vangef/BBGradebookOrganiser
1
1
Fork 0
Code Issues Releases Activity

'inspect submission files by hash' feature v0.1

Browse Source
This commit is contained in:
vangef
2023-02-28 22:48:34 +00:00
parent 00cf555efb
commit 980281c5fd
3 changed files with 79 additions and 44 deletions
Download Patch File Download Diff File Expand all files Collapse all files

44
hash_submissions.py
View File

@@ -1,44 +0,0 @@
### TESTING
### feature to hash all gradebook submission files, and check for duplicates across all students / submissions
### not fully implemented yet - only creates hashes and outputs to csv for manual inspection
import os, sys
from datetime import datetime
import csv
import hashlib
def hash_files_in_dir(dir_path: str, csv_suffix: str):
os.makedirs('csv', exist_ok=True)
csv_file_name = f'file_hashes_{csv_suffix}_{datetime.now().strftime("%Y%m%d-%H%M%S")}.csv'
csv_file = os.path.join('csv', csv_file_name)
with open(csv_file, 'w', newline='') as csvfile: # Open the output CSV file for writing
fieldnames = ['Student ID', 'file', 'sha256 hash']
writer = csv.DictWriter(csvfile, fieldnames=fieldnames)
writer.writeheader()
for subdir, dirs, files in os.walk(dir_path): # Loop through all files in the directory and generate hashes
for file in files:
if 'README.md' not in file:
directories = [d for d in os.path.abspath(subdir).split(os.path.sep)] # list of directories in the file path
student_id = directories[directories.index(csv_suffix)+1] # use the index of 'csv_suffix' which is the gradebook name, and get the next directory which is the student id
filepath = os.path.join(subdir, file)
with open(filepath, 'rb') as f:
filehash = hashlib.sha256(f.read()).hexdigest()
writer.writerow({'Student ID': student_id, 'file': filepath, 'sha256 hash': filehash})
def main():
submissions_dir_name = ' '.join(sys.argv[1:]) if len(sys.argv) > 1 else exit(f'\nNo submissions dir name given. Provide the name as an argument.\n\nUsage: python {sys.argv[0]} [submissions dir name]\n')
submissions_dir = os.path.join('BB_submissions', submissions_dir_name) # dir with extracted submissions
if os.path.isdir(submissions_dir):
hash_files_in_dir(submissions_dir, submissions_dir_name)
else:
exit(f'Directory {submissions_dir} does not exist.\nMake sure "{submissions_dir_name}" exists in "BB_submissions".')
if __name__ == '__main__':
main()

24
inspect_submissions.py Normal file
View File

@@ -0,0 +1,24 @@
import os, sys
import pandas as pd
from datetime import datetime
from utils.inspector import hash_submissions, suspicious_by_hash
def main():
submissions_dir_name = ' '.join(sys.argv[1:]) if len(sys.argv) > 1 else exit(f'\nNo submissions dir name given. Provide the name as an argument.\n\nUsage: python {sys.argv[0]} [submissions dir name]\nExample: python {sys.argv[0]} AssignmentX\n')
submissions_dir_path = os.path.join('BB_submissions', submissions_dir_name)
if not os.path.isdir(submissions_dir_path):
exit(f'Directory {submissions_dir_path} does not exist.\nMake sure "{submissions_dir_name}" exists in "BB_submissions".')
else:
hashes_csv_file_path = hash_submissions(submissions_dir_path)
csv = pd.read_csv(hashes_csv_file_path)
df = pd.DataFrame(csv) # df with all files and their hashes
df_suspicious = suspicious_by_hash(df) # df with all files with duplicate hash, excludes files from the same student id
csv_name = f'{submissions_dir_name}_suspicious_{datetime.now().strftime("%Y%m%d-%H%M%S")}.csv'
csv_out = os.path.join('csv', csv_name)
df_suspicious.to_csv(csv_out, index=False)
if __name__ == '__main__':
main()

55
utils/inspector.py Normal file
View File

@@ -0,0 +1,55 @@
import os
from datetime import datetime
import csv
import hashlib
import pandas as pd
CSV_DIR = os.path.join(os.getcwd(), 'csv')
def get_hashes_in_dir(dir_path: str) -> list:
hash_list = []
for subdir, dirs, files in os.walk(dir_path): # Loop through all files in the directory and generate hashes
for file in files:
filepath = os.path.join(subdir, file)
with open(filepath, 'rb') as f:
filehash = hashlib.sha256(f.read()).hexdigest()
hash_list.append({ 'file': filepath, 'sha256 hash': filehash})
return hash_list
def hash_submissions(submissions_dir_path: str):
os.makedirs(CSV_DIR, exist_ok=True)
submissions_dir_name = os.path.abspath(submissions_dir_path).split(os.path.sep)[-1]
csv_file_name = f'{submissions_dir_name}_file_hashes_{datetime.now().strftime("%Y%m%d-%H%M%S")}.csv'
csv_file_path = os.path.join(CSV_DIR, csv_file_name)
with open(csv_file_path, 'w', newline='') as csvfile: # Open the output CSV file for writing
fieldnames = ['Student ID', 'file', 'sha256 hash']
writer = csv.DictWriter(csvfile, fieldnames=fieldnames)
writer.writeheader()
for student_dir_name in os.listdir(submissions_dir_path):
student_dir_path = os.path.join(submissions_dir_path, student_dir_name)
hashes_dict = get_hashes_in_dir(student_dir_path)
for d in hashes_dict:
d.update({'Student ID': student_dir_name}) # update hash records with student id
writer.writerows(hashes_dict)
return csv_file_path
def get_suspicious_hashes(df: pd.DataFrame) -> list:
drop_columns = ['file']
df = df.drop(columns=drop_columns).sort_values('sha256 hash') # clear not needed colums & sort by hash
duplicate_hash = df.loc[df.duplicated(subset=['sha256 hash'], keep=False), :] # all files with duplicate hash - incl. files from the same student id
hash_with_multiple_student_ids = duplicate_hash.groupby('sha256 hash').agg(lambda x: len(x.unique())>1) # true if more than 1 unique student ids (= multiple student ids with same hash), false if unique (= same student id re-submitting with the same hash)
suspicious_hashes_list = hash_with_multiple_student_ids[hash_with_multiple_student_ids['Student ID']==True].index.to_list() # list with duplicate hashes - only if different student id (doesn't include attempts from same student id)
return suspicious_hashes_list
def suspicious_by_hash(df: pd.DataFrame) -> pd.DataFrame:
suspicious_hashes_list = get_suspicious_hashes(df)
files_with_suspicious_hash = df[df['sha256 hash'].isin(suspicious_hashes_list)] # excluding duplicate from same student id
return files_with_suspicious_hash.sort_values(['sha256 hash', 'Student ID'])