Quentin Pradet created LIBCLOUD-916:
---------------------------------------
Summary: s3.upload_object_via_stream no longer accepts unicode
Key: LIBCLOUD-916
URL: https://issues.apache.org/jira/browse/LIBCLOUD-916
Project: Libcloud
Issue Type: Bug
Reporter: Quentin Pradet
Consider this snippet:
{code}
import io
from libcloud.storage.base import Container
from libcloud.storage.drivers.google_storage import GoogleStorageDriver
driver = GoogleStorageDriver(key='replaceme', secret='replaceme')
container = Container(name='container', driver=driver, extra={})
driver.upload_object_via_stream(io.StringIO(' '), container, 'path')
{code}
In libcloud 1.5, upload_object_via_stream was calling
libcloud.utils.files.read_in_chunks who encodes the data in UTF-8. In libcloud
2.0, the unicode data is directly sent to requests which requires a bytes body
when sent in chunks.
Should we work around the issue in libcloud or decide that bytes are required
in the iterator?
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
