vangef/BBGradebookOrganiser
1
1
Fork 0
Code Issues Releases Activity

skip "empty" (based on MIN FILE SIZE) files from inspection

Browse Source
This commit is contained in:
vangef
2025-02-21 17:47:55 +00:00
parent 5f91e08b00
commit 5a2d03db7d
1 changed files with 7 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

11
utils/inspector.py
View File

@@ -5,7 +5,7 @@ import hashlib
import pandas as pd import pandas as pd
from functools import partial from functools import partial
from utils.settings import CSV_DIR, BB_GRADEBOOKS_DIR, BB_SUBMISSIONS_DIR from utils.settings import CSV_DIR, BB_GRADEBOOKS_DIR, BB_SUBMISSIONS_DIR, MIN_FILESIZE_IN_BYTES
def load_excluded_filenames(submissions_dir_name: str) -> list[str]: # helper function for hashing all files def load_excluded_filenames(submissions_dir_name: str) -> list[str]: # helper function for hashing all files
@@ -31,10 +31,13 @@ def get_hashes_in_dir(dir_path: str, excluded_filenames: list = []) -> list: #
for filename in files: for filename in files:
if filename.lower() not in excluded_filenames: # convert to lowercase for comparison with excluded files & do not hash if in the excluded list if filename.lower() not in excluded_filenames: # convert to lowercase for comparison with excluded files & do not hash if in the excluded list
filepath = os.path.join(subdir, filename) filepath = os.path.join(subdir, filename)
with open(filepath, 'rb') as f: if os.path.getsize(filepath) > MIN_FILESIZE_IN_BYTES: # file size more than MIN_FILESIZE_IN_BYTES (as set in settings.py)
filehash = hashlib.sha256(f.read()).hexdigest() with open(filepath, 'rb') as f:
if filehash != 'e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855': # do not include hashes of empty files filehash = hashlib.sha256(f.read()).hexdigest()
#if filehash != 'e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855': # do not include hashes of empty files
hash_list.append({ 'filepath': filepath, 'filename': filename, 'sha256 hash': filehash}) hash_list.append({ 'filepath': filepath, 'filename': filename, 'sha256 hash': filehash})
# else:
# print(f'size: {os.path.getsize(filepath)}B, {filepath}')
return hash_list return hash_list
def generate_hashes_gradebook(gradebook_dir_path: str) -> str: # main function for hashing all files in gradebook def generate_hashes_gradebook(gradebook_dir_path: str) -> str: # main function for hashing all files in gradebook