Skip to content

Latest commit

 

History

History
102 lines (75 loc) · 3.03 KB

README.md

File metadata and controls

102 lines (75 loc) · 3.03 KB

sux

Python 2 compatibility layer for Python 3

For those times when you really need to use a non-python-3-compliant module but don't want to downgrade your whole system to python 2.

Are you crazy?

Yes, a little, but this actually works!

How could this possibly work?

Simply install your python2 packages as you would normally, in a python2 (virtual) environment. Then, in your python3 space, you will install and invoke sux, and it will provide a "tunnel" through to python2, relaying your interactions between the two environments.

Quickstart

In a python3 (virtual) env:

pip3 install sux

Create your python2 virtualenv and install the naughty package(s):

export PY2_VIRTUAL_ENV="/tmp/sux-to-use-python2"
virtualenv -ppython2 "$PY2_VIRTUAL_ENV"
"$PY2_VIRTUAL_ENV"/bin/pip install boto

Now, with PY2_VIRTUAL_ENV still set, fire up python3 and get to work:

import sux
connection = sux.to_use('boto.s3.connection')
conn = connection.S3Connection('abc', '123')
conn.create_bucket("foo")

The final command will attempt to connect to S3, and will raise sux.exceptions.S3ResponseError.

Note that the to_use operator expects a package name to be passed. e.g. with the pyxero package, you might choose to do the following:

PublicCredentials = sux.to_use('xero.auth').PublicCredentials
Xero = sux.to_use('xero').Xero

where in a native python2 environment you would have instead used:

from xero.auth import PublicCredentials
from xero import Xero

Tell me More!

CI status

What works?

  • Importing a python2 package
  • Accessing attributes in that package (recursively)
  • Invoking functions and methods on the python2 objects
  • Handling exceptions
  • Garbage collection

What doesn't (yet) work?

  • Simultaneous operations from multiple threads.
  • Passing python3 objects to python2 space. (e.g. registering a 'hook' function with a python2 object)
  • Accessing the special class methods starting with double-underscore.

Is it thread-safe?

Yes. However, only one thread can be operating on a python2 object at a time. So if you make a long-running call via sux, if another thread attempts to use a python2 object, it will block.

How are exceptions handled?

Any exception raised in the python2 environment will cause an exception of the same name (to be raised in the python3 space. See tests.test_boto for an example.

I want to control the environment of the python2 instance!

OK:

bf = sux.to_use('bioformats', env=os.environ)

or, if you are not afraid of accessing private attributes:

bf = sux.to_use('bioformats', env=os.environ.__dict__['_data'])

Is this fast?

Every interaction between python2 and python3 uses pickle for transport and is not optimised for speed. Then again, the python2 instance runs in a separate process, with a separate GIL.