feat: Fix closing file after chunked upload (#761)

Closes: SDK-2611

Author: lukaszsocha2

boxsdk/object/folder.py

@@ -45,6 +45,7 @@ class _CollaborationType(TextEnum):
 
 class _Collaborator:
     """This helper class represents a collaborator on Box. A Collaborator can be a User, Group, or an email address"""
+
     def __init__(self, collaborator: Any):
         if isinstance(collaborator, User):
             self._setup(user=collaborator)
@@ -138,21 +139,29 @@ def create_upload_session(self, file_size: int, file_name: str) -> 'UploadSessio
         )
 
     @api_call
-    def get_chunked_uploader(self, file_path: str) -> 'ChunkedUploader':
+    def get_chunked_uploader(self, file_path: str, file_name: Optional[str] = None) -> 'ChunkedUploader':
         # pylint: disable=consider-using-with
         """
         Instantiate the chunked upload instance and create upload session with path to file.
 
         :param file_path:
             The local path to the file you wish to upload.
+         :param file_name:
+            The name with extention of the file that will be uploaded, e.g. new_file_name.zip.
+            If not specified, the name from the local system is used.
         :returns:
             A :class:`ChunkedUploader` object.
         """
         total_size = os.stat(file_path).st_size
+        upload_file_name = file_name if file_name else os.path.basename(file_path)
         content_stream = open(file_path, 'rb')
-        file_name = os.path.basename(file_path)
-        upload_session = self.create_upload_session(total_size, file_name)
-        return upload_session.get_chunked_uploader_for_stream(content_stream, total_size)
+
+        try:
+            upload_session = self.create_upload_session(total_size, upload_file_name)
+            return upload_session.get_chunked_uploader_for_stream(content_stream, total_size)
+        except Exception:
+            content_stream.close()
+            raise
 
     def _get_accelerator_upload_url_fow_new_uploads(self) -> Optional[str]:
         """
@@ -310,7 +319,9 @@ def upload_stream(
             headers['Content-MD5'] = sha1
         if not headers:
             headers = None
-        file_response = self._session.post(url, data=data, files=files, expect_json_response=False, headers=headers).json()
+        file_response = self._session.post(
+            url, data=data, files=files, expect_json_response=False, headers=headers
+        ).json()
         if 'entries' in file_response:
             file_response = file_response['entries'][0]
         return self.translator.translate(

boxsdk/util/chunked_uploader.py

@@ -9,7 +9,7 @@
 
 class ChunkedUploader:
 
-    def __init__(self, upload_session: 'UploadSession', content_stream: IO, file_size: int):
+    def __init__(self, upload_session: 'UploadSession', content_stream: IO[bytes], file_size: int):
         """
         The initializer for the :class:`ChunkedUploader`
 
@@ -42,8 +42,7 @@ def start(self) -> Optional['File']:
         if self._is_aborted:
             raise BoxException('The upload has been previously aborted. Please retry upload with a new upload session.')
         self._upload()
-        content_sha1 = self._sha1.digest()
-        return self._upload_session.commit(content_sha1=content_sha1, parts=self._part_array)
+        return self._commit_and_erase_stream_reference_when_succeed()
 
     def resume(self) -> Optional['File']:
         """
@@ -68,8 +67,7 @@ def resume(self) -> Optional['File']:
                 self._inflight_part = None
             self._part_definitions[part['offset']] = part
         self._upload()
-        content_sha1 = self._sha1.digest()
-        return self._upload_session.commit(content_sha1=content_sha1, parts=self._part_array)
+        return self._commit_and_erase_stream_reference_when_succeed()
 
     def abort(self) -> bool:
         """
@@ -86,7 +84,7 @@ def abort(self) -> bool:
 
     def _upload(self) -> None:
         """
-        Utility function for looping through all parts of of the upload session and uploading them.
+        Utility function for looping through all parts of the upload session and uploading them.
         """
         while len(self._part_array) < self._upload_session.total_parts:
             # Retrieve the part inflight if it exists, if it does not exist then get the next part from the stream.
@@ -124,6 +122,14 @@ def _get_next_part(self) -> 'InflightPart':
             copied_length += len(bytes_read)
         return InflightPart(offset, chunk, self._upload_session, self._file_size)
 
+    def _commit_and_erase_stream_reference_when_succeed(self):
+        content_sha1 = self._sha1.digest()
+        commit_result = self._upload_session.commit(content_sha1=content_sha1, parts=self._part_array)
+        # Remove file stream reference when uploading file succeeded
+        if commit_result is not None:
+            self._content_stream = None
+        return commit_result
+
 
 class InflightPart:
 

docs/usage/files.md

@@ -17,6 +17,7 @@ file's contents, upload new versions, and perform other common file operations
   - [Automatic Uploader](#automatic-uploader)
     - [Upload new file](#upload-new-file)
     - [Upload new file version](#upload-new-file-version)
+    - [Preflight check before upload](#preflight-check-before-upload)
     - [Resume Upload](#resume-upload)
     - [Abort Chunked Upload](#abort-chunked-upload)
   - [Manual Process](#manual-process)
@@ -200,39 +201,39 @@ without aborting the entire upload, and failed parts can then be retried.
 
 #### Upload new file
 
-The SDK provides a method of automatically handling a chunked upload. First get a folder you want to upload the file to. Then call [`folder.get_chunked_uploader(file_path, rename_file=False)`][get_chunked_uploader_for_file] to retrieve a [`ChunkedUploader`][chunked_uploader_class] object. Calling the method [`chunked_upload.start()`][start] will kick off the chunked upload process and return the [File][file_class] 
+The SDK provides a method of automatically handling a chunked upload. First get a folder you want to upload the file to.
+Then call [`folder.get_chunked_uploader(file_path, rename_file=False)`][get_chunked_uploader_for_file] to retrieve
+a [`ChunkedUploader`][chunked_uploader_class] object. Calling the method [`chunked_upload.start()`][start] will
+kick off the chunked upload process and return the [File][file_class] 
 object that was uploaded.
 
 <!-- samples x_chunked_uploads automatic -->
 ```python
 # uploads large file to a root folder 
-chunked_uploader = client.folder('0').get_chunked_uploader('/path/to/file')
+chunked_uploader = client.folder('0').get_chunked_uploader(file_path='/path/to/file.txt', file_name='new_name.txt')
 uploaded_file = chunked_uploader.start()
 print(f'File "{uploaded_file.name}" uploaded to Box with file ID {uploaded_file.id}')
 ```
 
-You can also return a [`ChunkedUploader`][chunked_uploader_class] object by creating a [`UploadSession`][upload_session_class] object first 
-and calling the method [`upload_session.get_chunked_upload(file_path)`][get_chunked_uploader] or 
-[`upload_session.get_chunked_uploader_for_stream(content_stream, file_size)`][get_chunked_uploader_for_stream].
-
-```python
-chunked_uploader = client.upload_session('56781').get_chunked_uploader('/path/to/file')
-uploaded_file = chunked_uploader.start()
-print(f'File "{uploaded_file.name}" uploaded to Box with file ID {uploaded_file.id}')
-```
+You can also upload file stream by creating a [`UploadSession`][upload_session_class] first and then calling the 
+method [`upload_session.get_chunked_uploader_for_stream(content_stream, file_size)`][get_chunked_uploader_for_stream].
 
 ```python
 test_file_path = '/path/to/large_file.mp4'
-content_stream = open(test_file_path, 'rb')
-total_size = os.stat(test_file_path).st_size
-chunked_uploader = client.upload_session('56781').get_chunked_uploader_for_stream(content_stream, total_size)
-uploaded_file = chunked_uploader.start()
-print(f'File "{uploaded_file.name}" uploaded to Box with file ID {uploaded_file.id}')
+with open(test_file_path, 'rb') as content_stream:
+  total_size = os.stat(test_file_path).st_size
+  upload_session = client.folder('0').create_upload_session(file_size=total_size, file_name='large_file.mp4')
+  chunked_uploader = upload_session.get_chunked_uploader_for_stream(content_stream=content_stream, file_size=total_size)
+  uploaded_file = chunked_uploader.start()
+  print(f'File "{uploaded_file.name}" uploaded to Box with file ID {uploaded_file.id}')
 ```
 
 #### Upload new file version
 
-To upload a new file version for a large file, first get a file you want to replace. Then call [`file.get_chunked_uploader(file_path)`][get_chunked_uploader_for_version] to retrieve a [`ChunkedUploader`][chunked_uploader_class] object. Calling the method [`chunked_upload.start()`][start] will kick off the chunked upload process and return the updated [File][file_class].
+To upload a new file version for a large file, first get a file you want to replace.
+Then call [`file.get_chunked_uploader(file_path)`][get_chunked_uploader_for_version]
+to retrieve a [`ChunkedUploader`][chunked_uploader_class] object. Calling the method [`chunked_upload.start()`][start]
+will kick off the chunked upload process and return the updated [File][file_class].
 
 <!-- samples x_chunked_uploads automatic_new_version -->
 ```python
@@ -243,13 +244,31 @@ print(f'File "{uploaded_file.name}" uploaded to Box with file ID {uploaded_file.
 # the uploaded_file.id will be the same as 'existing_big_file_id'
 ```
 
+#### Preflight check before upload
+
+To check if a file can be uploaded with given name to a specific folder call 
+[`folder.preflight_check(size, name)`][preflight_check]. If the check did not pass, this method will raise an exception
+including some details on why it did not pass.
+
+<!-- samples x_chunked_uploads automatic_new_version -->
+```python
+file_name = 'large_file.mp4'
+test_file_path = '/path/to/large_file.mp4'
+total_size = os.stat(test_file_path).st_size
+destination_folder_id = '0'
+try:
+  client.folder(destination_folder_id).preflight_check(size=total_size, name=file_name)
+except BoxAPIException as e:
+  print(f'File {file_name} cannot be uploaded to folder with id: {destination_folder_id}. Reason: {e.message}')
+```
+
 [start]: https://box-python-sdk.readthedocs.io/en/latest/boxsdk.object.html#boxsdk.object.chunked_uploader.ChunkedUploader.start
 [chunked_uploader_class]: https://box-python-sdk.readthedocs.io/en/latest/boxsdk.object.html#boxsdk.object.chunked_uploader.ChunkedUploader
 [get_chunked_uploader_for_version]: https://box-python-sdk.readthedocs.io/en/latest/boxsdk.object.html#boxsdk.object.file.File.get_chunked_uploader
 [get_chunked_uploader_for_file]: https://box-python-sdk.readthedocs.io/en/latest/boxsdk.object.html#boxsdk.object.folder.Folder.get_chunked_uploader
 [upload_session_class]: https://box-python-sdk.readthedocs.io/en/latest/boxsdk.object.html#boxsdk.object.upload_session.UploadSession
-[get_chunked_uploader]: https://box-python-sdk.readthedocs.io/en/latest/boxsdk.object.html#boxsdk.object.upload_session.UploadSession.get_chunked_uploader
 [get_chunked_uploader_for_stream]: https://box-python-sdk.readthedocs.io/en/latest/boxsdk.object.html#boxsdk.object.upload_session.UploadSession.get_chunked_uploader_for_stream
+[preflight_check]: https://box-python-sdk.readthedocs.io/en/latest/boxsdk.object.html#boxsdk.object.folder.Folder.preflight_check
 
 #### Resume Upload