grading

Atomic python scripts to make grading C++ Programming for Majors students' homework submissions less time consuming.

Once run.x is executed, various penalties will be enforced, automatically deducting points and listing them in a generated <username>_comments.txt.

Traversing through the created directory tree and reviewing the generated output files, diff logs, and source code will enable a much quicker manual grading for the rest of the grading rubric.

After the automatic and manual grading is complete, grade.py can be executed to calculate the total grade per student for easy uploading.

• extract.py - Only extracts students from a tarball listed in roster.csv and organizes them by section.
• format.py - Only extracts students' submissions from a zipfile and reformats them into a tarball expected by the rest of the scripts
• test.py - Using various input/output files and a duedate.txt, the extracted source code is further organized and tested
• grade.py - Iterates over all <username>_comments.txt, printing the total grade per student in a GRADES.csv
• run.x - A bash script that executes extract.py then test.py
• pseudorand.c - A short C overload of the STL rand function which allows passing a sequence of numbers to the homework executable

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.

Prerequisites

What things you need to install the software and how to install them

python3
g++

Installing

Install python3

sudo apt-get install python3

Test Installation

python3 --version

Should be python3.5+

Install g++

sudo apt-get install g++

Test Installation

g++ --version

Check to see if any version appears

Requirements

After installing python3 and cloning the repo, the scripts expect certain files to be included within the grading directory.

Files:

roster.csv - example

duedate.txt - example

• First line is the due date for full points.
• Second line is the due date for the late penalty.

input.txt

output.txt

<SUBMISSIONS>.tar

• The submissions within the tarball are expected as <username>_<assignmentFileName>.cpp
• usernames in roster.csv should be a subset of those in the files in the tarball

Optional:

rand1.txt

• Each rand file must be associated with a matching pseudorand.c with the same integer suffix

Directory Structure

• hw directory should contain an integer suffix denoting which assignment it is if more than one hw directories exist in the repo.
• ex: hw1/ hw2/ hw3/ etc
• input.txt and output.txt should contain an integer suffix differentiating the test suites and maintaing the mapping of each input file with its output file.
• ex: input1.txt input2.txt input3.txt ... output1.txt output2.txt output3.txt ...
• rand1.txt should only be included if you will be using it. Also if you have multiple sequences, they should each be in their own rand#.txt
• Note: that if you have multiple rand files you also need multiple pseudorand.c with a matching integer suffix. Make sure to edite the file name in pseudorand.c as well.

grading/
  extract.py
  format.py
  grade.py
  pseudorand.c
  hw/
    duedate.txt
    example.tar
    input.txt
    output.txt
    rand1.txt
  roster.csv
  run.x
  test.py

Running

Method 1

./run.x

Method 2

For COP3014 only:

./format.py

For COP3363 & COP3014:

./execute.py
./test.py

Method 3

For COP3014 only:

python3 format.py

For COP3363 & COP3014:

python3 execute.py
python3 test.py

After manual grading

Method 1

./grade.py

Method 2

python3 grade.py

Output

Logging is produced to stdout showing at what stage the script is at as well as anomalies such as:

• Late penalties
• Compilation failures and warnings
• Detected infinite loops

After successful completion of the scripts, several files and directories will be generated within hw/:

• section?/
• ? is substituted with an integer representing a specific section.
• section?/compiled - source code that compiles with or without warnings are moved here
• section?/failed - source code that fails to compile is moved here
• section?/<username>_comments.txt - automatically generated but must be edited during manual grading
• GRADES.csv - generated based on the docked or added points in the <username>_comments.txt files after manual grading for easy updating of the grade book