Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,9 @@
from django.contrib.sites.models import Site
from django.core import mail
from django.core.cache import cache
from django.db import connection
from django.test import TestCase
from django.test.utils import CaptureQueriesContext
from django.urls import reverse
from enterprise.models import (
EnterpriseCourseEnrollment,
Expand Down Expand Up @@ -1078,9 +1080,86 @@ def cleanup_and_assert_status(self, data=None, expected_status=status.HTTP_204_N
assert response.status_code == expected_status
return response

def test_simple_success(self):
self.cleanup_and_assert_status()
assert not UserRetirementStatus.objects.all()
def _assert_redacted_update_delete_queries(self, queries, redacted_username, redacted_email, redacted_name):
"""
Helper method to verify UPDATE and DELETE queries contain correct field-value assignments.

Args:
queries: List of captured query dicts from CaptureQueriesContext
redacted_username: Expected redacted username value
redacted_email: Expected redacted email value
redacted_name: Expected redacted name value
"""
update_queries = [q for q in queries if 'UPDATE' in q['sql'] and 'user_api_userretirementstatus' in q['sql']]
delete_queries = [q for q in queries if 'DELETE' in q['sql'] and 'user_api_userretirementstatus' in q['sql']]

# Should have exactly 1 bulk UPDATE and 1 bulk DELETE query (not individual per-record queries)
assert len(update_queries) == 1, f"Expected 1 UPDATE query, found {len(update_queries)}"
assert len(delete_queries) == 1, f"Expected 1 DELETE query, found {len(delete_queries)}"

# Verify UPDATE query redacts records with the correct field-value assignments
for update_query in update_queries:
sql_lower = update_query['sql']
# Ensure original_username, original_email, and original_name are set to redacted values
assert "original_username" in sql_lower and f"= '{redacted_username}'" in sql_lower, (
f"UPDATE query missing 'original_username = {redacted_username}': {sql_lower}"
)
assert "original_email" in sql_lower and f"= '{redacted_email}'" in sql_lower, (
f"UPDATE query missing 'original_email = {redacted_email}': {sql_lower}"
)
assert "original_name" in sql_lower and f"= '{redacted_name}'" in sql_lower, (
f"UPDATE query missing 'original_name = {redacted_name}': {sql_lower}"
)

# Verify DELETE query is a single bulk operation (uses ID-based deletion after the update)
# Django optimizes by deleting the records using their IDs rather than re-filtering
for delete_query in delete_queries:
sql_lower = delete_query['sql']
# Ensure it's a single DELETE statement for the UserRetirementStatus table
assert 'DELETE FROM' in sql_lower and 'user_api_userretirementstatus' in sql_lower, (
f"DELETE query doesn't match expected pattern: {sql_lower}"
)

def test_default_redacted_values(self):
"""
Test basic cleanup with default redacted values.
Verify that redaction (UPDATE) happens before deletion (DELETE).
Captures actual SQL queries to ensure UPDATE queries contain correct field-value assignments.
"""
with CaptureQueriesContext(connection) as context:
self.cleanup_and_assert_status()

# Verify records are deleted after redaction
retirements = UserRetirementStatus.objects.all()
assert retirements.count() == 0

# Verify UPDATE and DELETE queries with default 'redacted' value
self._assert_redacted_update_delete_queries(context.captured_queries, 'redacted', 'redacted', 'redacted')

def test_custom_redacted_values(self):
"""Test that custom redacted values are applied before deletion."""
custom_username = 'username-redacted-12345'
custom_email = 'email-redacted-67890'
custom_name = 'name-redacted-abcde'

data = {
'usernames': self.usernames,
'redacted_username': custom_username,
'redacted_email': custom_email,
'redacted_name': custom_name
}

with CaptureQueriesContext(connection) as context:
self.cleanup_and_assert_status(data=data)

# Verify records are deleted after redaction
retirements = UserRetirementStatus.objects.all()
assert retirements.count() == 0

# Verify UPDATE and DELETE queries with custom redacted values
self._assert_redacted_update_delete_queries(
context.captured_queries, custom_username, custom_email, custom_name
)

def test_leaves_other_users(self):
remaining_usernames = []
Expand Down
26 changes: 23 additions & 3 deletions openedx/core/djangoapps/user_api/accounts/views.py
Original file line number Diff line number Diff line change
Expand Up @@ -1024,14 +1024,20 @@ def cleanup(self, request):

```
{
'usernames': ['user1', 'user2', ...]
'usernames': ['user1', 'user2', ...],
'redacted_username': 'Value to store in username field',
'redacted_email': 'Value to store in email field',
'redacted_name': 'Value to store in name field'
}
```

Deletes a batch of retirement requests by username.
Redacts a batch of retirement requests by redacting PII fields.
"""
try:
usernames = request.data["usernames"]
redacted_username = request.data.get("redacted_username", "redacted")
redacted_email = request.data.get("redacted_email", "redacted")
redacted_name = request.data.get("redacted_name", "redacted")

if not isinstance(usernames, list):
raise TypeError("Usernames should be an array.")
Expand All @@ -1045,7 +1051,21 @@ def cleanup(self, request):
if len(usernames) != len(retirements):
raise UserRetirementStatus.DoesNotExist("Not all usernames exist in the COMPLETE state.")

retirements.delete()
# Redact PII fields first, then delete. In case an ETL tool is syncing data
# to a downstream data warehouse, and treats the deletes as soft-deletes,
# the data will have first been redacted, protecting the sensitive PII.
retirements.update(
original_username=redacted_username,
original_email=redacted_email,
original_name=redacted_name
)
UserRetirementStatus.objects.filter(
original_username=redacted_username,
original_email=redacted_email,
original_name=redacted_name,
current_state=complete_state
).delete()

return Response(status=status.HTTP_204_NO_CONTENT)
except (RetirementStateError, UserRetirementStatus.DoesNotExist, TypeError) as exc:
return Response(str(exc), status=status.HTTP_400_BAD_REQUEST)
Expand Down
Loading