database connection established");   // Read row returns for one column. while (rset.next()) { System.out.println(rset.getString(1) + ", " + rset.getString(2)); }   } catch (SQLException e) { System.err.println ("Cannot connect to database server (SQLException):"); System.out.println(e.getMessage()); } catch (ClassNotFoundException e) { System.err.println ("Cannot connect to database server (ClassNotFoundException)"); System.out.println(e.getMessage()); } catch (InstantiationException e) { System.err.println ("Cannot connect to database server (InstantiationException)"); System.out.println(e.getMessage()); } catch (IllegalAccessException e) { System.err.println ("Cannot connect to database server (IllegalAccesException)"); System.out.println(e.getMessage()); } finally { if (conn != null) { try { rset.close(); stmt.close(); conn.close(); System.out.println ("Database connection terminated"); } catch (Exception e) { /* ignore close errors */ } } } } /* Unit test. */ public static void main(String args[]) { new MySQLResult(); } }

    /* Import the java.sql.* package. */ import java.sql.*; /* You can't include the following on Linux without raising an exception. */ // import com.mysql.jdbc.Driver; public class MySQLResult { public MySQLResult() { /* Declare variables that require explicit assignments because they're addressed in the finally block. */ Connection conn = null; Statement stmt = null; ResultSet rset = null; /* Declare other variables. */ String url; String username = "student"; String password = "student"; String database = "studentdb"; String hostname = "localhost"; String port = "3306"; String sql; /* Attempt a connection. */ try { // Set URL. url = "jdbc:mysql://" + hostname + ":" + port + "/" + database; // Create instance of MySQL. Class.forName ("com.mysql.jdbc.Driver").newInstance(); conn = DriverManager.getConnection (url, username, password); // Query the version of the database, relies on *_ri2.sql scripts. sql = "SELECT i.item_title, ra.rating FROM item i INNER JOIN rating_agency ra ON i.item_rating_id = ra.rating_agency_id"; stmt = conn.createStatement(); rset = stmt.executeQuery(sql); System.out.println ("Database connection established"); // Read row returns for one column. while (rset.next()) { System.out.println(rset.getString(1) + ", " + rset.getString(2)); } } catch (SQLException e) { System.err.println ("Cannot connect to database server (SQLException):"); System.out.println(e.getMessage()); } catch (ClassNotFoundException e) { System.err.println ("Cannot connect to database server (ClassNotFoundException)"); System.out.println(e.getMessage()); } catch (InstantiationException e) { System.err.println ("Cannot connect to database server (InstantiationException)"); System.out.println(e.getMessage()); } catch (IllegalAccessException e) { System.err.println ("Cannot connect to database server (IllegalAccesException)"); System.out.println(e.getMessage()); } finally { if (conn != null) { try { rset.close(); stmt.close(); conn.close(); System.out.println ("Database connection terminated"); } catch (Exception e) { /* ignore close errors */ } } } } /* Unit test. */ public static void main(String args[]) { new MySQLResult(); } }

    After you click the green arrow or the Run menu and Run menu option to compile and run the program, you should see the following output. That is if you’re using my create_mysql_store_ri2.sql and seed_mysql_store_ri2.sql files.

    Database connection established I Remember Mama, NR Tora! Tora! Tora!, G A Man for All Seasons, G Around the World in 80 Days, G Camelot, G Christmas Carol, G I Remember Mama, G The Hunt for Red October, PG Star Wars I, PG Star Wars II, PG Star Wars II, PG The Chronicles of Narnia, PG Beau Geste, PG Hook, PG Harry Potter and the Sorcerer's Stone, PG Scrooge, PG Harry Potter and the Sorcer's Stone, PG Harry Potter and the Sorcer's Stone, PG Harry Potter and the Chamber of Secrets, PG Harry Potter and the Chamber of Secrets, PG Harry Potter and the Prisoner of Azkaban, PG Harry Potter and the Prisoner of Azkaban, PG Harry Potter and the Half Blood Prince, PG Star Wars III, PG-13 Casino Royale, PG-13 Casino Royale, PG-13 Die Another Day, PG-13 Die Another Day, PG-13 Die Another Day, PG-13 Golden Eye, PG-13 Golden Eye, PG-13 Tomorrow Never Dies, PG-13 Tomorrow Never Dies, PG-13 The World Is Not Enough, PG-13 Clear and Present Danger, PG-13 Clear and Present Danger, PG-13 Harry Potter and the Goblet of Fire, PG-13 Harry Potter and the Goblet of Fire, PG-13 Harry Potter and the Goblet of Fire, PG-13 Harry Potter and the Order of the Phoenix, PG-13 Harry Potter and the Deathly Hallows, Part 1, PG-13 Harry Potter and the Deathly Hallows, Part 2, PG-13 Brave Heart, R The Chronicles of Narnia, E MarioKart, E Need for Speed, E Cars, E RoboCop, M Pirates of the Caribbean, T Splinter Cell, T The DaVinci Code, T Database connection terminated

    Database connection established I Remember Mama, NR Tora! Tora! Tora!, G A Man for All Seasons, G Around the World in 80 Days, G Camelot, G Christmas Carol, G I Remember Mama, G The Hunt for Red October, PG Star Wars I, PG Star Wars II, PG Star Wars II, PG The Chronicles of Narnia, PG Beau Geste, PG Hook, PG Harry Potter and the Sorcerer's Stone, PG Scrooge, PG Harry Potter and the Sorcer's Stone, PG Harry Potter and the Sorcer's Stone, PG Harry Potter and the Chamber of Secrets, PG Harry Potter and the Chamber of Secrets, PG Harry Potter and the Prisoner of Azkaban, PG Harry Potter and the Prisoner of Azkaban, PG Harry Potter and the Half Blood Prince, PG Star Wars III, PG-13 Casino Royale, PG-13 Casino Royale, PG-13 Die Another Day, PG-13 Die Another Day, PG-13 Die Another Day, PG-13 Golden Eye, PG-13 Golden Eye, PG-13 Tomorrow Never Dies, PG-13 Tomorrow Never Dies, PG-13 The World Is Not Enough, PG-13 Clear and Present Danger, PG-13 Clear and Present Danger, PG-13 Harry Potter and the Goblet of Fire, PG-13 Harry Potter and the Goblet of Fire, PG-13 Harry Potter and the Goblet of Fire, PG-13 Harry Potter and the Order of the Phoenix, PG-13 Harry Potter and the Deathly Hallows, Part 1, PG-13 Harry Potter and the Deathly Hallows, Part 2, PG-13 Brave Heart, R The Chronicles of Narnia, E MarioKart, E Need for Speed, E Cars, E RoboCop, M Pirates of the Caribbean, T Splinter Cell, T The DaVinci Code, T Database connection terminated

As always, I hope the note helps those trying to work with the Eclipse product.

I wrote a shared C library example to demonstrate external procedures in the Oracle Database 11g PL/SQL Programming book. I also reused the same example to demonstrate Oracle’s external procedures in the Oracle Database 12c PL/SQL Advanced Programming Techniques book last year. The example uses a C Shared Library but a PL/SQL wrapper and PL/SQL test case.

One of my students asked me to simplify the unit test case example by writing the complete unit test in the C Progamming Language. The student request seemed like a good idea, and while poking around on the web it appears there’s a strong case for a set of simple shared C library examples. This blog post isn’t meant to replace the C Programming web site and C Programming Tutorial web site, which I recommend as a great reference point.

Like most things, the best place to start is with basics of C programming because some readers may be very new to C programming. I’ll start with basic standalone programs and how to use the gcc compiler before showing you how to use shared C libraries.

The most basic program is a hello.c program that serves as a “Hello World!” program:

1
2
3
4
5

#include <stdio.h>
 
int main() {
  printf("Hello World!\n");
  return(0); }

Assuming you put the C source files in a src subdirectory and the executable files in a bin subdirectory. You compile the program with the gcc program from the parent directory of the src and bin subdirectories, as follows:

gcc -o bin/hello src/hello.c

Then, you execute the hello executable program from the parent directory as follows:

bin/hello

It prints:

Hello World!

You can modify the basic Hello World! program to accept a single input word, like this hello_whom.c program:

1
2
3
4
5
6
7
8
9
10
11
12

#include <stdio.h>
 
/* The executable main method. */
int main() {
  // Declare a character array to hold an input value.
  char whom[30];
 
  /* Print a question and read a string input. */
  printf("Who are you? ");
  scanf("%s", whom);
  printf("Hello %s!\n", whom);
  return(0); }

You can compile the hello_whom.c program as follows:

gcc -o bin/hello_whom src/hello_whom.c

Then, you execute the hello_whom executable program from the parent directory as follows:

bin/hello_whom
Who are you? Stuart

It prints:

Hello Stuart!

Alternatively, you can modify the hello_whom.c program to accept a stream of text, like the following hello_string.c program:

1
2
3
4
5
6
7
8
9
10
11
12

#include <stdio.h>
 
/* The executable main method. */
int main() {
  // Declare a character array to hold an input name.
  char phrase[4000];
 
  /* Print a question and read a string input. */
  printf("Hello? ");
  scanf("%[^\n]%*c", phrase);
  printf("Hello %s!\n", phrase);
  return(0); }

The [] is the scan set character. The [^\n] on line 10 defines the input as not a newline with a white space, and the %*c reads the newline character from the input buffer. After you compile the program you can call it like this:

bin/hello_string
Hello? there, it reads like a C++ stream

It would print:

Hello there, it reads like a C++ stream!

These example, like the previous examples, assume the source files are in a src subdirectory and the executable files are in the bin subdirectory. All compilation commands are run from the parent directory of the src and bin subdirectories.

The first example puts everything into a single writingstr.c file. It defines a writestr() function prototype before the main() function and the writestr() function after the main() function.

The code follows below:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22

#include <stdio.h>
 
/* Declare two integer variables. */
char path[255], message[4000];
 
/* Define a prototype for the writestr() function. */
void writestr(char *path, char *message);
 
/* The executable main method. */
int main() {
  printf("Enter file name and message: ");
  scanf("%s %[^\n]%*c", &path, &message);
  printf("File name:    %s\n", path);
  printf("File content: %s\n", message);
  writestr(path, message);
  return(0); }
 
void writestr(char *path, char *message) {
  FILE *file_name;
  file_name = fopen(path,"w");
  fprintf(file_name,"%s\n",message);
  fclose(file_name); }

You can compile the writingstr.c function with the following syntax:

gcc -o bin/writingstr src/writingstr.c

You can run the writingstr executable file with the following syntax:

bin/writingstr
Enter file name and message: /home/student/Code/c/test.txt A string for a file.
File name:    /home/student/Code/c/test.txt
File content: A string for a file.

You’ll find a test.txt file written to the /home/student/Code/C directory. The file contains only the single sentence fragment entered above.

Now, let’s create a writestr.h header file, a writestr.c shared object file, and a main.c testing file. You should note a pattern between the self-contained code and the approach for using shared libraries. The prototype of the writestr() function becomes the definition of the writestr.h file, and the implementation of the writestr() function becomes the writestr.so shared library.

The main.c file contains the only the main() function from the writingstr.c file. The main() function uses the standard scanf() function to read a fully qualified file name (also known as a path) as a string and then a text stream for the content of the file.

You define the writestr.h header file as:

1
2
3
4
5
6

#ifndef writestr_h__
#define writestr_h__
 
extern void writestr(char *path, char *message);
 
#endif

You define the writestr.c shared library, which differs from the example in the book. The difference is the #include statement of the writestr.h header file. The source code follows:

1
2
3
4
5
6
7
8

#include <stdio.h>
#include "writestr.h"
 
void writestr(char *path, char *message) {
  FILE *file_name;
  file_name = fopen(path,"w");
  fprintf(file_name,"%s\n",message);
  fclose(file_name); }

You define the main.c testing program as:

1
2
3
4
5
6
7
8
9
10
11
12

#include <stdio.h>
#include "writestr.h"
 
/* Declare two integer variables. */
char path[255], message[4000];
 
/* The executable main method. */
int main() {
  printf("Enter file name and message: ");
  scanf("%s %[^\n]%*c", &path, &message);
  writestr(path, message);
  return(0); }

Before you begin the process to compile these, you should create an environment file that sets the $LD_LIBRARY_PATH environment variable or add it to your .bashrc file. You should point the $LD_LIBRARY_PATH variable to the directory where you’ve put your shared libraries.

# Set the LD_LIBRARY_PATH environment variable.
export LD_LIBRARY_PATH=/home/student/Code/c/trylib/libfile

With programs defined, you need to first compile the writestr.c shared library first. You use the following syntax from the parent directory of the src and bin subdirectories.

gcc -shared -fPIC -o bin/writestr.so src/writestr.c

If you haven’t set the $LD_LIBRARY_PATH, you may raise an exception. There’s also an alternative to setting the $LD_LIBRARY_PATH before you call the gcc executable. You can use the -L option set the $LD_LIBRARY_PATH for a given all to the gcc executable, like:

gcc -L /home/student/Code/c/trylib/libfile -shared -fPIC -o bin/writestr.so src/writestr.c

Then, you compile the main.c program. You must put the writestr.so shared library before you designate the main target object and main.c source files, like this:

gcc bin/writestr.so -o bin/main src/main.c

Now, you can perform a C-only unit test case by calling the main executable. However, you must have set the $LD_LIBRARY_PATH environment variable at runtime too. You see the following reply to the “Enter file name and message” question when you run the main program unit:

bin/main
Enter file name and message: /home/student/Code/c/trylib/libfile/test.txt A long native string is the second input to this C program.

You can now see that the a new test.txt file has been written to the target directory, and that it contains the following string:

A long native string is the second input to this C program.

As always, I hope this helps those you want to write shared libraries in the C programming language.