bump to version 0.4.2

[jakevdp]

No description provided.

[conda-forge-linter]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe) and found it was in an excellent condition.

[jakevdp]

So... travis is failing because Python 2.7 apparently uses yield from in its own site-packages:

Compiling /Users/travis/miniconda3/envs/_build/lib/python2.7/site-packages/pexpect/async.py ...
  File "/Users/travis/miniconda3/envs/_build/lib/python2.7/site-packages/pexpect/async.py", line 16
    transport, pw = yield from asyncio.get_event_loop()\
                             ^
SyntaxError: invalid syntax

That can't be good...

[jakevdp]

Looks like it's from the pexpect package, in the pexpect/async.py module. Looking at the code, it's designed so that the file should never be imported in the normal use of the package (imports are conditional on asyncio existing, which is a python 3-only module). But it appears some part of the conda build process is automatically compiling all .py files in the package, and in the course of that runs into a syntax error on yield from.

@jakirkam – any ideas here?

[jakevdp]

Note this is also caught in the conda-build process for pexpect, but for some reason the error is ignored there: https://travis-ci.org/conda-forge/pexpect-feedstock/jobs/148997040#L516

[ocefpaf]

Jake rerendering the feedstock might fix this b/c it will update
conda-build.

I'll send more info later. Only online on the phone today.

On Aug 12, 2016 11:50 AM, "Jake Vanderplas" notifications@github.com
wrote:

Looks like it's from the pexpect package, in the pexpect/async.py module.
Looking at the code, it's designed so that the file should never be
imported in the normal use of the package (imports are conditional on
asyncio existing, which is a python 3-only module). But it appears some
part of the conda build process is automatically compiling all .py files
in the package, and in the course of that runs into a syntax error on yield
from.

@jakirkam – any ideas here?

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#5 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AA6BLwPYVwJnWKHNcIPnVvRPSZ_VBIToks5qfIhQgaJpZM4JjIX8
.

[jakevdp]

Thanks @ocefpaf – I'm not sure how to re-render the feedstock. I found conda-render, but running it on the recipe here doesn't change anything.

[jakevdp]

(googling "conda re-render feedstock" mostly turns-up conda-forge commit messages)

[ocefpaf]

You need to install conda-smithy and then type conda smithy rerender from the recipe root directory.
Sorry, I was on my cell when I sent that message and I got lazy 😜

[jakevdp]

Thanks!

[ocefpaf]

BTW some background on the issue: old versions of conda-build failed when byte-compiling Python 3 code while in Python 2 env. That broke builds instead of just skipping a feature that would not be available on Python 2.

Up until recently we were patching packages to remove the Python 3 code in that situation but, since conda/conda-build#1146 , we can leave the code untouched.

vega was not at fault here but vega-feedstock was still using an old conda-build and the re-rendering should fix that.

[jakevdp]

We're seeing the same error after the re-render. Any thoughts?

[jakirkham]

There should have been some changes to .travis.yml that were not included in the last commit. Try re-rendering again and make sure everything gets committed.

[jakevdp]

Ah, thanks! Should have done git add . -u rather than git add * -u.

[jakirkham]

No worries. Happens to me sometimes too. Particularly when I add a new file like a patch. 😒

[jakevdp]

Looks like things are fixed. Once the travis build finishes, I'll merge. Thanks for the help, @ocefpaf & @jakirkham!