[Storage][Blob][Fix]fix get_block_list bug (#18751)

#16314
Azure · Jun 28, 2021 · ee0af9b · ee0af9b
1 parent 88a742c
commit ee0af9b
Show file tree

Hide file tree

Showing 3 changed files with 22 additions and 5 deletions.
diff --git a/sdk/storage/azure-storage-blob/azure/storage/blob/_models.py b/sdk/storage/azure-storage-blob/azure/storage/blob/_models.py
@@ -12,7 +12,7 @@
 from azure.core.exceptions import HttpResponseError
 from ._generated.models import ArrowField
 
-from ._shared import decode_base64_to_text
+from ._shared import decode_base64_to_bytes
 from ._shared.response_handlers import return_context_and_deserialized, process_storage_error
 from ._shared.models import DictMixin, get_enum_value
 from ._generated.models import Logging as GeneratedLogging
@@ -739,7 +739,15 @@ def __init__(self, block_id, state=BlockState.Latest):
 
     @classmethod
     def _from_generated(cls, generated):
-        block = cls(decode_base64_to_text(generated.name))
+        try:
+            decoded_bytes = decode_base64_to_bytes(generated.name)
+            block_id = decoded_bytes.decode('utf-8')
+        # this is to fix a bug. When large blocks are uploaded through upload_blob the block id isn't base64 encoded
+        # while service expected block id is base64 encoded, so when we get block_id if we cannot base64 decode, it
+        # means we didn't base64 encode it when stage the block, we want to use the returned block_id directly.
+        except UnicodeDecodeError:
+            block_id = generated.name
+        block = cls(block_id)
         block.size = generated.size
         return block
 

diff --git a/sdk/storage/azure-storage-blob/tests/test_large_block_blob.py b/sdk/storage/azure-storage-blob/tests/test_large_block_blob.py
@@ -51,7 +51,10 @@ def _setup(self, storage_account, key):
         self.container_name = self.get_resource_name('utcontainer')
 
         if self.is_live:
-            self.bsc.create_container(self.container_name)
+            try:
+                self.bsc.create_container(self.container_name)
+            except:
+                pass
 
     def _teardown(self, file_name):
         if path.isfile(file_name):
@@ -170,9 +173,12 @@ def test_create_large_blob_from_path(self, resource_group, location, storage_acc
 
         # Act
         with open(FILE_PATH, 'rb') as stream:
-            blob.upload_blob(stream, max_concurrency=2)
+            blob.upload_blob(stream, max_concurrency=2, overwrite=True)
+
+        block_list = blob.get_block_list()
 
         # Assert
+        self.assertIsNot(len(block_list), 0)
         self.assertBlobEqual(self.container_name, blob_name, data)
         self._teardown(FILE_PATH)
 

diff --git a/sdk/storage/azure-storage-blob/tests/test_large_block_blob_async.py b/sdk/storage/azure-storage-blob/tests/test_large_block_blob_async.py
@@ -199,9 +199,12 @@ async def test_create_large_blob_from_path_async(self, resource_group, location,
         # Act
         try:
             with open(FILE_PATH, 'rb') as stream:
-                await blob.upload_blob(stream, max_concurrency=2)
+                await blob.upload_blob(stream, max_concurrency=2, overwrite=True)
+
+            block_list = await blob.get_block_list()
 
             # Assert
+            self.assertIsNot(len(block_list), 0)
             await self.assertBlobEqual(self.container_name, blob_name, data)
         finally:
             self._teardown(FILE_PATH)