Difference between revisions of "OPS435 Assignment 2 for Section A"
Eric.brauer (talk | contribs) (→Second Milestone (due August 2)) |
Eric.brauer (talk | contribs) (→Second Milestone (due August 2)) |
||
(7 intermediate revisions by the same user not shown) | |||
Line 14: | Line 14: | ||
An example of the finished code your script might produce is this: | An example of the finished code your script might produce is this: | ||
− | <code><b>user@host ~ $ ./duim.py /usr/local/lib</b></code> | + | <code><b>user@host ~ $ ./duim.py -H /usr/local/lib</b></code> |
<pre> | <pre> | ||
− | 61 % [============ ] 160.2 | + | 61 % [============ ] 160.2 M /usr/local/lib/heroku |
− | 4 % [= ] 10.8 | + | 4 % [= ] 10.8 M /usr/local/lib/python2.7 |
− | 34 % [======= ] 90.4 | + | 34 % [======= ] 90.4 M /usr/local/lib/node_modules |
− | 0 % [ ] 8.0 | + | 0 % [ ] 8.0 K /usr/local/lib/python3.8 |
− | Total: 261.4 | + | Total: 261.4 M /usr/local/lib |
</pre> | </pre> | ||
− | The | + | Notice that total size of the target directory (/usr/local/lib) is around 260 Megabytes. Of that 260 Megabytes, 160 Megabytes can be found in the heroku subdirectory. |
+ | |||
+ | 160 MB represents 61% of the total 160 MB. The percentages don't have to add up to 100%, since with these arguments we are excluding files in the target directory. You may choose to add an option to your script to print files as well. | ||
+ | |||
+ | The bar chart in this example is 20 characters long, but this must be dynamic. The 20 characters does <i>not</i> the square brackets. The resolution of the bar chart must become more accurate as you increase the total size. For example, if the user specifies a length of 100 total characters, in this example 61 of those characters would be equal signs and 39 would be spaces. | ||
+ | |||
+ | The output of each subdirectory should include percentage, size in bytes (or Human readable if the user uses the -H option), the bar chart and the name of the subdirectory. | ||
+ | Specific formatting of the final output will be up to you, but should be formatted in such a way that the output is easy to read. (ie. use columns!) | ||
+ | |||
+ | You will be required to fulfill some specific requirements before completing your script. Read on... | ||
= Assignment Requirements = | = Assignment Requirements = | ||
+ | |||
+ | == Starting Code == | ||
+ | The first step for the assignment will be to accept the assignment using the invite code provided by your instructor. You will need to create a GitHub account to do this. (If you already have a GitHub account, you may use this). | ||
+ | |||
+ | In your repository you will find a file called <b>duim.py</b>. This file contains starting code. You will complete the assignment inside duim.py. <b>Do not rename this file or the functions inside, unit tests will fail and you will lose marks!</b> | ||
+ | |||
+ | Also in your repository you will find <b>checkA2.py</b>. You can use this check script to check your work. | ||
+ | |||
== Permitted Modules == | == Permitted Modules == | ||
<b><font color='blue'>Your python script is allowed to import only the <u>os, subprocess, argparse and sys</u> modules from the standard library.</font></b> | <b><font color='blue'>Your python script is allowed to import only the <u>os, subprocess, argparse and sys</u> modules from the standard library.</font></b> | ||
Line 121: | Line 138: | ||
To run the check script, enter the following: | To run the check script, enter the following: | ||
− | <code> | + | <code>python3 checkA2.py -f -v TestDirDict</code> |
You will be using a module in the standard library called <b>Argparse</b>. This will help handle more complex sets of options and arguments than simply using sys.argv. | You will be using a module in the standard library called <b>Argparse</b>. This will help handle more complex sets of options and arguments than simply using sys.argv. | ||
Line 135: | Line 152: | ||
<code><b>user@host ~ $ python3 duim.py -h</b></code> | <code><b>user@host ~ $ python3 duim.py -h</b></code> | ||
<pre> | <pre> | ||
− | usage: duim.py [-h] [-H] [-l LENGTH] target | + | usage: duim.py [-h] [-H] [-l LENGTH] [target] |
DU Improved -- See Disk Usage Report with bar charts | DU Improved -- See Disk Usage Report with bar charts | ||
Line 157: | Line 174: | ||
== Minimum Viable Product == | == Minimum Viable Product == | ||
Once you have achieved the Milestones, you will have to do the following to get a minimum viable product: | Once you have achieved the Milestones, you will have to do the following to get a minimum viable product: | ||
− | * In your <code>if __name__ == '__main__'</code> block, you will have to | + | * In your <code>if __name__ == '__main__'</code> block, you will have to call the parse_command_args function. Experiment with print statements so that you understand how each option and argument are stored. |
** If the user has entered more than one argument, or their argument isn't a valid directory, print an error message. | ** If the user has entered more than one argument, or their argument isn't a valid directory, print an error message. | ||
+ | ** If the user doesn't specify any target, use the current directory. | ||
* Call <code>call_du_sub</code> with the target directory. | * Call <code>call_du_sub</code> with the target directory. | ||
* Pass the return value from that function to <code>create_dir_dict</code> | * Pass the return value from that function to <code>create_dir_dict</code> |
Latest revision as of 10:56, 15 July 2021
Contents
Overview: du Improved
du
is a tool for inspecting directories. It will return the contents of a directory along with how much drive space they are using. However, it can be parse its output quickly, as it usually returns file sizes as a number of bytes:
user@host ~ $ du --max-depth 1 /usr/local/lib
164028 /usr/local/lib/heroku 11072 /usr/local/lib/python2.7 92608 /usr/local/lib/node_modules 8 /usr/local/lib/python3.8 267720 /usr/local/lib
You will therefore be creating a tool called duim (du improved). Your script will call du and return the contents of a specified directory, and generate a bar graph for each subdirectory. The bar graph will represent the drive space as percent of the total drive space for the specified directory. An example of the finished code your script might produce is this:
user@host ~ $ ./duim.py -H /usr/local/lib
61 % [============ ] 160.2 M /usr/local/lib/heroku 4 % [= ] 10.8 M /usr/local/lib/python2.7 34 % [======= ] 90.4 M /usr/local/lib/node_modules 0 % [ ] 8.0 K /usr/local/lib/python3.8 Total: 261.4 M /usr/local/lib
Notice that total size of the target directory (/usr/local/lib) is around 260 Megabytes. Of that 260 Megabytes, 160 Megabytes can be found in the heroku subdirectory.
160 MB represents 61% of the total 160 MB. The percentages don't have to add up to 100%, since with these arguments we are excluding files in the target directory. You may choose to add an option to your script to print files as well.
The bar chart in this example is 20 characters long, but this must be dynamic. The 20 characters does not the square brackets. The resolution of the bar chart must become more accurate as you increase the total size. For example, if the user specifies a length of 100 total characters, in this example 61 of those characters would be equal signs and 39 would be spaces.
The output of each subdirectory should include percentage, size in bytes (or Human readable if the user uses the -H option), the bar chart and the name of the subdirectory. Specific formatting of the final output will be up to you, but should be formatted in such a way that the output is easy to read. (ie. use columns!)
You will be required to fulfill some specific requirements before completing your script. Read on...
Assignment Requirements
Starting Code
The first step for the assignment will be to accept the assignment using the invite code provided by your instructor. You will need to create a GitHub account to do this. (If you already have a GitHub account, you may use this).
In your repository you will find a file called duim.py. This file contains starting code. You will complete the assignment inside duim.py. Do not rename this file or the functions inside, unit tests will fail and you will lose marks!
Also in your repository you will find checkA2.py. You can use this check script to check your work.
Permitted Modules
Your python script is allowed to import only the os, subprocess, argparse and sys modules from the standard library.
Required Functions
You will need to complete the functions inside the provided file called duim.py
. The provided checkA2.py
will be used to test these functions.
-
call_du_sub()
should take the target directory as an argument and return a list of strings returned by the command du -d 1<target directory>.- Use subprocess.Popen.
- '-d 1' specifies a max depth of 1. Your list shouldn't include files, just a list of subdirectories in the target directory.
- Your list should not contain newline characters.
-
percent_to_graph()
should take two arguments: percent and the total chars. It should return a 'bar graph' as a string.- Your function should check that the percent argument is a valid number between 0 and 100. It should fail if it isn't. You can
raise ValueError
in this case. - total chars refers to the total number of characters that the bar graph will be composed of. You can use equal signs
=
or any other character that makes sense, but the empty space must be composed of spaces, at least until you have passed the first milestone. - The string returned by this function should only be composed of these two characters. For example, calling
percent_to_graph(50, 10)
should return:
- Your function should check that the percent argument is a valid number between 0 and 100. It should fail if it isn't. You can
'===== '
Please note that the characters should not be part of the output, they are here to indicate that this is a string!
-
create_dir_dict
should take a list as the argument, and should return a dictionary.- The list can be the list returned by
call_du_sub()
. - The dictionary that you return should have the full directory name as key, and the number of bytes in the directory as the value. This value should be an integer. For example, using the example of /usr/local/lib, the function would return:
- The list can be the list returned by
{'/usr/local/lib/heroku': 164028, '/usr/local/lib/python2.7': 11072, ...}
Additional Functions
You may create any other functions that you think appropriate, especially when you begin to build additional functionality. Part of your evaluation will be on how "re-usable" your functions are, and sensible use of arguments and return values.
Use of GitHub
You will be graded partly on the quality of your Github commits. You may make as many commits as you wish, it will have no impact on your grade. The only exception to this is assignments with very few commits. These will receive low marks for GitHub use and may be flagged for possible academic integrity violations. Assignments that do not adhere to these requirements may not be accepted.
Professionals generally follow these guidelines:
- commit their code after every significant change,
- the code should hopefully run without errors after each commit, and
- every commit has a descriptive commit message.
After completing each function, make a commit and push your code.
After fixing a problem, make a commit and push your code.
GitHub is your backup and your proof of work.
These guidelines are not always possible, but you will be expected to follow these guidelines as much as possible. Break your problem into smaller pieces, and work iteratively to solve each small problem. Test your code after each small change you make, and address errors as soon as they arise. It will make your life easier!
Coding Standard
Your python script must follow the following coding guide:
Documentation
There are three types of comments in programming and your assignment should contain each:
- The top-level docstring should contain information about what your script does. This is included in the duim.py file. Please complete the top-level docstring.
- Use Python's function docstrings to document how each of the functions work. The docstring should describe what each function does.
- Your script should also include in-line comments to explain anything that isn't immediately obvious to a beginner programmer. For these comments, it's always better to explain why your code is doing what it does rather than what it's doing. Also: It is expected that you will be able to explain how each part of your code works in detail.
Authorship Declaration
All your Python code for this assignment must be placed in the provided Python file called duim.py. Do not change the name of this file. Please complete the declaration as part of the top-level docstring in your Python source code file (replace "Student Name" with your own name).
Submission Guidelines and Process
Clone Your Repo (ASAP)
The first step will be to clone the Assignment 2 repository. The invite link will be provided to you by your professor. You will need a free GitHub account to complete this assignment. If you already have an existing GitHub account, you may use it.
The repo will contain a check script, a README file, and the file where you will enter your code.
The First Milestone (due July 26)
For the first milestone you will have two functions to complete.
-
call_du_sub
will take one argument and return a list. The argument is a target directory. The function will usesubprocess.Popen
to run the command du -d l <target_directory>. -
percent_to_graph
will take two arguments and return a string.
In order to complete percent_to_graph()
, it's helpful to know the equation for converting a number from one scale to another.
In this equation, ``x`` refers to your input value percent and ``y`` will refer to the number of symbols to print. The max of percent is 100 and the min of percent is 0. Be sure that you are rounding to an integer, and then print that number of symbols to represent the percentage. The number of spaces that you print will be the inverse.
Test your functions with the Python interpreter. Use python3
, then:
import duim duim.percent_to_graph(50, 10)
To test with the check script, run the following:
python3 checkA2.py -f -v TestPercent
python3 checkA2.py -f -v TestDuSub
Second Milestone (due August 2)
For the second milestone you will have two more functions to complete.
-
create_dir_dict
will take your list fromcall_du_sub
and return a dictionary.- Every item in your list should create a key in your dictionary.
- Your dictionary values should be a number of bytes.
For example: {'/usr/lib/local': 33400}
- Again, test using your Python interpreter or the check script.
To run the check script, enter the following:
python3 checkA2.py -f -v TestDirDict
You will be using a module in the standard library called Argparse. This will help handle more complex sets of options and arguments than simply using sys.argv.
Refer to the argparse documentation to complete the parse_command_args
function. At minimum, your assignment should handle the following options and arguments:
- -h will print a usage message. This will automatically be created by argparse itself, you will not need to implement this. However, refer carefully to the sample output and ensure that your help message matches the required output.
- -H will print file sizes in Human readable format. For example, 1024 bytes will be printed as 1K, 1024 kilobytes will be printed as 1M, and so on.
- -l <number> will set the maximum length of the bar graph. The default should be 20 character. This option will require an option argument that is an integer.
- Your script will also require one positional argument which contains the target directory for scanning.
Your assignment should be able to produce the following:
user@host ~ $ python3 duim.py -h
usage: duim.py [-h] [-H] [-l LENGTH] [target] DU Improved -- See Disk Usage Report with bar charts positional arguments: target The directory to scan. optional arguments: -h, --help show this help message and exit -H, --human-readable print sizes in human readable format (e.g. 1K 23M 2G) -l LENGTH, --length LENGTH Specify the length of the graph. Default is 20. Copyright 2021
Use the following to test your code:
python3 checkA2.py -f -v TestArgs
Minimum Viable Product
Once you have achieved the Milestones, you will have to do the following to get a minimum viable product:
- In your
if __name__ == '__main__'
block, you will have to call the parse_command_args function. Experiment with print statements so that you understand how each option and argument are stored.- If the user has entered more than one argument, or their argument isn't a valid directory, print an error message.
- If the user doesn't specify any target, use the current directory.
- Call
call_du_sub
with the target directory. - Pass the return value from that function to
create_dir_dict
- You may wish to create one or more functions to do the following:
- Use the total size of the target directory to calculate percentage.
- For each subdirectory of target directory, you will need to calculate a percentage, using the total of the target directory.
- Once you've calculated percentage, call
percent_to_graph
with a max_size of your choice. - For every subdirectory, print at least the percent, the bar graph, and the name of the subdirectory.
- The target directory should not have a bar graph.
Additional Features
After completing the above, you are expected to add some additional features. Some improvements you could make are:
- Format the output in a way that is easy to read.
- Add colour to the output.
- Include files in the output.
- Include a threshold, so that results that are less than a user-specified size get excluded from results.
- Add more error checking, print a usage message to the user.
- Accept more options from the user.
- Sort the output by percentage, or by filename.
It is expected that the additional features you provided should be useful, non-trivial, they should not require super-user privileges and should not require the installation of additional packages to work. (ie: I shouldn't have to run pip to make your assignment work).
The Assignment (due August 6, 11:59pm)
- Be sure to make your final commit before the deadline. Don't forget to also use
git push
to push your code into the online repository! - Then, copy the contents of your duim.py file into a Word document, and submit it to Blackboard. I will use GitHub to evaluate your deadline, but submitting to Blackboard tells me that you wish to be evaluated.
Rubric
Task | Maximum mark | Actual mark |
---|---|---|
Program Authorship Declaration | 5 | |
required functions design | 5 | |
required functions readability | 5 | |
main loop design | 10 | |
main loop readability | 10 | |
output function design | 5 | |
output function readability | 5 | |
additional features implemented | 20 | |
docstrings and comments | 5 | |
First Milestone | 10 | |
Second Milestone | 10 | |
github.com repository: Commit messages and use | 10 | |
Total | 100 |
Due Date and Final Submission requirement
Please submit the following files by the due date:
- [ ] your python script, named as 'duim.py', in your repository, and also submitted to Blackboard, by August 6 at 11:59pm.