amorfo77/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2026-09-27 20:00:31 +00:00
Code Issues Projects Releases Wiki Activity

Security: validate remote OCR endpoint against internal SSRF (#13897)

Browse Source 
* Security: validate remote OCR endpoint against internal SSRF

Adds PAPERLESS_REMOTE_OCR_ALLOW_INTERNAL_ENDPOINTS (default true)
and validates remote_ocr_endpoint via validate_outbound_http_url
on the config serializer, matching the existing LLM endpoint handling.

* Validates te outbound url again right before use

* cover empty-value branch of validate_remote_ocr_endpoint because coverage

* re-validate remote OCR endpoint on every outbound request
This commit is contained in:
Trenton H
2026-09-01 20:22:10 +00:00
committed by GitHub
parent 5c5b1ee6b5
commit d78754bff1
5 changed files with 140 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/documents/tests/test_api_app_config.py
+76
View File
@@ -1063,3 +1063,79 @@ class TestApiAppConfig(DirectoriesMixin, APITestCase):
)
self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
self.assertIn("non-public address", str(response.data).lower())
@override_settings(REMOTE_OCR_ALLOW_INTERNAL_ENDPOINTS=False)
def test_update_remote_ocr_endpoint_blocks_internal_endpoint_when_disallowed(
self,
) -> None:
"""
GIVEN:
- Internal remote OCR endpoints are disallowed
WHEN:
- The config is updated with a remote OCR endpoint resolving internally
THEN:
- The request is rejected
"""
response = self.client.patch(
f"{self.ENDPOINT}1/",
json.dumps(
{
"remote_ocr_endpoint": "http://127.0.0.1:5000",
},
),
content_type="application/json",
)
self.assertEqual(response.status_code, status.HTTP_400_BAD_REQUEST)
self.assertIn("non-public address", str(response.data).lower())
@override_settings(REMOTE_OCR_ALLOW_INTERNAL_ENDPOINTS=True)
def test_update_remote_ocr_endpoint_allows_internal_endpoint_by_default(
self,
) -> None:
"""
GIVEN:
- Internal remote OCR endpoints are allowed (the default)
WHEN:
- The config is updated with a remote OCR endpoint resolving internally
THEN:
- The request is accepted, preserving existing self-hosted deployments
"""
response = self.client.patch(
f"{self.ENDPOINT}1/",
json.dumps(
{
"remote_ocr_endpoint": "http://127.0.0.1:5000",
},
),
content_type="application/json",
)
self.assertEqual(response.status_code, status.HTTP_200_OK)
self.assertEqual(
response.data["remote_ocr_endpoint"],
"http://127.0.0.1:5000",
)
@override_settings(REMOTE_OCR_ALLOW_INTERNAL_ENDPOINTS=False)
def test_update_remote_ocr_endpoint_empty_value_skips_validation(
self,
) -> None:
"""
GIVEN:
- Internal remote OCR endpoints are disallowed
WHEN:
- The config is updated with an empty remote OCR endpoint
THEN:
- The request is accepted; clearing the field never needs
outbound URL validation
"""
response = self.client.patch(
f"{self.ENDPOINT}1/",
json.dumps(
{
"remote_ocr_endpoint": "",
},
),
content_type="application/json",
)
self.assertEqual(response.status_code, status.HTTP_200_OK)
self.assertEqual(response.data["remote_ocr_endpoint"], "")