Retry logic on 404s in some text-analytics API calls

dariosky · dariosky · commit ed311d0b5332 · 2020-12-09T15:21:57.000+01:00
diff --git a/ravenpackapi/core.py b/ravenpackapi/core.py
@@ -2,7 +2,7 @@
 import os
 
 from ravenpackapi import Dataset
-from ravenpackapi.exceptions import APIException
+from ravenpackapi.exceptions import get_exception
 from ravenpackapi.models.dataset_list import DatasetList
 from ravenpackapi.models.job import Job
 from ravenpackapi.models.mapping import RPMappingResults
@@ -15,7 +15,7 @@
 from ravenpackapi.utils.dynamic_sessions import DynamicSession
 
 _VALID_METHODS = ('get', 'post', 'put', 'delete', 'patch')
-VERSION = '1.0.43'
+VERSION = '1.0.44'
 
 logger = logging.getLogger("ravenpack.core")
 
@@ -88,11 +88,7 @@ def request(self, endpoint, data=None, json=None,
             logger.info("API query to %s" % to_curl(response.request))
         if except_on_fail and response.status_code != 200:
             logger.error("Error calling the API, we tried: %s" % to_curl(response.request))
-            raise APIException(
-                'Got an error {status}: body was \'{error_message}\''.format(
-                    status=response.status_code,
-                    error_message=response.text
-                ), response=response)
+            raise get_exception(response)
         return response
 
     def list_datasets(self, scope=None, tags=None):
diff --git a/ravenpackapi/exceptions.py b/ravenpackapi/exceptions.py
@@ -3,13 +3,31 @@
 import requests
 
 
+def get_exception(response):
+    """ Return an appropriate APIException given an API response """
+    if response.status_code == 404:
+        Exception_class = APIException404
+    else:
+        Exception_class = APIException
+
+    return Exception_class(
+        'Got an error {status}: body was \'{error_message}\''.format(
+            status=response.status_code,
+            error_message=response.text
+        ), response=response)
+
+
 class APIException(Exception):
     def __init__(self, *args, **kwargs):
         response = kwargs.pop('response', None)
         super(APIException, self).__init__(*args)
         self.response = response
 
 
+class APIException404(APIException):
+    pass
+
+
 class MissingAPIException(Exception):
     def __str__(self):
         return "Define the api parameter with your own APIKEY " \
diff --git a/ravenpackapi/tests/test_job_cancellation.py b/ravenpackapi/tests/test_job_cancellation.py
diff --git a/ravenpackapi/upload/models.py b/ravenpackapi/upload/models.py
@@ -1,6 +1,7 @@
 from time import sleep
 
 from ravenpackapi.exceptions import api_method
+from ravenpackapi.upload.upload_utils import retry_404
 
 FILE_FIELDS = (
     'file_id', 'file_name', 'folder_id',
@@ -64,33 +65,36 @@ def get_metadata(self, force_refresh=False):
 
     @api_method
     def save_original(self, filename):
-        response = self.api.request('%s/files/%s' % (self.api._UPLOAD_BASE_URL, self.file_id),
-                                    stream=True)
+        response = retry_404(self.api.request,
+                             '%s/files/%s' % (self.api._UPLOAD_BASE_URL, self.file_id),
+                             stream=True)
         with open(filename, 'wb') as f:
             for chunk in response.iter_content(chunk_size=self.api._CHUNK_SIZE):
                 f.write(chunk)
 
     @api_method
     def save_analytics(self, filename, output_format='application/json'):
         self.wait_for_completion()
-        response = self.api.request('%s/files/%s/analytics' % (self.api._UPLOAD_BASE_URL, self.file_id,),
-                                    headers=dict(
-                                        Accept=output_format,
-                                        **self.api.headers
-                                    ),
-                                    stream=True)
+        response = retry_404(self.api.request,
+                             '%s/files/%s/analytics' % (self.api._UPLOAD_BASE_URL, self.file_id,),
+                             headers=dict(
+                                 Accept=output_format,
+                                 **self.api.headers
+                             ),
+                             stream=True)
         with open(filename, 'wb') as f:
             for chunk in response.iter_content(chunk_size=self.api._CHUNK_SIZE):
                 f.write(chunk)
 
     @api_method
     def get_analytics(self, output_format='application/json'):
         self.wait_for_completion()
-        response = self.api.request('%s/files/%s/analytics' % (self.api._UPLOAD_BASE_URL, self.file_id,),
-                                    headers=dict(
-                                        Accept=output_format,
-                                        **self.api.headers
-                                    ))
+        response = retry_404(self.api.request,
+                             '%s/files/%s/analytics' % (self.api._UPLOAD_BASE_URL, self.file_id,),
+                             headers=dict(
+                                 Accept=output_format,
+                                 **self.api.headers
+                             ))
         if output_format == 'application/json':
             return response.json()
         else:
@@ -99,22 +103,27 @@ def get_analytics(self, output_format='application/json'):
     @api_method
     def save_annotated(self, filename):
         self.wait_for_completion()
-        response = self.api.request('%s/files/%s/annotated' % (self.api._UPLOAD_BASE_URL, self.file_id),
-                                    stream=True)
+        response = retry_404(self.api.request,
+                             '%s/files/%s/annotated' % (self.api._UPLOAD_BASE_URL, self.file_id),
+                             stream=True)
         with open(filename, 'wb') as f:
             for chunk in response.iter_content(chunk_size=self.api._CHUNK_SIZE):
                 f.write(chunk)
 
     @api_method
     def get_annotated(self):
         self.wait_for_completion()
-        response = self.api.request('%s/files/%s/annotated' % (self.api._UPLOAD_BASE_URL, self.file_id))
+        response = retry_404(self.api.request,
+                             '%s/files/%s/annotated' % (self.api._UPLOAD_BASE_URL, self.file_id)
+                             )
         return response.text
 
     @api_method
     def delete(self):
-        response = self.api.request('%s/files/%s' % (self.api._UPLOAD_BASE_URL, self.file_id),
-                                    method='delete')
+        response = retry_404(self.api.request,
+                             '%s/files/%s' % (self.api._UPLOAD_BASE_URL, self.file_id),
+                             method='delete'
+                             )
         return response
 
     @api_method
diff --git a/ravenpackapi/upload/tests/test_text_analytics.py b/ravenpackapi/upload/tests/test_text_analytics.py
@@ -0,0 +1,16 @@
+import os
+
+from ravenpackapi import RPApi
+
+
+class TestRecentAnalyticsRetried:
+    api = RPApi()
+
+    def test_upload_delete_retry(self):
+        """ When we delete immediately after creation we get a 404
+            The API should silently retry for some time
+        """
+        api = self.api
+        filename = "upload_sample.txt"
+        f = api.upload.file(os.path.join(os.path.dirname(__file__), filename))
+        f.delete()
diff --git a/ravenpackapi/upload/tests/upload_sample.txt b/ravenpackapi/upload/tests/upload_sample.txt
@@ -0,0 +1,2 @@
+Apple Inc.
+IBM
diff --git a/ravenpackapi/upload/upload_utils.py b/ravenpackapi/upload/upload_utils.py
@@ -0,0 +1,15 @@
+import logging
+
+from retry.api import retry_call
+
+from ravenpackapi.exceptions import APIException404
+
+logger = logging.getLogger('ravenpack.upload.retry')
+
+
+def retry_404(func, *args, **kwargs):
+    response = retry_call(func, fargs=args, fkwargs=kwargs,
+                          exceptions=(APIException404,),
+                          delay=3, backoff=2, jitter=(1, 5), tries=10,
+                          logger=logger)
+    return response
diff --git a/requirements.txt b/requirements.txt
@@ -2,3 +2,4 @@ requests[security]
 future
 six
 python-dateutil
+retry
diff --git a/setup.py b/setup.py
@@ -1,6 +1,6 @@
 from setuptools import setup, find_packages
 
-VERSION = '1.0.43'
+VERSION = '1.0.44'
 
 with open('README.rst') as readme_file:
     readme = readme_file.read()
@@ -35,5 +35,5 @@
     ],
 
     keywords='python analytics api rest news data',
-    install_requires=['requests[security]', 'future', 'python-dateutil', 'six'],
+    install_requires=['requests[security]', 'future', 'python-dateutil', 'six', 'retry'],
 )

-Original file line number
+Diff line change
 future
 six
 python-dateutil
 +retry