MacLochlainns Weblog

Michael McLaughlin's Technical Blog

Site Admin

Archive for October, 2016

Basic Python Object

with one comment

One of my students wanted a quick example of a Python object with getters and setters. So, I wrote a little example that I’ll share.

You define this file in a physical directory that is in your $PYTHONPATH, like this:

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
# Define Coordinate class.
class Coordinate:
  # The method that initializes the Coordinate class.
  def __init__ (self, x, y):
    self.x = x
    self.y = y
 
  # Gets the x value from the instance.
  def getX (self):
    return self.x
 
  # Gets the y value from the instance.
  def getY (self):
    return self.y
 
  # Sets the x value in the instance.
  def setX (self, x):
    self.x = x
 
  # Sets the y value in the instance.    
  def setY (self, y):
    self.y = y
 
  # Prints the coordinate pair.
  def printCoordinate(self):
    print (self.x, self.y)

Assuming the file name is Coordinate.py, you can put it into the Python Idle environment with the following command:

from Coordinate import Coordinate

You initialize the class, like this:

g = Coordinate(49,49)

Then, you can access the variables of the class or it’s methods as shown below:

# Print the retrieved value of x from the g instance of the Coordinate class.
print g.getX()
 
# Print the formatted and retrieved value of x from the g instance of the Coordinate class.
print "[" + str(g.getX()) + "]"
 
# Set the value of x inside the g instance of the Coordinate class.
print g.setX(39)
 
# Print the Coordinates as a set.
g.printCoordinate()

You would see the following results:

49
[49]
(39,49)

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

Written by maclochlainn

October 30th, 2016 at 6:23 pm

Create a Python Module

without comments

Sometime formal programming documentation is less than clear. At least, it’s less than clear until you’ve written your first solution. The Modules section of the Python language is one of those that takes a few moments to digest.

Chapters 22 and 23 in Learning Python gives some additional details but not a clear step-by-step approach to implementing Python modules. This post is designed to present the steps to write, import, and call a Python module. I figured that it would be helpful to write one for my students, and posting it in the blog seemed like the best idea.

I wrote the module to parse an Oracle version string into what we’d commonly expect to see, like the release number, an “R”, a release version, and then the full version number. The module name is more or less equivalent to a package name, and the file name is effectively the module name. The file name is strVersionOracle.py, which makes the strVersionOracle the module name.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
# Parse and format Oracle version.
def formatVersion(s):
 
  # Split string into collection.
  list = s.split(".")
 
  # Iterate through the result set.
  for i, l in enumerate(list):
    if i == 0 and list[i] == "11":
      label = str(l) + "g"
    elif i == 0 and list[i] == "12":
      label = label + str(l) + "c"
    elif i == 1:
      label = label + "R" + list[i] + " (" + s + ")"
 
  # Return the formatted string.
  return label

You can put this in any directory as long as you add it to the Python path. There are two Python paths to maintain. One is in the file system and the other is in Python’s interactive IDLE environment. You can check the contents of the IDLE path with the following interactive commands:

import sys
print sys.path

It prints the following:

['', '/usr/lib64/python27.zip', '/usr/lib64/python2.7', '/usr/lib64/python2.7/plat-linux2', '/usr/lib64/python2.7/lib-tk', '/usr/lib64/python2.7/lib-old', '/usr/lib64/python2.7/lib-dynload', '/usr/lib64/python2.7/site-packages', '/usr/lib64/python2.7/site-packages/gtk-2.0', '/usr/lib/python2.7/site-packages']

You can append to the IDLE path using the following command:

sys.path.append("/home/student/Code/python")

After putting the module in the runtime path, you can test the code in the IDLE environment:

1
2
3
import cx_Oracle
db = cx_Oracle.connect("student/student@xe")
print strVersionOracle.formatVersion(db.version)

Line 3 prints the result by calling the formatVersion function inside the strVersionOracle module. It prints the following:

11gR2 (11.2.0.2.0)

You can test the program outside of the runtime environment with the following oracleConnection.py file. It runs

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
# Import the Oracle library.
import cx_Oracle
import strVersionOracle
 
try:
  # Create a connection.
  db = cx_Oracle.connect("student/student@xe")
 
  # Print a message.
  print "Connected to the Oracle " + strVersionOracle.formatVersion(db.version) + " database."
 
except cx_Oracle.DatabaseError, e:
  error, = e.args
  print >> sys.stderr, "Oracle-Error-Code:", error.code
  print >> sys.stderr, "Oracle-Error-Message:", error.message
 
finally:
  # Close connection. 
  db.close()

You can call the formatVersion() function rather than a combination of module and function names when you write a more qualified import statement on line 3, like:

3
from strVersionOracle import formatVersion

Then, you can call the formatVersion() function like this on line 10:

10
  print "Connected to the Oracle " + formatVersion(db.version) + " database."

It works because you told it to import a function from a Python module. The first example imports a module that may contain one to many functions, and that style requires you to qualify the location of functions inside imported modules.

The oracleConnection.py program works when you call it from the Bash shell provided you do so from the same directory where the oracleConnection.py and strVersionOracle.py files (or Python modules) are located. If you call the oracleConnection.py file from a different directory, the reference to the library raises the following error:

Traceback (most recent call last):
  File "oracleConnection.py", line 3, in <module>
    import strVersionOracle
ImportError: No module named strVersionOracle

You can fix this error by adding the directory where the strVersionOracle.py file exists, like

export set PYTHONPATH=/home/student/Code/python

Then, you can call successfully the oracleConnection.py file from any directory:

python oracleConnection.py

The program will connect to the Oracle database as the student user, and print the following message to the console:

Connected to the Oracle 11gR2 (11.2.0.2.0) database.

I hope this helps those trying to create and use Python modules.

Written by maclochlainn

October 19th, 2016 at 11:50 pm

Python for loops

without comments

It’s always interesting to explain a new programming language to students. Python does presents some challenges to that learning process. I think for-loops can be a bit of a challenge until you understand them. Many students are most familiar with the traditional for loop like Java:

for (i = 0; i < 5; i++) { ... }

Python supports three types of for-loops – a range for loop, a for-each expression, and a for-loop with enumeration. Below are examples of each of these loops.

  1. A range for-loop goes from a low numerical value to a high numerical value, like:
  2. for i in range(0,3):
      print i

    It prints the following range values:

    0
    1
    2
  1. A for-each loop goes from the first to the last item while ignoring indexes, like:
  2. list = ['a','b','c']
    for i in list:
      print i

    It prints the following elements of the list:

    a
    b
    c
  1. A for-loop with enumeration goes from the first to the last item while ignoring indexes, like:
  2. list = ['a','b','c']
      for i, e in enumerate(list):
        print "[" + str(i) + "][" + list[i] + "]"

    The i represents the index values and the e represents the elements of a list. The str() function casts the numeric value to a string.

    It prints the following:

    [0][a]
    [1][b]
    [2][c]

This should help my students and I hope it helps you if you’re trying to sort out how to use for loops in Python.

Written by maclochlainn

October 19th, 2016 at 9:02 pm

Install cx_Oracle for Python

without comments

This shows you how to install the cx_Oracle library for Python 2.7 on Fedora Linux. If Fedora has it on the server you can download it with the following yum command:

yum install -y cx_Oracle-5.2.1-11g-py27-1.x86_64.rpm

Currently, you’ll get the following failure because it’s not available in the Fedora repository:

Loaded plugins: langpacks, refresh-packagekit
mysql-connectors-community                                      | 2.5 kB  00:00:00     
mysql-tools-community                                           | 2.5 kB  00:00:00     
mysql56-community                                               | 2.5 kB  00:00:00     
pgdg93                                                          | 3.6 kB  00:00:00     
updates/20/x86_64/metalink                                      | 2.3 kB  00:00:00     
No package cx_Oracle-5.2.1-11g-py27-1.x86_64.rpm available.
Error: Nothing to do

You can download the cx_Oracle library from the Python web site. The cx_Oracle documentation qualifies module interfaces, objects, and connections. Assuming your Linux user’s name is student, you download the cx_Oracle library into the /home/student/Downloads directory. Then, you use the su or sudo command to become the root user.

As the root user, run the following yum command:

yum install -y ~student/Downloads/cx_Oracle-5.2.1-11g-py27-1.x86_64.rpm

You should see the following output:

Loaded plugins: langpacks, refresh-packagekit
Examining /home/student/Downloads/cx_Oracle-5.2.1-11g-py27-1.x86_64.rpm: cx_Oracle-5.2.1-1.x86_64
Marking /home/student/Downloads/cx_Oracle-5.2.1-11g-py27-1.x86_64.rpm to be installed
Resolving Dependencies
--> Running transaction check
---> Package cx_Oracle.x86_64 0:5.2.1-1 will be installed
--> Finished Dependency Resolution
 
Dependencies Resolved
 
=======================================================================================
 Package        Arch        Version      Repository                               Size
=======================================================================================
Installing:
 cx_Oracle      x86_64      5.2.1-1      /cx_Oracle-5.2.1-11g-py27-1.x86_64      717 k
 
Transaction Summary
=======================================================================================
Install  1 Package
 
Total size: 717 k
Installed size: 717 k
Downloading packages:
Running transaction check
Running transaction test
Transaction test succeeded
Running transaction (shutdown inhibited)
  Installing : cx_Oracle-5.2.1-1.x86_64                                            1/1 
  Verifying  : cx_Oracle-5.2.1-1.x86_64                                            1/1 
 
Installed:
  cx_Oracle.x86_64 0:5.2.1-1                                                           
 
Complete!

After you install the cx_Oracle-5.2.1-1.x86_64 package, you can find the installed files with this rpm command:

rpm -ql cx_Oracle-5.2.1-1.x86_64

It lists:

/usr/lib64/python2.7/site-packages/cx_Oracle-5.2.1-py2.7.egg-info
/usr/lib64/python2.7/site-packages/cx_Oracle-5.2.1-py2.7.egg-info/PKG-INFO
/usr/lib64/python2.7/site-packages/cx_Oracle-5.2.1-py2.7.egg-info/SOURCES.txt
/usr/lib64/python2.7/site-packages/cx_Oracle-5.2.1-py2.7.egg-info/dependency_links.txt
/usr/lib64/python2.7/site-packages/cx_Oracle-5.2.1-py2.7.egg-info/top_level.txt
/usr/lib64/python2.7/site-packages/cx_Oracle.so
/usr/share/doc/cx_Oracle-5.2.1
/usr/share/doc/cx_Oracle-5.2.1/BUILD.txt
/usr/share/doc/cx_Oracle-5.2.1/README.txt
/usr/share/doc/cx_Oracle-5.2.1/samples
/usr/share/doc/cx_Oracle-5.2.1/samples/DatabaseChangeNotification.py
/usr/share/doc/cx_Oracle-5.2.1/samples/DatabaseShutdown.py
/usr/share/doc/cx_Oracle-5.2.1/samples/DatabaseStartup.py
/usr/share/doc/cx_Oracle-5.2.1/samples/ReturnLongs.py
/usr/share/doc/cx_Oracle-5.2.1/samples/ReturnUnicode.py
/usr/share/doc/cx_Oracle-5.2.1/samples/RowsAsInstance.py
/usr/share/doc/cx_Oracle-5.2.1/test
/usr/share/doc/cx_Oracle-5.2.1/test/3kArrayDMLBatchError.py
/usr/share/doc/cx_Oracle-5.2.1/test/3kNumberVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/3kStringVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/ArrayDMLBatchError.py
/usr/share/doc/cx_Oracle-5.2.1/test/BooleanVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/Connection.py
/usr/share/doc/cx_Oracle-5.2.1/test/Cursor.py
/usr/share/doc/cx_Oracle-5.2.1/test/CursorVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/DateTimeVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/IntervalVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/LobVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/LongVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/NCharVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/NumberVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/ObjectVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/SessionPool.py
/usr/share/doc/cx_Oracle-5.2.1/test/SetupTest.sql
/usr/share/doc/cx_Oracle-5.2.1/test/StringVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/TestEnv.py
/usr/share/doc/cx_Oracle-5.2.1/test/TimestampVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/test.py
/usr/share/doc/cx_Oracle-5.2.1/test/test3k.py
/usr/share/doc/cx_Oracle-5.2.1/test/test_dbapi20.py
/usr/share/doc/cx_Oracle-5.2.1/test/uArrayDMLBatchError.py
/usr/share/doc/cx_Oracle-5.2.1/test/uConnection.py
/usr/share/doc/cx_Oracle-5.2.1/test/uCursor.py
/usr/share/doc/cx_Oracle-5.2.1/test/uCursorVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/uDateTimeVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/uIntervalVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/uLobVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/uLongVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/uNumberVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/uObjectVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/uSessionPool.py
/usr/share/doc/cx_Oracle-5.2.1/test/uStringVar.py
/usr/share/doc/cx_Oracle-5.2.1/test/uTimestampVar.py

After you installed the software, you can test whether inside Python’s IDLE environment with the import command, like this:

Python 2.7.5 (default, Apr 10 2015, 08:09:05) 
[GCC 4.8.3 20140911 (Red Hat 4.8.3-7)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>> import cx_Oracle
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: libclntsh.so.11.1: cannot open shared object file: No such file or directory

This error indicates that Oracle Client software isn’t installed, which is true in this case. I only installed the Oracle Database 11g Express Edition. You need to download the Oracle Client software and install it as the root user.

You download the Oracle Client software from the Oracle web site. Assuming your Linux user’s name is student, you download the cx_Oracle library into the /home/student/Downloads directory. Then, you use the su or sudo command to become the root user.

As the root user, run the following yum command:

yum install -y ~student/Downloads/oracle-instantclient11.2-basic-11.2.0.4.0-1.x86_64.rpm

You should see the following output:

Loaded plugins: langpacks, refresh-packagekit
Examining /home/student/Downloads/oracle-instantclient11.2-basic-11.2.0.4.0-1.x86_64.rpm: oracle-instantclient11.2-basic-11.2.0.4.0-1.x86_64
Marking /home/student/Downloads/oracle-instantclient11.2-basic-11.2.0.4.0-1.x86_64.rpm to be installed
Resolving Dependencies
--> Running transaction check
---> Package oracle-instantclient11.2-basic.x86_64 0:11.2.0.4.0-1 will be installed
--> Finished Dependency Resolution
 
Dependencies Resolved
 
================================================================================
 Package
        Arch   Version
                      Repository                                           Size
================================================================================
Installing:
 oracle-instantclient11.2-basic
        x86_64 11.2.0.4.0-1
                      /oracle-instantclient11.2-basic-11.2.0.4.0-1.x86_64 179 M
 
Transaction Summary
================================================================================
Install  1 Package
 
Total size: 179 M
Installed size: 179 M
Downloading packages:
Running transaction check
Running transaction test
Transaction test succeeded
Running transaction (shutdown inhibited)
  Installing : oracle-instantclient11.2-basic-11.2.0.4.0-1.x86_64           1/1 
  Verifying  : oracle-instantclient11.2-basic-11.2.0.4.0-1.x86_64           1/1 
 
Installed:
  oracle-instantclient11.2-basic.x86_64 0:11.2.0.4.0-1                          
 
Complete!

You can create a Python program that checks your ability to connect to the Oracle database, like the following oracleConnection.py file:

# Import the Oracle library.
import cx_Oracle
 
try:
  # Create a connection.
  db = cx_Oracle.connect("student/student@xe")
 
  # Print a message.
  print "Connected to the  Oracle " + db.version + " database."
 
except cx_Oracle.DatabaseError, e:
  error, = e.args
  print >> sys.stderr, "Oracle-Error-Code:", error.code
  print >> sys.stderr, "Oracle-Error-Message:", error.message
 
finally:
  # Close cursor. 
  db.close()

You can run this from the Linux command line with the following syntax:

python oracleConnection.py

It should return the following string:

Connected to the Oracle 11.2.0.2.0 database.

Now, you can create a Python program that reads data from the Oracle database. The following oracleString.py file reads a string literal from the pseudo table dual:

# Import the Oracle library.
import cx_Oracle
 
try:
  # Create a connection.
  db = cx_Oracle.connect("student/student@xe")
 
  # Create a cursor.
  cursor = db.cursor()
 
  # Execute a query.
  cursor.execute("SELECT 'Hello world!' FROM dual")
 
  # Read the contents of the cursor.
  for row in cursor:
    print (row[0]) 
 
except cx_Oracle.DatabaseError, e:
  error, = e.args
  print >> sys.stderr, "Oracle-Error-Code:", error.code
  print >> sys.stderr, "Oracle-Error-Message:", error.message
 
finally:
  # Close cursor and connection. 
  cursor.close()
  db.close()

You can run this from the Linux command line with the following syntax:

python oracleString.py

It should return the following string:

Hello world!

Now, you can create a Python program that reads actual table data from the Oracle database (assuming you have a copy of my video store database). The following oracleTable.py file reads a string literal from the pseudo table dual:

# Import the Oracle library.
import cx_Oracle
 
try:
  # Create a connection.
  db = cx_Oracle.connect("student/student@xe")
 
  # Create a cursor.
  cursor = db.cursor()
 
  # Execute a query.
  cursor.execute("SELECT item_title, item_subtitle FROM item")
 
  # Read the contents of the cursor.
  for row in cursor:
    print (row[0], row[1]) 
 
except cx_Oracle.DatabaseError, e:
  error, = e.args
  print >> sys.stderr, "Oracle-Error-Code:", error.code
  print >> sys.stderr, "Oracle-Error-Message:", error.message
 
finally:
  # Close cursor and connection. 
  cursor.close()
  db.close()

You can run this from the Linux command line with the following syntax:

python oracleTable.py

It should return the following strings (only a subset of the returned values):

("Harry Potter and the Sorcer's Stone", 'Two-Disc Special Edition')
('Harry Potter and the Chamber of Secrets', 'Two-Disc Special Edition')
('Harry Potter and the Prisoner of Azkaban', 'Two-Disc Special Edition')
('Harry Potter and the Chamber of Secrets', None)
('Harry Potter and the Goblet of Fire', 'Widescreen Edition')
('Harry Potter and the Goblet of Fire', 'Two-Disc Special Edition')
('Harry Potter and the Order of the Phoenix', 'Widescreen Edition')
('The Lord of the Rings - Fellowship of the Ring', 'Widescreen Edition')
('The Lord of the Rings - Fellowship of the Ring', 'Platinum Series Special Extended Edition')
('The Lord of the Rings - Two Towers', 'Widescreen Edition')
('The Lord of the Rings - Two Towers', 'Platinum Series Special Extended Edition')
('The Lord of the Rings - The Return of the King', 'Widescreen Edition')
('The Lord of the Rings - The Return of the King', 'Platinum Series Special Extended Edition')
('Star Wars - Episode I', 'The Phantom Menace')
('Star Wars - Episode II', 'Attack of the Clones')
('Star Wars - Episode III', 'Revenge of the Sith')
('Star Wars - Episode IV', 'A New Hope')
('Star Wars - Episode V', 'The Empire Strikes Back')
('Star Wars - Episode VI', 'Return of the Jedi')

As always, I hope this helps others who want to work with Python and the Oracle database.

Written by maclochlainn

October 19th, 2016 at 1:47 am

Oracle Segment Fails

with 2 comments

The instance that I’ve built for my students in a Fedora VM is quite stable except for one feature. The feature is the hibernation process of the base operating system. Sometimes when the base operating system hibernates, it causes the Oracle shared memory segment to fail. When that happens you get the following error:

ERROR:
ORA-01034: ORACLE NOT available
ORA-27101: shared memory realm does NOT exist
Linux-x86_64 Error: 2: No such FILE OR DIRECTORY
Process ID: 0
SESSION ID: 0 Serial NUMBER: 0

I created the master sudoer account as the student user. The oracle user is configured so that you can’t log in to the Linux OS with it. To restart the instance you can do the following in a default Oracle 11g XE installation:

su - root

or, you can do this:

sudo sh

Then as the root user, you can sign on to the oracle user’s account by using the su command without a password, like:

su - oracle

As the user who installed the Oracle instance, you can connect to the database without a password after you source the environment file. The standard Oracle 11g XE environment file can be sources like this:

. /u01/app/oracle/product/11.2.0/xe/bin/oracle_env.sh

Alternatively, for my students there is a .bashrc file that they can manually source. It contains the following:

# Source global definitions
if [ -f /etc/bashrc ]; then
  . /etc/bashrc
fi
 
# Uncomment the following line if you don't like systemctl's auto-paging feature:
# export SYSTEMD_PAGER=
 
# User specific aliases and functions
. /u01/app/oracle/product/11.2.0/xe/bin/oracle_env.sh
 
# Wrap sqlplus with rlwrap to edit prior lines with the
# up, down, left and right keys.
sqlplus () 
{
    # Discover the fully qualified program name. 
    path=`which rlwrap 2>/dev/null`
    file=''
 
    # Parse the program name from the path.
    if [ -n ${path} ]; then
        file=${path##/*/}
    fi;
 
    # Wrap when there is a file and it is rewrap.
    if [ -n ${file} ] && [[ ${file} = "rlwrap" ]]; then
        rlwrap sqlplus "${@}"
    else
        echo "Command-line history unavailable: Install the rlwrap package."
        $ORACLE_HOME/bin/sqlplus "${@}"
    fi
}
 
# Set vi as a command line editor.
set -o vi

You can source the oracle user’s .bashrc account, like this:

. .bashrc

After you’ve sourced the environment, you can connect as the internal user with the following syntax:

sqlplus / AS sysdba

Connected as the internal user, run these two commands in sequence:

shutdown IMMEDIATE
startup

Then, you should be able to connect as the student user or another ordinary user with the following syntax:

CONNECT student/student

Hope this helps my students and those who want to know how to restart the Oracle instance.

Written by maclochlainn

October 16th, 2016 at 10:29 pm

Multitenant Architecture 12c

without comments

I thought this illustration was an interesting view of Oracle Database 12c’s Multitenant Architecture. It posted on ToadWorld.com today in a new article by Deiby Gomez and I thought it might be interesting for others.

multitenantarchitecture12c

You can see how to provision a pluggable database in this article on provisioning a pluggable database. As always, I hope it helps.

Written by maclochlainn

October 11th, 2016 at 10:07 am