0.2.0

• Removed google.resumable_media.exceptions module (moved into google.resumable_media.common) (feed0a0)
• Making sure we pass unicode to json.loads() rather than bytes. (#13)

PyPI: https://pypi.org/project/google-resumable-media/0.2.0/
Docs: https://googlecloudplatform.github.io/google-resumable-media-python/0.2.0/

0.1.1

• Renamed constants module as common

• Added RetryStrategy class so individual upload and download instances could have fine-grained control over how they attempt retries.

• Added non-public _retry_strategy attribute (e.g.) to all upload and download classes. For now, this cannot be modified via the public interface however, it can still be used, e.g.

  from google import resumable_media
  from google.resumable_media.requests import SimpleUpload
  
  upload_url = u'https://some-url.invald'
  upload = SimpleUpload(upload_url)
  
  upload._retry_strategy = resumable_media.RetryStrategy(max_retries=10)

PyPI: https://pypi.python.org/pypi/google-resumable-media/0.1.1
Docs: https://googlecloudplatform.github.io/google-resumable-media-python/0.1.1/

0.1.0

• Bumped minor version from 0.0.{micro} to 0.1.{micro}
• Bugfix: Actually passing total_bytes to the low-level _prepare* method in requests.ResumableUpload.initiate()

PyPI: https://pypi.python.org/pypi/google-resumable-media/0.1.0
Docs: https://googlecloudplatform.github.io/google-resumable-media-python/0.1.0/

0.0.5

• Adding optional total_bytes argument to ResumableUpload.initiate(). This can be used to explicitly declare the size of the upload. Until this release, the upload size was determined implicitly by calling stream.seek(0, os.SEEK_END) (i.e. seeking until the end of the stream).
• Adding optional stream_final arguments to ResumableUpload.initiate(). When stream_final=False (and total_bytes is unset), the upload will be started with an unknown size. This can occur, e.g. if the resource is generated on the fly, such as application logs. Such uploads are considered "finished" when the stream is exhausted and at that point the upload will send a content range header indicating completion to the server.

This feature was mentioned in the initial release notes as:

• Support for resumable uploads with size unknown when initiate()-d

PyPI: https://pypi.python.org/pypi/google-resumable-media/0.0.5
Docs: https://googlecloudplatform.github.io/google-resumable-media-python/0.0.5/

0.0.4

• Re-organized implementation from google.resumable_media into google.resumable_media.requests.
• This way, the subpackages google.resumable_media.FOO can explicitly support a given transport library FOO.
• Currently, the only one with explicit support is requests.

This re-org was mentioned in the initial release notes as:

Upload / download classes that explicitly support a particular transport

It's worth noting that the majority of the behavior has been moved into partially abstract classes in the non-public google.resumable_media._download and google.resumable_media._upload modules. This way, support for new transports can be added with little effort.

PyPI: https://pypi.python.org/pypi/google-resumable-media/0.0.4
Docs: https://googlecloudplatform.github.io/google-resumable-media-python/0.0.4/

0.0.3

Added automatic retries for any 429, 500, 502, 503 or 504 error responses.

The retry algorithm will sleep for 1, 2, 4, 8, 16, 32, 64, 64, ... seconds, doubling the wait time for each new failure until a maximum of 64 seconds is reached. (There will be some random jitter added to each sleep time to avoid server overload from multiple clients.) If 10 total minutes of wait time have accrued, there are no more attempts to retry.

This re-org was mentioned in the initial release notes as:

Automatic retries with exponential backoff for retry-able errors (429, 500, 502, 503, 504)

PyPI: https://pypi.python.org/pypi/google-resumable-media/0.0.3
Docs: https://googlecloudplatform.github.io/google-resumable-media-python/0.0.3/

0.0.2

Two changes have been added between releases

• Adding optional headers to all upload and download classes. These headers are sent in addition to the required headers for uploading, downloading and initiating requests
• Adding ChunkedDownload.invalid and making download instances invalid on some failed calls to consume_next_chunk

PyPI: https://pypi.python.org/pypi/google-resumable-media/0.0.2
Docs: https://googlecloudplatform.github.io/google-resumable-media-python/0.0.2/

0.0.1

Initial release. Five classes provided as core interface:

• resumable_media.Download
• resumable_media.ChunkedDownload
• resumable_media.SimpleUpload
• resumable_media.MultipartUpload
• resumable_media.ResumableUpload

and a single exception type for failures:

• resumable_media.InvalidResponse

Hop on the landing page for an overview of how to use each of the provided types.

Features missing from this release:

• Automatic retries with exponential backoff for retry-able errors (429, 500, 502, 503, 504)
• Upload / download classes that explicitly support a particular transport (e.g. requests)
• Support for resumable uploads with size unknown when initiate()-d

PyPI: https://pypi.python.org/pypi/google-resumable-media/0.0.1
Docs: https://googlecloudplatform.github.io/google-resumable-media-python/0.0.1/