Rchan sandbox

From CDOT Wiki
Revision as of 11:40, 3 November 2019 by Rchan (talk | contribs) (Investigation III)
Jump to: navigation, search

OPS435 Python Lab 7

OBJECTIVES

  1. Create functions to process programmer-defined type objects.
  2. Bind functions into methods for programmer-defined type objects.

Overview

Object-oriented programming is conceptually one level higher than simply structured programming style as you've experienced in Bash or C. In this lab, we're going to study some features of classes and programmer-defined objects by looking at a few object-oriented programming examples using the Python language.

Reference

  • Time object from Think Python Chapter 16 and 17
  • Date object from OPS435 Assignment 1

Investigation I: Objects and Functions

Part 1 - Simple Object Class with external functions

In this part, we consider a time object which has three data attributes, namely: hour, minute, and second. The following Python script lab7a.py provides the blue print for building such a time object and also defines three external functions: format_time(), sum_times(), and valid_time(), which can manipulate the time object.

#!/usr/bin/env python3

class Time:
   """Simple object type for time of the day.
      data attributes: hour, minute, second
   """
   def __init__(self,hour=12,minute=0,second=0):
       """constructor for time object""" 
       self.hour = hour
       self.minute = minute
       self.second = second

def format_time(t):
    """Return time object (t) as a formatted string"""
    return '%.2d:%.2d:%.2d' % (t.hour, t.minute, t.second)

def sum_times(t1, t2):
    """Add two time objests and return the sum."""
    sum = Time(0,0,0)
    sum.hour = t1.hour + t2.hour
    sum.minute = t1.minute + t2.minute
    sum.second = t1.second + t2.second
    
    return sum

def valid_time(t):
    """check for the validity of the time object attributes:
       24 > hour > 0, 60 > minute > 0, 60 > second > 0 """
    if t.hour < 0 or t.minute < 0 or t.second < 0:
       return False
    if t.minute >= 60 or t.second >= 60 or t.hour >= 24:
       return False
    return True
Perform the following steps:
1. Download or create the above Pythone script lab7a.py in your ~/ops435/lab7 directory.
2. Create a new Python script named lab7a1.py in the lab7 directory:
cd ~/ops435/lab7
vi ~/ops435/lab7/lab7a1.py
3. Place the following content inside the new python file lab7a1.py and save it:
#!/usr/bin/env python3
# Student ID: [seneca_id]
from lab7a import *
t1 = Time(8,0,0)
t2 = Time(8,55,0)
t3 = Time(9,50,0)

td = Time(0,50,0)

tsum1 = sum_times(t1,td)
tsum2 = sum_times(t2,td)
tsum3 = sum_times(t3,td)   

ft = format_time
print(ft(t1),'+',ft(td),'-->',ft(tsum1))
print(ft(t2),'+',ft(td),'-->',ft(tsum2))
print(ft(t3),'+',ft(td),'-->',ft(tsum3))
4. The above python script lab7a1.py should produce the following output when executed:
[rchan@centos7 lab7]$ python3 lab7a1.py
08:00:00 + 00:50:00 --> 08:50:00
08:55:00 + 00:50:00 --> 08:105:00
09:50:00 + 00:50:00 --> 09:100:00
5. If you pay a little bit attention on the outputs, you should notice that the sum_times() function did not carry 60 minutes over to hour. It also will not carry 60 seconds over to minute.
6. To fix that problem, perform the following:
(a) Edit the file lab7a.py and modify the function sum_times(). Add code to check the values of the time object's three attributes. If the value of the minute attribute is more than 60, carry it over to the hour until it is less than 60. Do the same for the value of the second attribute.
(b) The illustration below indicates where you should perform the carry over operation in the sum_times() function in lab7a.py:
...
...
def sum_times(t1, t2):
    """Add two time objests and return the sum."""
    sum = Time(0,0,0)
    sum.hour = t1.hour + t2.hour
    sum.minute = t1.minute + t2.minute
    sum.second = t1.second + t2.second
    
    #[ insert python code here to check for minute and second 
    #[ attribute here, and carry over when necessary
    #[
    return sum
...
...
(c) Save the file.
(d) Test run your script lab7a1.py again and make sure that it produces the output to match the following exactly:
[rchan@centos7 lab7]$ python3 lab7a1.py
08:00:00 + 00:50:00 --> 08:50:00
08:55:00 + 00:50:00 --> 09:45:00
09:50:00 + 00:50:00 --> 10:40:00
7. Download the checking script and check your work. Enter the following commands from the bash shell.
cd ~/ops435/lab7/
pwd #confirm that you are in the right directory
ls CheckLab7.py || wget https://ict.senecacollege.ca/~raymond.chan/ops435/labs/LabCheckScripts/CheckLab7.py
python3 ./CheckLab7.py -f -v lab7a
8. Before proceeding, make certain that you identify all errors in lab7a.py. When the checking script tells you everything is OK - proceed to the next step.



Part 2 - Pure Function and Modifiers

In Part 1, the sum_times() function will not make any change to any of the two time objects passed to it. This is called a pure function. In this part, we are going to create a function which changes the value (the three attributes) of a time object based on the integer value passed to the function. We call such type of function a modifier. The function we are going to add is called change_time(time, seconds): where time is the time object the function is going to modify, and seconds is the number of seconds the function will add to the time object.

1. Make a copy of lab7a.py and name it as lab7b.py in the ~/ops435/lab7 directory
cd ~/ops435/lab7
cp lab7a.py lab7b.py
2. Add the following new function called change_time(time, seconds) after the sum_times() function into lab7b.py:
def change_time(time, seconds):
    time.second += seconds
    if valid_time(time) != True:
        while time.second >= 60:
             time.second -= 60
             time.minute +=1
        while time.minute >= 60:
             time.minute -= 60
             time.hour += 1
    return None
3. Save the file.
4. Bring up an interactive Python3 shell to test the new function.
Testing change_time(time, seconds) with import
[rchan@centos7 lab7]$ python3
Python 3.4.9 (default, Aug 14 2018, 21:28:57) 
[GCC 4.8.5 20150623 (Red Hat 4.8.5-28)] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from lab7b import *
>>> time1 = Time(9,50,0)
>>> format_time(time1)
'09:50:00'
>>> seconds = 1800
>>> seconds
1800
>>> change_time(time1, seconds)
>>> format_time(time1)
'10:20:00'
5. If you encounter any syntax error or exception, please fix the error before moving on to the next step.
6. Now try to set seconds to a negative value, e.g. -1800, and then call the change_time() function. Will it change the time1 object back to '09:50:00'? Let's try it out:
>>> format_time(time1)
'10:20:00'
>>> seconds = -1800
>>> change_time(time1, seconds)
>>> format_time(time1)
'10:20:-1800'
7. The result is not correct! So we have to modify and update the change_time() function so that it can handle positive and negative value correctly. The following steps are necessary:
(a) We must first add code to check on the second attribute of the time object to make sure that it is not less than zero. If it is, we must borrow one from the minute attribute and add 60 to the second attribute. We have to repeat this checking until the second attribute is no long less than zero.
          while time.second < 0:
                time.minute -= 1
                time.second += 60
(b) We then have to add code to check on the minute attribute. If it is less than zero, we must borrow 1 from the 'hour' attribute and add 60 to the minute attribute. We have to repeat this checking until the minute attribute is no long less than zero.
          while time.minute < 0:
                time.hour -= 1
                time.minute += 60
8. After updating the change_time() function with the above additional code, save the file and test it again in a Python interactive shell by importing the new version of the function.
[rchan@centos7 lab7]$ python3
Python 3.4.9 (default, Aug 14 2018, 21:28:57) 
[GCC 4.8.5 20150623 (Red Hat 4.8.5-28)] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from lab7b import *
>>> time1 = Time(9,50,0)
>>> format_time(time1)
'09:50:00'
>>> seconds = 1800
>>> change_time(time1, seconds)
>>> format_time(time1)
'10:20:00'
>>> seconds = -1800
>>> change_time(time1, seconds)
>>> format_time(time1)
'09:50:00'
9. Make sure that your change_time() function works correctly for both positive and negative values of 'seconds'.
10. Download the checking script and check your work. Enter the following commands from the bash shell.
cd ~/ops435/lab7/
pwd #confirm that you are in the right directory
ls CheckLab7.py || wget https://ict.senecacollege.ca/~raymond.chan/ops435/labs/LabCheckScripts/CheckLab7.py
python3 ./CheckLab7.py -f -v lab7b
11. Before proceeding, make certain that you identify all errors in lab7b.py. When the checking script tells you everything is OK - proceed to the next step.



Part 3 - Another approach to perform operation on time objects

In previous parts, we recognise the fact that a Time object is essentially a three-digit number in base 60. The "second" attribute is the "ones colume", the "minute" attribute is the "sixties column", and the hour attribute is the "60 x 60" column. This is why we have to carry over "second" to "minute" when the value of "second" exceeds 60, and we have to do the same for the "minute" attribute.

In this part, we are going to use a different approach to manage calculation on Time object. We know that there are 86400 (24 x 60 x 60) seconds in a day. It is relatively easy to convert a Time object in the form of hour:minute:second into an integer value which represent the time in seconds from mid-night. It is equally straight forward to convert a integer number of seconds between 0 and 86399 into hour:minute:second format. Here are two functions: sec_to_time(seconds) and time_to_sec(time) which can do just that:
...
...
def time_to_sec(time):
    '''convert a time object to a single integer representing the number of seconds from mid-night'''
    minutes = time.hour * 60 + time.minute
    seconds = minutes * 60 + time.second
    return seconds

def sec_to_time(seconds):   
    '''convert a given number of seconds to a time object in hour,minute,second format'''
    time = Time()
    minutes, time.second = divmod(seconds, 60)
    time.hour, time.minute = divmod(minutes,60)
    return time
...
...
1. Make a copy of lab7b.py and name it as lab7c.py in the ~/ops435/lab7 directory:
cd ~/ops435/lab7
cp lab7b.py lab7c.py
2. Add the above two new functions: time_to_sec() and sec_to_time() to lab7c.py after the change_time() function.
3. Save the file, and test the two function in an Python interactive shell:
[rchan@centos7 lab7]$ python3
Python 3.4.9 (default, Aug 14 2018, 21:28:57) 
[GCC 4.8.5 20150623 (Red Hat 4.8.5-28)] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> from lab7c import *
>>> t1 = Time(9,50,0)
>>> sec_t1 = time_to_sec(t1)
>>> sec_t1
35400
>>> t1_sec = sec_to_time(sec_t1)
>>> t1_sec
<lab7c.Time object at 0x7f89f656e048>
>>> format_time(t1_sec)
'09:50:00'
>>>
4. You should get the same result as shown above. If not, please troubleshoot and fix any errors.
5. Edit your lab7c.py again. Modify the other two functions: sum_times() and change_time() to use the two new functions to perform the computation operation.
6. Create a new Python script named lab7c1.py and place the following content in it:
#!/usr/bin/env python3
# Student ID: [seneca_id]
from lab7c import *
t1 = Time(8,0,0)
t2 = Time(8,55,0)
t3 = Time(9,50,0)

td = Time(0,50,0)

tsum1 = sum_times(t1,td)
tsum2 = sum_times(t2,td)
t3None = change_time(t3,1800)   

ft = format_time
print(ft(t1),'+',ft(td),'-->',ft(tsum1))
print(ft(t2),'+',ft(td),'-->',ft(tsum2))
print('09:50:00 + 1800 sec','-->',ft(t3))
7. Test run your script lab7c1.py, it should produce the following output:
[rchan@centos7 lab7]$ python3 lab7c1.py
08:00:00 + 00:50:00 --> 08:50:00
08:55:00 + 00:50:00 --> 09:45:00
09:50:00 + 1800 sec --> 10:20:00
8. Download the checking script and check your work. Enter the following commands from the bash shell.
cd ~/ops435/lab7/
pwd #confirm that you are in the right directory
ls CheckLab7.py || wget https://ict.senecacollege.ca/~raymond.chan/ops435/labs/LabCheckScripts/CheckLab7.py
python3 ./CheckLab7.py -f -v lab7a
9. Before proceeding, make certain that you identify all errors in lab7a.py. When the checking script tells you everything is OK - proceed to the next step.



Investigation II

In the previous investigation, the functions that were defined for manipulating our time object are not tied directly to our time object. Given our time object alone, we won't be able to tell that there exist a function called sum_times() which can be used to add two time objects and return their sum. To tie up those functions to our time objects, we only need to move those functions definition under the class block which define our Time object.

Part 1 - Classes and Methods for our Time objects

The following illustration shows how we can change external functions to become object methods (aka class functions) for our Time object. It is simply by moving the function definitions to be under the class definition for the Time object:

#!/usr/bin/env python3
# Student ID: rchan
class Time:
   """Simple object type for time of the day.
      data attributes: hour, minute, second
      function attributes: __init__, __str__, __repr__
                           time_to_sec, format_time,
                           change_time, sum_time
   """
   def __init__(self,hour=12,minute=0,second=0):
       """constructor for time object""" 
       self.hour = hour
       self.minute = minute
       self.second = second
   
   def format_time(self):
       """Return time object (t) as a formatted string"""
       return '%.2d:%.2d:%.2d' % (self.hour, self.minute, self.second)

   def sum_times(self, t2):
       """Add two time objests and return the sum."""
       sum = Time(0,0,0)
       sec1 = time_to_sec(self)
       sec2 = time_to_sec(self)
       sum = sec_to_time(sec1 + sec2)
       return sum

   def change_time(self, seconds):
       time_seconds = time_to_sec(self)
       nt = sec_to_time(time_seconds + seconds)
       time.hour, time.minute, time.second = nt.hour, nt.minute, nt.second 
       return None

   def time_to_sec(self):
       '''convert a time object to a single integer representing the 
       number of seconds from mid-night'''
       minutes = self.hour * 60 + self.minute
       seconds = minutes * 60 + self.second
       return seconds

   def valid_time(t):
       """check for the validity of the time object attributes:
       24 > hour > 0, 60 > minute > 0, 60 > second > 0 """
       if t.hour < 0 or t.minute < 0 or t.second < 0:
          return False
       if t.minute >= 60 or t.second >= 60 or t.hour >= 24:
          return False
       return True

def sec_to_time(seconds):
    '''convert a given number of seconds to a time object in 
       hour, minute, second format'''
    time = Time()
    minutes, time.second = divmod(seconds, 60)
    time.hour, time.minute = divmod(minutes, 60)
    return time

Please notice that the function named sec_to_time() did not get moved. It remains as an external function.

1. Create a new file and name it as lab7d.py and place the code above in it.


Part 2 - Special object methods

Part 3 - Operator overloading

Investigation III - Scope

Scope means where an object can be accessed, and how long that object exists.

In this investigation, we will look at the types of scope you're most likely to run into and which are most likely to cause trouble for you. We will not be explicitly looking at class scope, nested functions, built-in objects, constants, and mutability.

Part 1: Local Scope

In Python any object that is created inside a function has local scope. That means that it is accessible only by code inside that function, and is not accessible by any code outside of the function. In other languages this concept is related to block scope but that does not exist in python. Every block inside a function has the same scope.

Try the following code. Have each in a separate Python file.

lab7g.py - local scope:

#!/usr/bin/env python3
# Student ID: [seneca_id]
def function1():
    # This object 'a' is completely unrelated to the object 'a' in function2():
    a = 'object_function1'
    print('print() call on a in function1:',a)

def function2():
    # This variable 'a' is completely unrelated to the variable 'a' in function1():
    a = 'function2_object'
    print('print() call on a in function2:',a)

# Note that you cannot access the value of object '''a''' created in function1()
# or function2() with the print() functions after the calling function1() and function2()
# All the print() retrieved the value of object '''a''' defined in the main script.
a = 'object_in_main'
print('First print() call on a:',a)
function1()
print('Second print() call on a:',a)
function2()
print('Third print() call on a:',a)

Part 2: Global Scope

Sometimes you want to have an object accessible from anywhere in your program, including inside and outside any functions. Here's an example:

lab7h.py - global scope

#!/usr/bin/env python3
# Student ID: [seneca_id]
def function1():
    print('print() in function1 on schoolName:',schoolName)

def function2():
    print('print() in function2 on schoolName:',schoolName)

schoolName = 'Seneca College'
print('print() in main on schoolName:',schoolName)
function1()
print('print() in main on schoolName:',schoolName)
function2()
print('print() in main on schoolName:',schoolName)

Note that the same thing is printed over and over because the schoolName object is defined outside a function which makes it global which makes it accessible from anywhere.

Python has one weird quirk when it comes to global scope: if you assign something to an existing object inside a function - it will assume you want to create a new object in that function's local scope. That will hide the global object inside the function unless you declare it explicitly with the global keyword: lab7i.py - global keyword

#!/usr/bin/env python3
# Student ID: [seneca_id]
def function1():
    schoolName = 'SICT'
    print('print() in function1 on schoolName:',schoolName)

def function2():
    global schoolName
    schoolName = 'SSDO'
    print('print() in function2 on schoolName:',schoolName)

schoolName = 'Seneca'
print('print() in main on schoolName:',schoolName)
function1()
print('print() in main on schoolName:',schoolName)
function2()
print('print() in main on schoolName:',schoolName)

Note that the function1() call does not modify the global schoolName object but function2() does.

As an exercise: modify the example above so that it would print the following, by changing only the scope of some objects. Save the program as lab7i.py:

Seneca
SICT
SICT
SSDO
SICT

Object/Instance Scope

Every object can have attributes that exist for that object only. You create and access those attribute with the name of the object itself using the dot '.' notation. Note that these may not part of the class object's attribute. Each object has its own set of instance attributes. You will have seen that when you created objects in the Classes and Objects section above.

LAB 7 SIGN-OFF (SHOW INSTRUCTOR)

Have Ready to Show Your Instructor:
Output of: ./CheckLab7.py -f -v
Output of: cat lab7a.py lab7b.py lab7c.py lab7d.py
Files to be submitted individually to Blackboard:
Name the output of ./CheckLab7.py -f -v as lab7_[seneca_id].txt
Python script files for this lab: lab7a.py, lab7b.py, lab7c.py, and lab7d.py.

Lab Review

  1. What is a program-defined type object?
  2. What python keyword is used to create a blue-print for a new type of objects?
  3. What are the two types of attributes a Python object may have?
  4. Where does a program-defined python object get its methods from?
  5. What is the difference between a pure function and a modifier function?
  6. What does the time_to_sec() function do?
  7. What does the sec_to_time() function do?
  8. What is the reason for not making sec_to_time() a mothod for the Time object?
  9. How do you make an external function to be a method of a programmer-defined object?
  10. What does the special method __init__ do?
  11. What does the special method __str__ do?
  12. What does the special method __repr__ do?
  13. What does operator overloading refer to in Python?
  14. What does the term scope mean in Python?
  15. When do you want to create an object with global scope?
  16. What scope does an object have if it is created in a function?