/cs330e-diplomacy-grading-script

Primary LanguagePython

How to run:

  1. go to the same folder as the makefile
  2. download the group JSON submissions from Canvas and save them in group_JSON_submissions
  3. download the evaluation JSON submissions from Canvas and save them in evaluation_JSON_submissions
  4. in the same directory, run make
  5. you can simply make again after any update to the submissions files
  6. output can be found in output directory (email_output.txt should contain most of the vital information; check the invalid JSON files for submissions that did not get evaluated because of the invalid JSON)

Additional utilities during grading:

Uncomment the lines involving diff_message in DiplomacyVetting.py for easier grading (it records the difference in output if there is any, which helps pinpoint the bugs in student code). Also see CommonMistakes.md for comments that helps with grading.

other notes

Make sure to read TODO.md for features that are yet implemented; some quirks of the script can be explained by the lack of certain features

modified from cs330e-collatz-grading-script:

https://github.com/peter1357908/cs330e-collatz-grading-script/

Main differences (updates):

  1. no longer checks for optional files (Collatz had a sphere challenge optional file)
  2. checks for Git flow files like .gitignore and .gitlab-ci.yml.
  3. commented out some print statements
  4. checks for group number (a new field added to the JSON file at the time of this project)
  5. account for multiple acceptance test files
  6. overall more modularization (easier to adapt to different projects)
  7. hard-checks the number of unit tests (3 for this project)
  8. uses subprocess.run() where possible (replacing the call(), check_call(), check_output(), and popen() calls)
  9. added a simple progress tracker
  10. silenced a lot of commands by directing output to subprocess.DEVNULL as well as silencing makefile commands with @