[TIP] weird coverage problems with subprocesses - one back, both still to go :-(

Chris Withers chris at simplistix.co.uk
Thu May 28 04:26:11 PDT 2015


> On 28 May 2015, at 12:22, Ned Batchelder <ned at nedbatchelder.com> wrote:
> 
>> On 5/28/15 5:00 AM, Chris Withers wrote:
>>> On 28/05/2015 09:05, Chris Withers wrote:
>>>> On 28/05/2015 07:47, Chris Withers wrote:
>>>> When running the tests with nose locally, I get lots of:
>>>> 
>>>> nosetests --with-cov --cov=picky
>>>> 
>>>> 'picky.handlers',
>>>>  'ERROR',
>>>>  'pip gave errors: Error processing line 1 of /Users/chris/.local/lib/python2.7/site-packages/init_cov_core.pth:\n\n Traceback (most recent call last):\n    File "/Users/chris/anaconda/envs/picky27/lib/python2.7/site.py", line 161, in addpackage\n      exec line\n    File "<string>", line 1, in <module>\n  ImportError: No module named cov_core_init\n\nRemainder of file ignored\n'),
>>> Well, my guess is this had to do with cov-core being installed in the root conda rather than a conda environment.
>>> I've uninstalled and manually deleted the file; the problem is I have a feeling that I manually deleted that file before and it came back :-S
>> 
>> yep, I created a new conda environment, installed coverage using the pip from the new environment, and hey presto, the file is back. Why is coverage (cov-core?) picking /Users/chris/.local/lib/python2.7/site-packages/init_cov_core.pth to put this path rather than somewhere inside the conda environment?
> This sounds like a complicated configuration.  Any chance you can provide detailed reproduction instructions?

This should cover it:

http://picky.readthedocs.org/en/latest/development.html#setting-up-the-environment

Cheers,

Chris


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.idyll.org/pipermail/testing-in-python/attachments/20150528/ec11b402/attachment.html>


More information about the testing-in-python mailing list