Ensure that preview, thumbnail lambdas handle Glacier-style 403s (#1789)

Author: akarve

lambdas/preview/index.py

@@ -153,12 +153,13 @@ def lambda_handler(request):
             'info': info,
             'html': html,
         }
-
     else:
         ret_val = {
-            'error': resp.reason
+            'error': resp.reason,
+            'text': resp.text,
         }
-    return make_json_response(200, ret_val)
+
+    return make_json_response(resp.status_code, ret_val)
 
 
 def extract_csv(head, separator):

lambdas/preview/test/test_index.py

@@ -5,11 +5,10 @@
 import os
 from pathlib import Path
 import re
+from unittest.mock import ANY, patch
 
 import pyarrow.parquet as pq
 import responses
-from unittest.mock import ANY, patch
-
 
 from t4_lambda_shared.utils import read_body
 from .. import index
@@ -41,6 +40,22 @@ def _make_event(cls, query, headers=None):
             'isBase64Encoded': False,
         }
 
+    @responses.activate
+    def test_403(self):
+        """test 403 cases, such as Glacier"""
+        url = self.FILE_URL
+        responses.add(
+            responses.GET,
+            url=url,
+            status=403,
+        )
+        event = self._make_event({'url': url, 'input': 'txt'})
+        response = index.lambda_handler(event, None)
+        assert response["statusCode"] == 403
+        body = json.loads(response["body"])
+        assert "text" in body
+        assert "error" in body
+
     @responses.activate
     def test_fcs(self):
         """test fcs extraction

lambdas/thumbnail/index.py

@@ -342,6 +342,7 @@ def lambda_handler(request):
 
     # Errored, return error code
     ret_val = {
-        'error': resp.reason
+        'error': resp.reason,
+        'text': resp.text,
     }
     return make_json_response(resp.status_code, ret_val)

lambdas/thumbnail/requirements.txt

@@ -10,8 +10,8 @@ pdf2image==1.13.1
 Pillow==6.2.2
 psutil==5.7.0
 pyrsistent==0.14.11
-requests==2.21.0
+requests==2.24.0
 scipy==1.3.0
 tifffile==0.15.1
 six==1.12.0
-urllib3==1.24.2
+urllib3==1.25.10

lambdas/thumbnail/test-requirements.txt

@@ -5,4 +5,4 @@ py==1.8.0
 pytest==4.3.0
 pytest-cov==2.6.1
 pytest-raises==0.10
-responses==0.10.5
+responses==0.12.0

lambdas/thumbnail/tests/test_thumbnail.py

@@ -16,6 +16,16 @@
 from ..index import lambda_handler, set_pdf_env
 
 
+HEADER_403 = {
+    'x-amz-request-id': 'guid123',
+    'x-amz-id-2': 'some/dat/here/+xxxxx+=',
+    'Content-Type': 'application/xml',
+    'Transfer-Encoding': 'chunked',
+    'Date': 'Tue, 08 Sep 2020 00:01:06 GMT',
+    'Server': 'AmazonS3'
+}
+
+
 @pytest.fixture
 def data_dir():
     return Path(__file__).parent / 'data'
@@ -33,6 +43,28 @@ def _make_event(query, headers=None):
     }
 
 
+@responses.activate
+def test_403():
+    """test 403 cases, such as Glacier"""
+    url = "https://example.com/folder/file.ext"
+    responses.add(
+        responses.GET,
+        url=url,
+        status=403,
+        headers=HEADER_403,
+    )
+    params = {
+        "size": "w32h32"
+    }
+    event = _make_event({"url": url, **params})
+    # Get the response
+    response = lambda_handler(event, None)
+    assert response["statusCode"] == 403
+    body = json.loads(response["body"])
+    assert "text" in body
+    assert "error" in body
+
+
 @patch.dict(os.environ, {
     'LAMBDA_TASK_ROOT': str(Path('/var/task')),
     # need the real OS path so that we can find pdftoppm, but patch to avoid side-effects